R 203_The_Application_of_a_High Speed_Digital_Computer_to_the_Present Day_Air_Traffic_Control_System_Jan52 203 The Application Of A High Speed Digital Computer To Present Day Air Traffic Control System Jan52

R-203_The_Application_of_a_High-Speed_Digital_Computer_to_the_Present-Day_Air_Traffic_Control_System_Jan52 R-203_The_Application_of_a_High-Speed_Digital_Computer_to_the_Present-Day_Air_Traffic_Control_System_Jan52

User Manual: R-203_The_Application_of_a_High-Speed_Digital_Computer_to_the_Present-Day_Air_Traffic_Control_System_Jan52

Open the PDF directly: View PDF .
Page Count: 170

Download
Open PDF In Browser	View PDF

6R-203

THE APPLICATION OF A HIGH-SPEED DIGITAL COMPUTER
TO THE PRESENT.DAY AIR TRAFFIC CONTROL SYSTEM

De R. Israel

Report 11.-203

THE APPLICATION OF A HIGH-SPEED DIGITAL
COMPUTER TO THE PRESENT-DAY AIR TRAFFIC
CONTROL SYSTEM '

David Robinson Israel

DIGITAL COMPUT~ LABORATORY
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Cam.bridge 39, Massachusetts

15, 1952
(The f3 isDate: May 29, 1951)
Reprinted: May 21, 1956
January

Report R-203
FOREWORD

(January 15, 1952)

Because it furnishes an example of the use of a large-scale
digital computer in a control app1ioation of great current importance,
this th~sis report, which has had only limited distribution, is
being issued as'a Digital Computer Laboratory R-series report.
Grateful acknowledgement is made of th~ assistance of W. Gordon
Welchman, formerly of the Division of Industrial Cooperation at
the Massachusetts Institute of Technology, who carefully read the
draft of this thesis and offered numerous valuable 'suggestions for
its improvement.
The cooperation of Professor William K. Linvi11,
thesis supervisor, is also noted and appreciated.
The task of Writing this thesis was greatly simplified b.Y the
patient aid of the personnel of the Boston Air Route Traffic Control
Center.
AUTHOR'S INTRODUCTION TO 1956 REPRINT
This reprint, made almost five years after original submission of
the thesis,- is occasioned by the current revival or interest in expanding and improving our air traffic control system. In 1951, digitai
computers were first beooming available for practioa1 use, and the
possibility of their application to air traffic control ~s highly
speculative*; today the use of such equipment in a new system of
air traffic control seems to be an accepted fact and major questions
now relate chiefly to how soon and how much.
Considerable progress has been made in the past five years in
the design and construotion of digital computers, and extensive
study and experimentation has been carried out in the use of such
machines in real-time control systems. Specifically, the following
developments are significant and should 'be kept in mind while reading
this document:
a)

improvements in computer storage systems, principally those of '
magnetic cores and drums, now make large, high-speed memories
possible at relatively low costso Economw of computer storage
need not be a prinCipal factor in the design of computer
programs for'ai,. traffic control

significant advances have been made in in-out syste~and
technique's for transmi t t ing data to and from computers

the concept of semi-automatic systems using both men and
~ines

has developed significantly. Techniques by which the
men and machines can communicate ~ir~ctly are many-fold and
·-inc1ude. intervention switches; displays us'ing cathode ray tubes.
*This, to a large extent, accounts for the inclusion of qbapter III which
was intended \0 present the fundamental notions or a computer and how it
could be used for non~mathematical problems.

iii
c)

continued 0
Charaotrons, or similar devices; alarms; digital indicators;
high-speed printers g etc»

computer reliability is continually being improved, and the
use of duplexed or dual computer installations offers a
satisfactory solution for system operation on a continuous
basis.

These and other developments, together with those in the fields of
communication and navigation, significantly affect much of "what is
contained in this report~ previous estimates need revision; many
former problems have disappeared; previous possibilities and speculations (see Chapter X) now seem to be realities g and new and
exoiting possibilities have ariseno Nevertheless g a large majority
of the original ideas are valid, and the author hopes that this
report9 now largely historical 9 may serve as a useful background
and as one of the several startini points for the large effort of
system design and development which yet lies ahead.
My thanks is extended to the following pedple who have materially
assisted in the preparation of this document for republication~

Anne Catalano
Barbara Higgins

Frances Lapham ,.
Eleanor Lyon

16 May 1956

ReportR-203

ABSTRACT
THE APPLICATION OF 'A HIGH-SPEED DIGITAL COMPUTER

TO THE

PRES~T

-DAY AIR TRAFFIC CONTROL SYSTEM
by

DAVID ROBINSON ISRAEL
Submitted for'the degree of Master of Science
in the
Depart.ent of Electrical Engin"erinc
on May 29,.. 1951
The general purpose of this thesis has been to consider how
a hi~h-speed digital computer might be used in the m.echanizatio~ ot tlle
present system of air traffic control, without any essential changes in
the philosop~ of the sy~tem or in the .ethods of navigation and 00.munication. The study ha..s beed ade in particular reference to the
Whirlwind Coaputer, a digital machine presently in operation at the
Massachusetts Institute of Technolo&Y'o
·Attention has been focused chi.f~ upon the en-route phases
of traffic control, inasmuch as it is this part of the present system
which appears to be most amenable to mechanization. The general
approach taken is that the most satisfactor.y results and most efficient
system ~ll ~e achieved through careful use of the capabilities of bota
the human. operators and the computero
The main part of the stu~ is concerned with the standardizations in the present procedure's that are n.eessar.y for efficient computer mechanization, and with the construction of flow diagraas for the
fulfillment of the necessary traffic control functions. The pro~. . ot
data. stora.ge within the computer is discus_sed in det.ail, and the thesia
contains a brier outline of the system of co~cation necessar,y for
transmitting da.ta to 8,nd receiving data. frbm the computer. Certain re.arks of a general nature are made in the last chapter as an indication
,of the possible improvements which could be ma.de in the present systea
through the use of a computer.

Report R-203
v
CON~.

TABLE. OF

f!W.
FOREVORD

0 0 0 0 0 0 0 •

ABS'I'RACT" 0

CHAPTER I

0 0 0 •

0 0 0 0 0 •

0 • 0 0 "

0 0 0 0 0 0 0 0 0 0 0 0 0 •

0 0 0 0 0 0 0 0 0 0 0 0 0 "

INTRODUCTIO,.o

0 •

0 0 0 0 0 0 0 0 0 " 0 0 0 0 0 0 0 •

0 0 ·0 0 0 0 0 0

0 0.0.0.. 0 0.. 0 0 0 0 " 0 •

4 ..

0 0 " •

0 .. 0 0 0 0 •

0 " 0 0 0 0 •

0 0 0 •

0 0 0 0 0 "

•

0 0·0

0 0 0 0 •

0 0 0 0.0 0 0 0 0

i1
iii

A..

Bac:kgrollllil:.~·

The Mr·.Traffic . Con.trol Problem•••
2
.1.. His.t.or)r.
0 :.
2
2 .. Present State of Affairsoooooooooooooooooooo. 3

.Theall,o 0000·00·.0·0 • • " . " . 0 0 " . " • • • • • • • • •

0 0 0 0 0 11 0

0 0 0 .. "

The' 'Computero

D.
E.
F.

Extent or Use of

CHAPTER II

0 '0

0.0 .... 1;> 0

I;> "

0 .0 "

O.D •

00 "

0 0 0 0 •

•

0 0 0 0 0 O.

" 0 0 0 0 00 0 '0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 .. 0 0 0 0 0

t h e C o m . p u t e r o o o o o o o o o J o o o o ..

-Bi-acuss-ion-of-R-esults-o-.• o
Rela

'1;> '.'; ••.

tionship of Chapter

0 0 "

oooo~

0 0 0 ... 00" 0 0 0 00000000000

0 00000000" 0 •

PRESENT-DAY.·· AIR· TRAFFIC· CONTROL 0

" " 0.0" 0 " .. 0 0 0 0 • •

0 0 " .. 0 0 0 0 •

00 0 0 0 0 "

7
8
10

11\

J).

14
The Civil·Airways Systemoooooooooooooooooooooooo. lS
.communieati()n~Faciliti-e8o~';oooo .....
,,-o.ooooooooooo 16
Seope .of Pres.ent-Day Contr.Glo
16

E.
Fo

Air Rout.:eTr&tricControl • .ooooooooooooooooooooooo 18

A.
B.

Governm.en'talOrganizatlo11ooooo00000GOOOOOOooooooo

0 ..

'l'yp.es o.f- 'Control 0 •

.lo.
2.

0 '. "'0' 0

0 .0 "

0..0 0 00 0 0

0 •

0000" 0 0 0 0 0 0

General. 'Philosophy 0
-0"
D1:vis.ion '.ofContr.ol..
'0-'';'" "
Processing 'and Handling of n:lrght
""'''''.00

.0 0 0 0." 0 0 " 0 .0 .. 00

.'?

0 0 "

•

CHAPTER III

Na.vi·gationo
Communica tionoo

00 0 0 0 "

Tra:ffic" .GolltroI

0 "

\10 0

THE. _CO.MI>UT.ER..o,& •

...0

0 0 0 0

o o o o o o o o ..

" 0 0 0 0 0 0 0 0

0 "

0 0

0 0 0 0 0

23
2.4-

0 0 0 000

24.

0 0 0 0 0 .. 00 0 0 0 .. 0 Olio 0 .. 0 0 . " 0 .. 0 0 0 0 0 0 0 0 0

0 . • • "•.. 0

0 0

0 "

0 0 .. 0 0 "

0 .. 0 0 ..

0 "

a ....

0 0 00<) " 0

000 0 0

e .... -.•.• _••••

• . • • • • • • • • • • ____

•• 0

General Philosophy of Use··af A,· Computer it it
1. Elememtsot a ComputingSystemociooooooooooooo
2. Handling of a Simple Problemooooooooooooqoooo
3. Comments··onGomputer Orders.
~
0 0 0 0 0 0 00

4. A. ·.re~GeneralizedA.ttitu.de
5. The Problems of

0.00" 0 0

17
18
20

" 0 0 0 0 0 0

0 0 0 0 0 0 "

Traftlc·Contr.ol·.at the.Airports 000'9 <> o·~
Limitations of the Present-Day·System ••

••

0 0 0 0

0 " 0 0 ..0.0" ".0 0

0-0-'; 0 0'0 0 0 . . . .

:Enf-ortlla1ii-ono

0 0 0 0 0 •• " •

0 0" 0 0 0 0

26
26
28

0 0 0 0 0 0 0 0 0 0 0 0 0 0. 0 0 . ,

';,2

Programming.GOGOOOOOooooo • • • •

Report R-203
vi
TABLE· OF CONTENTS

-(c.ontinued-.).
Bo

The Whirlwind. C-OmputfJrooooooooooooo.oooo.-ooo ••• o • • • •
Repres.entationot Inrorma-tionoo. 0

10
2.
.3
G

40
50
CHAPTER IV

Pag.

37
37
Electronic Aspects . and the Storage El~t...... 38
The Control .Elem.ent·o 0
oe
39
The Arithmetic Elementoooooooooooooooo •• ooooo ••• 40
Input-Outputooooooooooooooooooooooooooooooo.oo •• 40
0 0 0 0·'- . . . . 0 . 0 . 0 . .

0 0

0 •

0 0

0 0 -0 •

0 •• .-.

THE COMPUDR-

_INPUT-QUTPUT -CQNSIDERAIIONS FOR
CONTROLLED

B Y 8 T E M o o ' 0 0 0 0 000·<>- 0 0 0 0 - 0 0 0 0 ( ; 0 0 0 0 0 " '0'.0.0 •

•

~...

Information-·HB.ndling in the Present-Day Syst~a•• ~...
An Ou.tline·ot an Input-Output 81'8'610 ................
Detailed' Consideration of the Use of Termi_~
Equipment.o

0 0 .0 0 0 0 0 _0 0 0 0 .0_ 0

0 0 0 .0 0 0 0

0 0 0 0

0 0 0 0 0

0 0 •

0 0 0

•

42
42

43
47

47
20 COBversion.··.o
48
.3 0 Message Form and - Standardization 0
51
49 C~ti'Oll"'"!!l1t··Hegard$,·-Inpttb-eutput... ••• 52
10 Reliability and Checkingoooooooooooo.ooo ••••••••

it.. .........

0 0'" 0 0 0 0 0 0 0 0 0 0 0 0 0 0 GO. 0 0 0 0 0 . 0 • • 0 . . . . .

it -.; 0

CHAPTER l'

STORAGE (; "0·

0 '0·

-Data. .storage

0 .; 0- 0 0 0

in~

'0-

the.

0 0 ·0· 0 0 0 0 0

0 0 "

,,-

0 0 "

• • • • -; •

54
54

~•

.. 0 • • • • 0 • • • • • 0 •

.Prese.nt.~..Dq..~.o <>-'" .. 0 6"0

t,. alld'

-Genera1..R-e1a~:B-atweml..:.Bt-orage~~Gapaei
Op.erat.i.ng~. Tj,.me.~.<>--.O- ° -i>~'o 0: '0.. .c; «> " 0 0- 0 ''; 0<> 00· 11:0.:.0 ..0;,. 0 0 • 0 0 it
0-

•

StOrage~TimeC-onsid-erationa .In .. the d~Jlt-Da7
-System....-o---o
<>
-0 <>
~
:-0. ~.
2. -Storage--TtmeConsiderat±ons' in the COlIlpUter-ControlledJ3ys:temo ~
oe
00;-0 0

-0 -; 0- ..;. 0

0 0

G..

0 (; t>

I) <)

0 •••• 0

0 0 0 0 0 0 ' 0 , 0 0 0 0 0 0 0 0 •..0 . • 0 0 .

0......

General Problems .Auoru.ated .withStorage ~ (i ~ ~.- 0- 0- •
1.- Characteri-stics-Ofthe -Stor-ed. -Inf--Or-.tione
0

• • • •

00 0 •

,3.;,

a: ... I)" ~ oeo
Hand11ng<>f Variabl.-6.-Length Data ••• 0

40-.

·.It---l~Rgth·-Stor-&ge-'.oo-~

0- 0-

0- 0

00.0

.-;000. ;

0 .0.0 0 0 0 0 0 0 0 0 0 0 0 0 0

;. ;.

I). I) I)...

It I) 0 . 0 . 0 0 .

00. • • -;' 0 • • 0

Applf-cat±on--of':progra-ms-to· Varl-ousBl-oclts of
Storage
0

D- 0

•

0 0 0 0 0 0 • • 0 0 0 0 0 0 0 0 _• • • 0

0 •

•

PDint and .Pa_th.Data-.
~
a -0<>0-<>
l~ Int-Ot"ma. tionto-beStor-ed
21. Foint-Path Relationship 00
00'
3" Illustra t1:on~ o-f'·Poin-t Data· Storage-.o·o
40 Flow Diagrams for Path Dataooooooooooooooooooo ••
J:> 0 -0

0 -00 A.o 00041 00041 •

000 0 0

•••••••

0 0 .. 0 • • 0 " «I 0 .. -0 «I • • • 0 • •
0

000000000 •• 0

0 o. 0 •

«10..

0 00 •

••

56
57
58
58
59

6i
62
~3

6-3
6~

64
_69

Report R-203

vii

TABLE· OF CONTENTS
. (continued)
Page

Handling of Flight Plan

CHAPTER VI

Use of" Ad:d:tticmal ·Externa;l stara:ge- S-pac" 0

0 0 0 0 • •

MUS AND AIRP~ 0

0 0

0 0 0 0 0 0

Areas

0 .0.0 0 00.

10
2.
.3.

0 0

0 ·0. 0 0

.1)

0 0 .0 0

0 0

0 0 0 0 0

0 0 CI 0

0 0

0 0 0 0

76
76

Controloooooooooooo ••

0 0 0 0

0 0

Area Coor-dination in' the .Present-Day Syst~ 76
Criteria'£or Area C.oordinationo
<»
78
Ar.ea .Goo:cdiDation.£.or-(1ompt'lterContrel.
79
Transmls-sion of Flight PlaM from Area to
Area
82
00004

0 0

0 0 0 0

0 0

0'0 0

0 <>'0 0

0 0

Present-Day C.oordinationa tthe Airports 0 • 0 S3
Requirements tor Computer-Controlled System 89

DIRECTION AND_APPROVAL REQUEST.

P R O G R A H > 0 0 0 0 0 0 4 1 41

Direction Progra.mo
ii
.1. .Me.ssage .and . Time-Ord-er-ed Programs 0 ~
'00 0 0 0 ';:' 0 0 0 0 0 0

2, 0

71
74

0 ••

The Problem at Airportsoooooooooooooooooooooooo
].4

CHAPTER VII

0 0

0 0 0 0

Elim.inationorSe~tor

Dataooooooooooooooooooo

Timing-o

0 0 0

0 0 '0 0 0

GO'o 0 0 0 0

0 0

0 0 00

0 0 0 0

0 11 .0 0·0 1> 0"0 0 '0 '0 0 0 0 0 o. 0 0 .0 0 0 0 0 • •

88
88
88
90

30 Read-In and Read-out Programsoooooooeoooooo 90
40 FlowDiag:ramooooooooooooooooooooooooooooooo
91
_.
,

Approval Request Programo·o

0000

Purpos.e . and Present Use 0 0

Initiati.onand .Termination o£ ContrJl1o

Appl1.cati.on-ot·Programoo-oooooooooooOOOOOOOOO

44 lhssage.ForlDS
50 Fl:ow Dia'g:ra;ms'o

GO 0

CHAPTER VII I

0 0000000000000041 0

SEP.ARATION

<> <> <>

<> <>

0 0 0 0 0

<> <> <> <>

0 0 000 0000 0 0 0

P R O G R A M o '0 0 0 0

0 0

-0

.0 0

000'-0:0' -00

Q.G.G 0

000 00 00 0 0

0 0

0 0 0 0

Separation Standards
00
~
14 F~.-.and. :Molting-B1eeks.Sys.temo.-o:<>-o
0

0'0 0

0 ..

0 0' 0

0 0'0 0

0 0

00 0 0

0000

0 0000

10.3.

'I'ypes of' ·Separ-at1Qn~

o£.-Program.o.o <> <>

(10

0 0

104
104

106

0 •

10'7

0 0 41 0 0

ExceptiQns to Use of !formal Separa tiqns

.Genera1~Pro1t.isions

O.Oc 0 0

0 0000000'0000000 0

(10

lit

0 0

00 0 . 0

~ -0- 0

0 0

91
91
93
95

ol~.. ~ 0041 0

.0 0 0.0 0

0 0

lOS

10 Path and Point Separationoooooooooooooooooo 107
2. Broken Flight PathsooooooooooooOOOOOOOOOOOOOO 109

Report R=20,3
v1i.1

TABLE·OF CONTENTS
(continued )
Details or Separa:tlonProgramo

109
General- Flow Diagramo ~
109
i.9ngltud1na.1SeparatioD.oooooooooo"1)OOOOOoooooo ••• 111
Al ti tude S.eparationo
0 00
III
Holding Sepa-r~tionoooooooooooooooooooooooooooo •• 120

1.
2.
.'30

CHAPTER Dr CLEARANCE PROGRAMo

l!.a

00.

~ _O.t:?

0 0

0" 0

0 0 0 0 0

o· 0 0 0 0 0 .., .., .., ..,

0 0 0 '" 0

0 0

.., 0

0 0

0 0 0

0 0

0 0 0 ••

•

••

0 0 ••

•

••

Process1ng·orClearancee.oooooooooooooooooooooo •••••
14 Present-Day MethQdso
00
0
2. Standardized Comput8r ProeedveaoooOOOOOOOOO ••••
3
Handl.ing -Coordiaat.d G·lea-ranee-a
0

'0 0

0'0 0

00 0

0 0

•

0 0 0 0 0 0

0 0

•

0 11 •

122

••

122

••

123
124

Conflict Eliminationoooooooooooooooooooooooooooo •••• 12;

Co
Do

Alt:i tude Assignment. °
128
Su1::division of. hq.gramoo-o.vooo.oooooooooooooooooo •••• 132
0

<).0

0 0

0 0 0

0 0

0 0 0 0 0 0

0 0 0

0 0

•

••

Relationship'" of Subprogrameo
2. Progress Report SubProgram",,,,oooooooooooooooooo ••
.3 0 Cruise Clearance . Subprogram
40 Asc-en:t-and Descent, Gl'ea-ranee Sub-programs.
10

0 0 0 0 0 0 0 0 0 0 0 00000000.

OHAPTER

122

..,

0';: 0

132

••

132
1.32

0 0 000.

136

•

GENERAL' SYSTEM' IMPROVEMENTS WITH' THE . USE' OF' call:
. PUTE!B,:o ~
e
~'I o. "o.~ o·

138

Introdue.tion ° 0 '0'0' o· 0
<>
<>
<>
Bi All....W.atherControloooooo<>oooooooooooooooooooooooo ••
C.. .All=Area··Gontr.ol C) .Q -0
<>
1».
Do Us e . of Ra.dar
o.
Eo' Sched.mi.~o
<>
<>
<>

13 8
1.38
13 9
140
UJ.

'CJ'",

0 .., '0' 0 .., 0 0

0 0 0 0 0 0

0 0

0 0 0

0 0 0 0 0 0

0 0

0 0 0 0

0 l The basic ability of these 'D,ISchines is
the automatic sequential performance of elementary aritbmetic, and lo~cal
operations 0 The application of machines with these capabilities to" the'
mechanization of tlte extensive computations arising in conneq:tionWith eel''' '
tain ~thematieal Problems is not onlY possible but is highly, des~.ble,
and already the prdSTess 8~c.mg these linEl's has been such as to ·permit further exploration andresear,ch intor previously !Daccess1b~e rel1'.n8 of, both
pure and applied st'$ien.ce" The use of th.se, neY machines as the 'centrel or
directtng ,elements in complex physical situations 'is similarly possible,
and desirable,_ and,1n such ,applications these maehines otfer~eat ~r01l18e
in permitting humau to m~e fully exploit and utilize recent, advances in
technology and scieMe
'~
0

The au.thor's first interest in the u.se of a eomput:l.ni machine
fQlO,a1:r t:t'affiG'eont~l ar_e in 1949 as a result of his emplO¥tllel!l.'t: at the
ServameCmlIl1Sms"'ta1;'oratory of ,the Massachusetts InstItute of ~ee~ology.
At that time the L$bCAt'atorJ" was completing construction of t~e high-speed
Whirlwind Com:pute~ aDd a projeet was initiated to study the pr.Sible' application of eomptt~e:f.S to! fllture air traffic control systems.
~

I
'

This study. project, sponsored by the Air Force, wes eoncerned
primarily With the .J~se of.' ft, eompute;r in a future system of tr,ltfig",control radically difr~nt trom that, "'hich is presently in use. 'SUtlP-',', '
system, it was p18U.a., ~qldemploy radar devices as well as ,imPrOVed
, navigation. aDdd~ota.m\m.ieation equipment in a11 attempt to hendleh1SbJtra~~
fie densities and "penJ.it high landing rates ,Inasmuch asthiawork was
concerned with the uae of "he cOmputer in a system which itself had _t
been formulated, same thought 'Was given to how the computer could be 'Used
,

10' The difference between digital and analog machines is expl'SiDed _

pg.

260
20

The reports of this project--Proje~~ 6673 of the DivisiQ. of Iadus~
trial Cooperation of the NBssachusetts Institute of Te~e~ar.
available at theP.roject Whirlwind Library at MoI.T~ J.i'o:r themest
part, the literature of the project is classified asRE8fR~C'lD.

Report Rc:0203
-2in the existing system of controlj and it was. decided that such an application of the computer would be undertaken and investigated as a thesis
topic 0

Shortly before embarking upon the study described ~~r~, the
author prepared a seminar report on air traffic control as a part of the
degree requirements at the MAssachusetts Institute of Technology.l This
seminar reviewed the past history, the present problems~ and the possible
future developments of air traffic eontrol in the United ~tateso In the
course of the preparation of that report, it became evident that. this
thesis would represent a. part of an at.tack on a problem.of ourrent importance as well as being an illustration of the application of a comput or 0 As shall be mentioned later in this chapter, the results of the
work indicate that the application is feasible and merits further and
more detailed consideration as aso~ution to the problem o
B

The Air

History

Tr~.ffic

Control Problem

Improvements in the reliability of aircraft as well as new developments and improvements in communication and navigation equipment paved
the way for a tremendous expansion of civil and commercial aviation in the
decade prior to the second Worid Waro The resultant increase in aircraft
flights during this period necessitated the organization and control of
this traffic both at the airports and while aircraft were en-route between
airports
The need first arose at airports -- the focal points of air
traffic =- where it was desirable to have coordinated control over the
many aircraft taxiing on the ground or flying in the vieinityo During the
1930 v s and through 1940 this type of control was provided exclusivel.y;by
means of control towers operated by private organizations and various
municipalitieso
0

By 19.33 it had become evident that some form of coordinated
traffic control was needed for aircraft en-route in flights between cities o
As a means or satisfying this need several airlines established a systeJQ..
of exchanging reports on the positions, altitudes, and speeds of their aircraft During 1935 and 1936 a number of airlines banded together to establish an experimental traffic control center at Newark, New Jerseyo
Late in 1936 the Federal Government assumed control of the three then existing centers and established additional centerso By 1941 a total of
12 traffic control centers were in operation with control extending over
20»000 miles of airway-so
0

10 Reference 10
items in the

(Reference numbers in the footnotes refer to numbered
Bibliograp~)~

-3During 1941 8 great expansion of air traffic resulted frOll mUitary preparations and large numbers of military aircraft were· operating
from ~ivilian iieldso As e means of establishing 8 unified control ot
air traffic, the government 'WaS authori;ed . to maintain 8lild operate traff,-e
eontrol towere at ma jor airports
Tbenumber of control towers and cont:rol
centers grew very rapidly during and after the. war, until by 1950 8
total of 170 Airport Traffic Control 'fowe:r:s and 30 Air Route 'f~ffic Control Centers were :in active operat1ono l i
0

Present state of Affairs

A general description of the presen~-day systemoftraffie coatrol is given in. Chapter n, flndmore detailed. 8speets of. the system are
described in following chapters
It should ..be understood that the present sy~tem does not represent 8 recently-devised effort at creating a
sat1sf8~tory means of control for currently existing conditions; the pregeat system is the culmination of a number Of ~ifications and improvementa which have been. made :in an. effort to k~e..p pace with the ~ess
in a 1r~raftaD.d air traffic
This present system issstisfactOry to the
ctent that is provides 8 vor'k.able and 'safe sy.ftltem of ccmtrol;· however 1
eert6in limitations of the system. are quite eV.1dent in the light of the
experience of tl1e past five years and the expectations for the future.
I)

At the present time commereia.l airlines carry some 15,000,000
passengers per year, this figure represst~5C1fo of the total first~~$S passenger travel in the United Statese • A further mereasea this
per~entage seems quite lik.ely inaunrru.ch as both.the tares and safety reG.ora.s of the airlines are fast approaching end in some eases bave alr.eildy p~uu:led the corresponding figures for first-class :rail and ship
tran$lportationo !file experience of the 'lI'st few years has clearly inti:i~ated the potentialities of low-fare .air-coach service, while the field
flit ~1r freight transportation, relativelJ' unde.ve lope d before the war ex·~pt for 8 limited amount of air mil scvice,.",nowseems capable of ccmaiderable growt'ho
Despite the rather optim1stie.pict~,e of possible expanSion,
the progresff» of commercial aviation is being re.tarded by 8 laek of dependable and reliable serviceo It is obvious that air travel 'Will not
rea~h it~ tne position as a means of transportation until people are
re(lsolUibly certain that they can arrive ,and dePart on timeo Similar
remarks apply to freight and cargo movements where the advantages of high
speed~ are quickly nullified by cancellations or dela;ys
Although 00111Jtlete fi~e$ on d.elays and eancellaticm.s in. a'I1 types of 'Weatheral'e l'1o"\r',
0

10
20

:Reference 2
Reference 3

Report R...,203

available 9 it is interesting to note the fi~es for a representative a~
line operating in and out of New York Cityo
During the relatively fairweather month of Junes 1947» 89% of the arriving aircraft of this airline
were 1ats,9 while 4($ were over one hour lats o or the departing aircraft
41% were late.9 with 16% being delayed over an hour o
The air traffic control problem has economic implications,
and is a. contributing factor in creating losses in revenueo Typical ligures resulting from a study of active and potential revenue losses of airlines in 1946 showed the following itenu9 g:

a)
b)
©)

Cancellations due to weather

Low lqad factor resulting from
unreliability
Congestion at 13 stations at
which stUdies were made

12»200,000
219100~OOO

The losses to private operations", military operations, and non-scheduled
services as well as losses at airports at which studies were not made
are not included in these totals o Although some improvement· in reduotioD
of delays and congestion has been made in the past few years, the situation is still far from satisfactor.yo
'The need for an improvement in air traffic control does
not result solely from t he problems of civil or commercial aviation;
considerations of national. defense and security require that the control
system be capable of meeting certain militar.y requirements. In the event
of a national emergency the United States wil~ quite likely becaa~ ia
active air=supply area and both civil and military aircraft will be used
to transport large quantities of men and supplieso The success of the
Berlin Airlift3 and present operations in Korea emphasizes the practicality and a.dvantages of large=scale air supply operations o Future operations of this nature may be expected to increase to ver.y large numbers of
a.ircraft and it is essentia.l that the traffic control system, be geared
to handle an increased level of activitYo
The present tra,ffic control system was not. designed to
ha.ndle the type of aircraft now in use or which will shortly· be coming
into use; this system was best suited for the years from 1936 to 1940,
the era of the DC-3 vs 9 a non-pressurized aircraft with an airspeed of

10 Refar~e 49 page 107
20 Reference 5, page 5
30 Reference 6

Report R=20,3

-5about 150 to 180 mph o At the present time there are about 100,000 private aircraft and about 7~OOO co~rcial aircraft (1,000 scheduled,
6»000 non=scheduled) in operation o
These aircraft range from an extreme of a one=passenger capacity and a cruising speed of around 100 mph
to larger commercial aircraft with capacities close to 100 and speeds
between 2;0 and 325 mph o The smaller aircraft with IlOll~prMlRlr"Zed cabins
are limited to altitudes' below 10,000 feet9 while with cabin pressurization it is possible for newer aircraft to cruise as high as 20,000 to
30,9000 feeto The airer~.ft are all powered with propellers and conventional piston-type engineso
Present developments and future plans point to jet-propelled
Although jet aircraft are already in extensive use by the
military~ there has as yet been relatively little development along commercial lines in the United States o The principal work on jet aircraft
for commercial use has been done by the British and Canadians, and each
countr,r has already flown models which may be in active passenger operation in the near future 2 Conside.rable speculation has been advanced
concerning future jet a.ircra.ft, and this coupled witli what is already
known about their operating principles indicates that these aircraft
will fly at altitudes between 20 9 000 and 50,000 feet, will cruise at
speeds between 300 and 6bO mph, and that high fuel consumption under certain conditions will require maximum flying time at high altitudes with
a minimum of delays at low altitudeso 3 The necessity for high speeds and
a minimum of delays especially at low altitudes make for stringent requirements as regards air traffic control, and create conditioRs which
the present-day system can barely cope witho Fortunately the military
services presently restrict their jet operations as much as possible to
fa.ir=weather conditions o Such restrictions could not be tolerated with.
eonnnercial jet a.ircraft which could fly above the.. en-route weather oDJ.7
to have their operations restricted due to poor weather and conges~ion
at the airports.
aircraft

The basic elements of an air traffic control system are fourfoldg

a)
b)
c)
d)

navigational aids
communications facilities
a carefully,;",organized pl.an for supervising the tra.f'fic
a supervisor,y or controlling element for implementing
c) above
0

10
20

.30

Reference 7
Reference 8
Reference 9, 10

Report R-203
-6=Each of these elements is capable of satisfactorily accommodating a certain limited amount of traffic, the capacity of the composite system, however~ is limited by that element with the smallest capacity.
A shor~ discussion of the limitations a.nd capacity of these elements is given at the
end of Cha.pter I10 As noted in that chapter, studies have shown that Olle
of the limiting factors in the operation of the current system with r.gard
to en-route flights is the supervisory or control element;l the basic
considera.tion is of the speed. and relia.bility with which human controllers
can perform certain basic 9 albeit simple, data-handling functions.
The capacities of the a~ove-listed elements are not completely
independent of each other 9 a.nd certainly with modifications and improvements in other parts of the system it would be pdssible to decrease the
work of the controlling element, thereby improviagthe system capaoity.
This» in essenee~ is the underlying philosophy governing the efforts
presently being made to synthesize a new and better system of control.
To this end a good dea.l of effort has been put into improving or developing new navigation and communication equipment as well as into organizing
the traffic in Bucll a way as to ease the burden on the human controllers.
Although the na.vigational aids,9 communications fa'cilities. and.
traffic plan each have definite limitations, it is fairly clear that definite improvements in present system operation and capacity would result
if a controlling element of a larger capacity were utilizedo This controlling element might be totally or in part a high-speed computing machine
such as Whirlwind$ and the general purpose of this thesis is that of determining how and where such a computing machine might fit into the framework of the present system o In so far as is possible, this study considers the utilization of the computer in performing the necessary control
functions» all other elements of the system remaining unchanged o
To the author's best knowledge, this thesis represents the first
consideration of the use of automatic machinery for performing most, if
not all~ of the duties presently performed by the controllers o There bas
been one nota.bIe attempt in the past to decrease the general burden. of the
work of the controllers by the use of automatic machinery, this being made
in 1940 at the control center in Washington, Do Co In an attempt to relieve human controllers of the routine data-handling chores and permit
them to fully concentrate on t he traffic situation, an e~ectrsolUechanical
posting system was insta.lled in the control center; data was automatically displayed and sequenced on boards similar to those used for the tabula.tion of' stock quotations, and use was made of a telet~pe ....like system for
the internal' handling and transmission of informatioll o
The equipment

10 Ref erenC$ 11
~.

Reference 129 13

-7-

Report B-203

was'W)l'e11able and was taken ou.t of service ju,t before the var," and. lUfttrther attempts at obtaining satisfactory operation have bee. 1I18de.On'
a much smaller scale, e,uipment has been designed and is presently in fuse
as an aid to h'timJaD. controllers in the handling of data in conaeetion with
aircraft landing at LaGuardia .A~ort iJa New YCI)rk Cityol
Co

arh.

Computer

The development. and utilization of pigh-speed eompwt1DS_DI8chmea ~s been so -rapid and of such recent datil that OXlly a l1m1t:ed number
of people a1"e aCQuainted With the principles of their,.se 811dope~t1.n.
These ~chines will be used in very widely different fields at' wo~k; ~
deed, one at the' major problems in effectively utilizing these IIJ8Ch~es
is in bridging the gap between two groups g those who are intimately acquainted with the _Chines and their operation, and those who desire to
use these macllines 'for a particular purpose, but who are lUlfala111ar with l
their abilities a:ad usage co As a means of bridging this gap, ellapter III
of this thesiS has been written with the express purpose of e~18iD.1ng
the basic ideas behind the use' of a computing maehineo Thai; cpsptel'
places special emphasis upon the use of a digital computing Jq'c1l11'1e1J1a
control ~tpplication, and devotes several pages to e descripti8 of the
Whirlwind Computer which is used as 8 prototype, in this studylt The silll11ar1ty between most existing or planned computers is 'so great that there
is- little-restriction in scope introduced by a specific study 1R relation
to the Whirlwind Computer; it is also convenien.t and valuQle to study
the problem. with regard to the characte:r,istics and capabilities of a presently-existing maehineo A computer designed speeif:tcaUy for' hal11dliBg
the air -traffic control problem 'Woul.~;t,lipt need the flexib1l1tJ which ex-i$ts in WhirlWind; however, the design of a speCial-purpose .chine for
this purpose would :t..nitially r~quire a 81tudy such as ~p begun here in
order to determine its necessary characjeristicsco
Although this thesis makes us. of the eharaejer1st,icS of an existing and operating computer,,2 it shoUl~l be understoqc1 that this study"
is a cademic to the extent that no computer presently e~ists which QcnUd,
be used tor this purpose, nor is it likeiy that the ne~essary machine ..
could be built and successfully put 1ntCl!operation Witpin _t,wo to five
years
To a great extent, this situati,n exists beea...e of the problem
of obtaining ~ont1nuous error-free oper"t1on of a high~speed computing
machine
lflle reliability of eleetroo.ic equipment has ~een. greatly
"P:t'ov-ed as a result of the introduction t! ney. teehniq'U,s,3 nevertheless
the reliability is obtain.ed only at -the ~xpense of reg.lar per:to4s of
0

i.-

10
20

Reference
Note-~

To a large extent the reaults of using ,hese characteristics
are not in bold evidence; they are im.pl~ed in the seneral
manner in 'Which the prob.f.'m is handled aRd in the m.ethods
used to peri'o:rm eerta in flmctions
Refe:rence 15
0

Report R-203

-8mint_nee and eheck;J.n.gc. Machines presently being. constneted 'bave not
beem de~1gned with the intent of contfnuousQperation far lang periods of
time; the accent hB s been pla ced to 8 greater exten.t upon satisfa ctory
error-tree operation for shorter periods ot time, ~8er1ficing for this end
a continuitY' of operation" ~is, however, is bu.t a, current situation
which 'reflects the relative infancy of the cbmput1q machille, aDd 'it ca.
be exp,ected that improvem.ents in electronic components and technique.
will sh(f)rtly permit -the construction of cO!JfPut1~ machines capable of the
contimtous error-free operation :necessary for the :p~oposed application.

Lest it seem. that the necessary improvements and teelmical advances .. ke the use of a computer far air traffic control only a thing of
the ~ture, it should be noted that at the p~~sent time the work of coatro~lerSl at trafticcontrol eel1ters and towers. virtually ceases dur1llg the
early morning hourso In the place of a fully-automatic system,' one might
@oUlpC»'Und a semi-automatic system in which the human~, controllers would supervise traffic dUring the early morning hours while the eomputer underweat
checking and. maintenancec. The main body of this the$is hypothesizes the
availability of a continuous error-free computer; scme discussion is made
in the next section. coneerning the coordination between a hu.man supervisor
IJnd the computing me chine ()
Do

~,ent

of the Use at the Computer

One of the first ,uestions which wst be answered in appra ising
the a~ic8tion of ra computer to the present-day traffic control system
is that dealing with the extent to which'the compu.ter will perform the
work of' the h'Ui'llBD controllers &On one band lies the possibility of eemplete1:l repla emg the humen controllers by a ,coulputer, en the other is the
possibility of utilizing the eomputer only as an aid to the human controller.
In dealing with this fundamental question. perhaps it is best to
first enumerate several important considerations dealing with each of the
two elements -- the h'U.'man controller and the computer" As shall be pointed out· in Chapter III, thecCDputer is 8 relatively 'fast machine quite
capable of handling and process1:tlg large amOlU.1ts of data in short periods
of timeo The computer can. easily perform simple arithmetic eperatlons aDd
can make eerta1n decisions for which it has been previously iastructed.
The computer has a memory of 8 limited Size, but this memory is an eff1c.ient one and information can be completely committed.. to memory or recalled
therefrom in 8 uniformly short timeo The human is not em.inently suited
for .rapid and yet 8 cc~te operation upon large IImounts of data, elld great
reductions in his operat1Bg speed are necessary to produce extremely reliable results The outstanding characteristic of the human controller
is his versatility, ability to meke decisions, and ability to imprOVise
0

=9,

in handling a.n unforseen situa.tion o The human has a relatively large
memory!p yet adequate short-term retention of information usually requires
the use of additional aids such as pencil and papero

A most desirable state of arfa.irs would be that in which every
situation arising in air tra.ffic control could be immediately handled by
a. specified standard procedure o Unfortunately this is not the state of
affairs at the present time, as is pointed out below,the standardizatiom
of the present system is fa.r from complete and it is often necessar,y for
a controller to improvise or follow other than specified procedures.
Nevertheless» if there exists a method. by which a human controller caa
reach a. decision as to what action is necessary in a particular situation,
it is possible for the machine to reach the same conclusioR o That is to
saY9 the computer qan reach the same degree of versatility as ~he human
controller if' the added costs and expense are deemed desirable. To this
end~ the question of using the computer for all the traffic control tuaetiona is somewha.t of an economic one, and the degree of meehanizat.ioa
must be bala.nced against the cost a.nd desirability of this mecha:nizatioa.
Thus the computer offers an opportunity for a completely automatic 8Y8t~ with increased capacity, but probably with less flexibility and with
cert8.in probl.ems of reliability
0

Realizing that the computer and the human each have definite
contributions to offer~ there exists the possibility of a joint manualautomatic system Such a system could utilize the computer for certaia
standa.rd routine jobsj) employing the human for special decision-makjq
functions~
both the computer and the human might perform certain checks
on each otherOs work c The control of aircraft at and in the vicinity of
airports ha.s peculiarities common to the individual airports and is undoubtedlymore suited under present conditions for huma~ supervision;
the decision as to how ·to create changes in the scheduIes of a number of
aircraft when a future conflict must be averted is a situation which can
be satisfactorily handled~ in its present limited state of complexity,
by human controllers~
0

The question of the advisability of completely automatic control by machinery is one which will plague engineers for ma~ years 1n
the future~ and this author does not see any clear=cut decision which can
be reached in the: particular application studied" For this 'reason, after
considerable thought and investigation it was decided to restrict the
consideration of the use of the computer solely to the control ot aircraften=route between airports~ assuming that the control of aircraft
at and around the airports would remain in the hands of the human operatorso This decision is justified in Chapter VII wherein a statement
of the functions of the. computer and human controllers in this situatioJ1
is made o Within the function of en-route traffic control, this.thesis
formulates plans for and. discusses how the computer would carry out the
necessary traffic control opera,tiolnls o It is assumed that in cert.aiR

Report, R=203

-10situations£) to be defined in succeeding chapters,9 the computer will present.
the situat.ion to a human operator for guidance and in~tructiono ' It is also assumed that a human controller will be able to insert a~ desired informa.tion or ins·tructions into the comput~r as well as being able to secure and inspect information stored and used by" the machine o
One of the most difficult tasks in this study has been the
assimilation and sorting out of the various rules p regulations and practices governing the operation of the tra.ffic control system. o There is
extremely little comprehensive litera,ture on the subject, and for the most
part that which does exist merely enumerates certain fundamental_regulations and procedureso 1
The present system is hardly what one woUld call eomplete~
standardized» and to a great degree the operatioR of t.he system. is based
on the choice of individual controllerso Presented with a given situation$ it would not be unusual for five controllers to each have a different method by which he would handle ito Even within the standardizatioa
which exists~ there are certain situations whi~h must be met and handled
as the occasion necessitates o An example is the fact that controllers
hesitate to use certain procedures with civil and militar.y traffic ina.smuch as the training~ experienee'~ and proficiency of the 5e pilots m.tq

be less tha.n' that' of commercial airline pilotso

The lack of rigidity in standardization is» of cOln"se, quite
the antithesis of the requirements of an automatic computero tor this
rea.son~ one of This chapter aims at briefly describing the m.ethods
and techniques utilized in this pre.sent-day system, and is intended
Chiefly as background material for furt)ler and more detailed disoussio.s
of the subject in succeeding chapterso The emphasis is concentrated upon
an explanation of the en-rout. phases of control, as it is in this part
of the geReral framework of the present-day system that it will be showa
that the use of a computer is most promising o
A

Governmenta.l Or ganizatiol'l

The two governmental agencie s directly concerned with the
present system of air traffic control a;re the Civil Aeronautics Board
(CAB) and the Civil Aeronautics Administration (ell)
These1;,wo orgaaizations technically constitute the Civil Aeronautics Authority of the
United States Govermnent o
.
.
0

The Civil Aeronautics Board is an independent, quasijl1d.icial
panel appointed by and directly responsible to the PresideR~. The Board
formulates Civil Air Regulations dealillg with the competency of pilots;
the airworthiness of aircra.ft, and the conditions of flighto Other
duties include the investigation of accidents and violations of the CiVil
Air Regulations, the estab1isnmentof air mail rates and contracts, and
the certification of routes, rates, and carriers for passenger and freigb,
tl"avelo
10 Ref'ereaee· 16

Report R-203

-15=
The Civil Aeronautics Administration is a branch of the Department of Commerce, and operates under the direction of an Administra~or ot
Civil Aeronautics o The maintenance and operation ot communicatioBs,
navigation,9 and air traffic control facilities was authorized in the
Civil Aeronautics Act of 1939, which empowered the administrator to:
promote the development and safety and provide for the
regulation of Civil Aeronautics
designate and
establish civil airways
a~qUir., establish, operate,
and maintain along such airways all necessar,y air
navigation facilities
provide necessary facilities
and personnel for the regulatio~ and protection of air
traffic moving in air commerce o
i
00 • • 0

00000

The Citil Airways System

A network of radio ranges known as '.Che civil airways form the
These
ranges are created by the radiation patterns of ground transmitter. operated 'by the CAAo Four narrow spoke-like beams -- also referred to as
courses or ranges -- are set up in space, extending outward fro. the
range sta.tiono These courses, each 3 0 wide, are separated by about, 90 0
in azimuth" The nature of the si~nal transmission ill the beams or
courses is such that a properly-tu~ed radio receiver will give an aural
and/or visual indication of whether the aircraft is to the left ot, to
the right of, or directly in the path of beam o
basic navigational aids for aircraft flights in the United States.

The radio range stations have generally been situated at' or' near
major airports so as to provide direct paths, or paths consisting of
straight-line segments, between most commercial or military airports.
The range stationsjl t hen, provide predetermined flight paths or t1road.."
which pilots can Itride tt by earphone or meter indications o
Radio ranges alQnejl however, o~ provide a course to be
flown; navigation im. poor weather and during low visibility conditiolls
requires the definite determination of position along these courses. A
radio fix or check-point is obtainable at the intersection of courses
of two adjacent ranges or in the signal-free cone of silence over a
range station In order to provide additional indications of position,
radio markers have been installed at specific points along the air~
way-sQ These radio markers are produced b,y high-frequency transmitters
which radiate vertically and produce a definite field pattern ove-,: a
small area across the airway-so
0

These facilities -= the radio ranges and radio markers-- provide a complete.9 although not perfect, en-route na.vigation ayst_ capable of being flown entirely by instrumento Figure 1 shows the pres.at.;.
d~y airways totalling 43 gOOO miles o
10 Referenee 17

A-450l1O

_--__
..-

..... ....

ava-AIRWAYS CIIMI"

-- -

---------• -=- -.:::.-.......
• Ii!l!!!D - - - -

-....--.-.,..--.
....... ..... ......

...-....--... - - .
.-.-...~

. . . . . . _ _ _ . .A'IC

--..
--.----~
_~III.-PI

_OIFClCII.-ct•
Gc:.

Report R-203
C

-16-

CommunicatioR Facilities

The bulk of communication between aircraft and grou.d stations
is carried out by means of two-way radio telephone. Traffic control hstructioRs from control towers at airports are sen~ to neighboriag aircraft over high-frequency channels assigned. each airport. Coatrol ...sages for en-route aircraft are sent by the traffic controllers through
the facilities of airline, military, or CAA-operated communicatioa stations e
Point-to-poin\ ground communication among control e enters, C08trol .towers, airline offices, airline radio stations, military dispa1c.b."
ers, militar,y communications stations, CAA communications stations, and
weather observation stations is handled over land-line wire circuits.
Two facilities are in general use: an extensive teletype circuit for the
transmission of weather reports or messages not requiring immediate handling, and an interphone (private telephone) circuit for the relaying of
air traffic controlmessageso
Do

Scope of Present-Day Control

The air traffic rules of the Civil Air Regulations specify two
types of flight conditions, with appropria.te regula.tions for each. ;Th.
Visual Flight Rul&s (VFR) apply to'clear-weather flights condueted UDder
visibility conditioBs better than specified minimums o These visibili\7
regulations are rather comp1$Xo Generall1 speaking a three-mile horizo.tal visibility is required and the aircraft must be at least 500 feet below or 2000 feet to the side of cloud formations. Flights which are
flown under visibility conditions'less than those certified for VFR are
governed by the Instrument Flight Rules (IFR)
Standard fligh\ procedures
and "right-of-way" rules for all flights, regardles~ of visibility, are
stipulated. in the Civil Air Regula.tio.so
0

En-route flights carried out under Visual Flight Ru+es are ~o~
controlled. by the facilities of the CAA There are two priD.ci~ rea'so••
for thiso Under the relatively low-density conditions presently e.~
countered ill en....route flight,!I when a. pilot can see tba t he is about
pass, overtake, or cross the path of aRother aircraft he can safe17 alter his course to prevent col1isioRo In full-visibility there also exists no urgent need for navigational aid. For these rea.sons pUo'hs are
permitted to select their flight paths independent of the civil airways,
and they are personally charged with maintainiag safe separation froaotaer
aircraft 0
fI

Flights conducted underIFR conditions require the en-route
navigatioRal aid aRel overall traffic coordinatioR of·f.red by the ell facilities; these flights must be flown on the civil airways and. are stric\11' controlled from. the grolUtd in accordance with methods described ia
succeeding.sect1oas e

-Report R-203

-17CAA supervision also extends to the control of aircraft in and
around a~rports, regardless of weather and visibility conditions, where
high traf£ic densities require careful co~trol for pUrposes of safety,
efficiency, and expedieneY6
Eo

Types of Control
I

The types of control exercised ill t.he pres••t,-day qat_ are

thre.-foldg:

1) air route traffic coatrol
2)
3)

airport traffic control
approach control

Air route traffic control, administered by cODlrolce.ters, is
specifically responsible for the safe and orderly flow of aircraft proceeding along the civil airways in IFR flight conditio.s o The prinCipal
functions are~
1)

To issue instructions to pilots regarding altitudes to be flown., routes to be followed,
speeds to be maintained, holding procedures
over specified locations, etc o
To advise pilots of hazardous conditions and
miscellaneous information which may effect the
safety of the flight, and suggest a change iR
flight planss
To maintain a progressive check of aircraft and
to initiate action for overdue aircraft o
To provide assistance to aircraft in difficult.y
and report accidents a
To report violations to the proper office. 1

It should be particularly noted that air route traffic control is designed

chiefly for application to en~route1FR traffic o For reasons to b. m.n~
tioRed in Chapter VII9 this service is not used by .,et pilots during
VFR conditions o
Airport Traffic control concerns itself with the safe and orderly supervision of aircraft which are taxiing, landing, taking-otf, or
flying in the immediate vicinity of an airport area The principal rUCtions areg
,
0

1)
2)

To issue-instructions to pilo\s for taxiing,
take-off, B-pproach for landiDg~ aDd landj tag of
aircraft 0
'
To inform pilots regarding field aad weather
eondit~o.s» air ,navigation facilities~ ~er...
gency landing areas, restrictioas to fath",

~1-o~R~e~t~e-r~en~c-.--~1~8-~~P~a-g~e~30~"

and other matters which may be of assistance \0 the pilo\.
3) To relay messages between pilots~ air carrier or mili\ar.y
operations of'rices~ conmumicatiollstations, and other
appropria'te agencies concerned with the operatio., control,
and dispatch of aircraf\o
4) To inaugurate emergency procedures whe. an accide.t or
ergency occurs on or in the viciRit7 of a landing are. o '

ra-

Airport traffic control, it is again noteds; is an. all-weather servia ••
A third type of service -- approach coatro1 -- is pr.s.Rt~prO
vided as a link between air route and airport traffic 00D\1"o1 ia Co. . .C-'~
tion with the handling and coordination of traff,i.e ill instrument weather
near airportso III the word~ of a eAA manualg
Approach control is a service whereb1 airport traffic
control toWel"SiS8Ue traffic clearuces to aircraft
bfi.:ng controlled in accordance with Instrumeat Flight. Rul.
standards by communicat~g directly with pilots over the
voice feature of the radio range~ or over a very high
frequency channel of the control-tower
CQordinatioa
of traffic arriving and depa,rting during adverse weather
conditions is vested in the approach controller who is in
a position to see the airport and aircraft in the vicinity
aDd is therefore able to take advantage of every opportunity to 5afel~ expedite the flow of traffic on and aroUDd
the airpor\o
I)

F'o

Air Rout. Traffic Co:ntrol

General Philosopat

0 0

,
Air traffic obviously cannot be controlled 11k. surfac. traffic;
an a.ircraf't cannot be stopped in flisht but must reman ill motioa at sutficientspeed. to ma.inta.in altitude and maneuverabilit.yo (I:u.sauch. a. "he
available navigational aids safely provide for ouy a siDgle lall(e -- oft.••
called a ttl single wall u1 -= of traffic. along a.n airWay, iJaediat. delays
of' en=l"'oute aircraft are possible only bY' flying the aircraft ill rac ....
tra.ck=like holding patterns anchored on these airways) e The desire for
the maximum. utilizatioR of the airways as well as the Ilinimum. aJIOuat Qf
'confusion and delay has' led to a guiding_ priBciple of' cOlltrolligth.
traffic by anticipatiolAo In accordance with this principle, the move..
ment of traffic·is organized in advance so that no danger of eollisioa
can arise if pilot$ proceed· according to in5tructions~ and if "he traffi© control is predicted on accurate and current iDfor.ma~10. of the
weather and aircraft move~nts.
l~

Reference
Reference

18~
16~

Pas.

21
Page 79

Report R-203

-19Whereas tower controllers are usually situated in a position or
full visibility at the airport under their supervision, air route controllers cannot possibly view the en-route aircraft and hence must be
supplied With a full knowledge of the positions and plans of these aircraft" This inform tion is supplied both by the flight planS which' ~t
be filed priOr to all IFR -flights and by means of progres's reports sent
by the pilots to the traffic controllers when ai~crart pa~s the parti-'
cular fixes designated as reporting points. Although these flight P:L~ns
and~progress 'reports 'dealexelus:tyelywith times and altitudes a.tthe,,'
disc~etereporting points, these reporting points are close enough --15
to 20 lftinutes flying time apart -- so that a trairiedcontroller can extrapolate upon the given data and supply-himself with a mental picture ~t
the traffic at the present'cr'at future timeso With these pictures"the
controller c};lecks for -safe separation between aircrafts; when it isapparent that proper separation will not exist~ the controller requests
changes in ~ flight plans of the appropriate aircraft.
Seps"J:'ation between aircraft is defined in ter~s of distances,
times, and altitudeso The separation standards -- lateral, longitudinal,
and vertical -- generally requtreany aircraft in IFR weather tOlba
either 1000 feet above or below anY' other aircraft on the same airWay,
or to be at least' ten minutes flying time a-way fromanot~~aircrar~ "at
the same al titudeo Tbestringency ofthese-standaros is 'pri~rilY due
to the limited 'nature of the navigational system and the . liInited'reliability and acctr.raey of barOmetric a'ltimeters· and aiTSJ>"eed indicators.
Variations 'ofSpeeds 9' altitudes ,and"routes ,as well as ,the' :hOlding of'.,
aircraft over specified points [# -canba used' to obtain pl-ripersepara tion
between aircraft Detai.ls -of the en. . -routesepara
.
ti-on standards and. possible variations therein permitted under particular flight conditions are
discussed in Chapt~ Vlllo
Q

'Before commencing any 'flight or portion of a'flight which will
come under the'jurisdictionof Instrument Flight Rules, pilots must ob':"
tain a clearance from the appropriat'e air route traffic control center.
The initial application'for a clearance is made 'by 'the prior filing ota
flight plan, indica tit).g full details or the proJeeted fligh~o' The'flight
plan must specify the flight'indentlfication,'type of aircraft,'proposed
airspeed~ point of'departure,'desiredaltftude,
propos'(inFrcUte; and
point of destinationo This initial filing' of a flight plan is 'generally
referred to-as ,'~ri'ApProValReguesto',' Ap~roval Req~stsarerorWarded to
the control-centers some tfmebefore proposed departure time by the airlines operations offices or the military ope~tions off;ceso
Clearances are issued to aircraft ona "first-come, firstbas is with due respect to up-to-date weather inforna tion and the
fullest possible knowledge of aircraft movementsQ The final' request for
a clearance, ,the Cleara.nce Request9 is' made by the tower operator at the
airport' when the aircraft' is actually prepared for departure
The
clearance ~ as such, is giverifor the aircraft to c-omri1ellcather'equested
flight; in actuality~ however, due to the fact that the weather arid
traffic inform tion at anyone time 1s know, only with a limited degree
served!~

Report R=203
-20=

of certainty, this clearance is based only on proper separatioas for a
first section of the flighto
After receiving their clearances, pilots are required to co.form. to their flight plans and to inform the controllers via progress reports of the passing of the reporting points o The progress report,s iIldieate the actual times at which the reporting poilts are passed, and
confirm the altitudes and speeds at these pointso On the basis of these
progress reports9 modified clearances can be given9 i f necessar.y, to
provid~ separation for further sections of the tlightso

Division of Control

Control of all aircraft on the civil airw~s of the Yait.d
from one central point is not feasible due to the high communications costs» and is unnecessary due to the localized nature of air
tl"'avel o (A eAA survey of 1948 revealed that almost half. of all airliD..
passengers travel less than 300 miles and over a fourth t~av.l less thaa
200 mile~ol)

State~

For thes~ reasons th. country has been divided up into 26 do~
mestic traffic control ~o This division has generally bee. made oa
the basis of efficiency and econoIDY,9 and so that one point of :major
tra,ffic co:ngestion and the airways serving that point compose an area o
The tra.ffic within each ar.ea i~ controlled by a centrally-located !it.
Route Traffic Control Centero
The amount of traffic within al~ one area presents. a prohl. .
of control which is too difficult for a single ~son to hamdle o For
this reason each control area is subdivided i.to seetorso The size ot
these sectors varies with the amount of traffic and complexity of the
routes
It has been found~ howeverJ' tha,t there exists a minim.'WI. size
for a sector from a practical viewpoint» since a controlled aircraf\
must remain in the area long enough to permit a controller to assimila'.
the i~formatio~$ analyze the traffic flow~ and issue the instructions.
Generally speaking, aircraft receive clearances through 0 •• or two 'sectors
a:t a t:im.e o
0

The subdivision of the traffic into areas and then sectors
creates serious coordination problems o Coordination is needed betw•••
adjacent sectors) betweeR areas~ between centers and towerss and betw•••
successive shifts of controll~rs working the same sectorso

30 Processing and Handling of Flight Informatio.
As a means of compiling and handling the iRiormatioR COReerRing the va,riouB flights within an a,re~$ the control centers are provided
with a number of flight progress boards ~pon which can be posted curre.t
f'~ight informa,t~on in the form of flight process strips?
A single fligh\
10

R~ferelm©e

Report R-203

-21-

progress· board is generally reserved for each rePorting point, the name
or abbreviation of that reporting point being displayed at the top of the
board~ Separate flight progress strips are
out arid posted for all
reporting points mentioned in·a flight plan, each strip essentially
carrying all of the information in the flight plan but prominently displaying the time· and altitude at which the point shall be passede

made

Adjacent reporting points are assigned, insofar as is possible,
to physically adjacent flight progress boardso A group of 5 ~ 10 contiguous flight progress boards constitute a sector under the control of a
controller The strips on a single board are generally kept in time sequence, and hence by referring to corresponding (flight) strips on successive boards and by comparing times and altitudes, controllers are able
to check for proper separation~
0

The CAA de£ines three positions of operation in a control center, these P9sitions and their duties being~
Fosition A~
a) To collect flight plan data via interpQone·and post
properly on flight progress (board) strips4
b) To maintain prIpared strips in proper sequences in
suspense bays.
c) To prepare strips for reporting fixes within the
sector assigned and prepare strips for the entry fix
(reporting point) of succeeding sectors as necessary"
d) To transmit over interphonefacilities assigned to the
position, flight plan data and control information to
other positions within the sector and adjacent sectors
as necessary~
e) To transfer prepared strips to the tiD" position (see
below) at appropriate fixes and at proper times.
f) To monitor reports and information being transmitted
and received by assoc"iated tiD" position of operation$
ackriowledging receipt of information for which acti()n
is required by the "Alt positi6no
g) To handle other communications via interphone facil;"
ities assigIled to the'position as instructed or autllorized by position uD
h) To remove and file flight progress strips from holdersno longer being used for control purposes.
PosttionD:
a) to issue trafficclearancresantiother: contral messages.,'
either d-ireotlyortmciughu other positions of operation.
b) To revise air route traffic control estimates as neces~
sary iIhmediately after a flight progress report has been
p~~~
,
c) To mintaiil fix postings on flight progress boards in
proper sequence.
.
. ..'
.
.
d) To :maiiitain current expected apprOach times as required
_~_____
on_r_l_i_g_h_t_p_r_.?gress boards assigned touthe position.
Reference 1. -Suspense bays are separate' flight progress bOards reserved
for strips of aircraft still on the ground or in adjacent areas.
lt "

Report R-203
-22...
TOrel'llOT8 'strips - f~m the flight progres-s boards when

theinf~ti'()'rl"eontained on the strips is no longer
required'for"'controlpurposes'o
f) To study-all weather reports , winds aloft· reports,
noti.e~:staf~en, ·and··-term:.t=-landai~way forecasts,
lime'· t with Fl.ight Advisory Weath,r Service personnel as necessaryo .
'
g) . ·to administer fi-ightasrlstanceto -eivll aircraft o

Pos,it~on

a)
b)

e)
d)

To' ·eoordinate·the'-em:ltrol·-of-a1r-~ratf:t.c between
'variou",seetOrs of ,the c-ontrolarea 'as; required.
To "cheek-' floy of" traffic between' sectors t·, insure
proper" s-eparattono
'
To advise' c'ontrollers as necessary' to 'i,sure efficient
and -safe' contro'l --of a ir traff±co
.' ,
.
,
To regulate th, flow' of air traffic in an efficient
and orderly- ms:,ner' c'onsistent-with the . ' operational
l~itin'r"'factors-involved-o
'
" . '. '.
".
'fo"stud) all' weather-reporls,Vinds aloft reports;
notices. to airmen';"andterll1inal and airway forecasts
immed-t~ely" uponrecrefp-:t'- of; such' reports. 1

Position 'D . duties . are filled- by--eontr0fierS, Position A duties
by assistant controllers -or ealeWators 9 while 'the supervision and coordinatiemof·th! 'E p-os1:lion is provided by a senior controll~r.
I

Go Traffic Contrpl" at" the Airp~s~_

Supervision-of':fl-tghtS"--a-t":-and--near'air}jorts-i:s--B4ared "by Airport
Traffic Oontrol"and- . ApprOaclt";e ontrol ":Ai:rport'fraffic 'Control'Ms been
estS-blished, . to·--provtd:e-·a:dequate'Supelvlrion'·"-of all traffie-'i~- -the movemellt'area'hand"all-ai~nfttfyfng:- in'~~D!l:'TeferriC'Ef-,to-the' :~ground in the
Tieinrty'-:'of"an--al:t port 0 '. ApPfOach' .Cont-rol g -a ''S-ervtce' '~rsta:li1:tsh,d in 1944
. am-still'''belng -'expanded"att'elllp'tlttci,
c~dftiQU . a-t"-atrt?ortsduriUg
ins t:r amen l:;-wather:bj":-'-direcrt;:--a1idins tantaneotis' 'communieati-on" between tower
-~~11erS-:'and'IPR-f-l:igIlt;s>.arrtvtng,·:a-t-?·, "de}yrati:Dg"'f~9 '-Or-holding irf
vicinity of the airports o At the present ttme over 100 of the CAA-operat-ed 'airport·· . towers- -of~er:'~1;m," appl oaeh "~o-~ servtce--.
0

_ft-

Duril1g- IFReOndittons g 'when-:aircraft:are'''T!ot'-1rilde"control of
a'irroUte--trs;ffic-·eontro-l-een,tel s 9 --ptlotlfgenel'ally-'-eontaet;'-tllJe towers: at
the'airport" of ·-thetr-~d'ERftliiatlon·wheiI--they-are'ufTom· 10- :t.'-=1JI!nutestly....
ingtime~awayo 'The . toWer-" contI oliers-establ-tshlanding "pi"b;erns . , suitable
tor-·'-the"'l'

an.,

'.·.'in,'l1ln!'-and':--tns Lx'uct-'·-the'--ptiots- 'u

·w·-trov-a11lJ i' when tl1e1."

should"proee.e·d wtth,"thef:rlandf:ngso :.: -The'se' instructions 9 .. at course 9 ' are

eOCJrdinated"wttlJ r,speet to taxiing

and'aire~ft

'aircrart~

other arriving aireraft g

-req,ueSting permission to tate otf.

Reference 22

Report R-203
The situation is somewhat more complicated at the airports during poor visibility eonditions o The lack of visibility aad the limited
aature of the navigational facilities are such that large s.paratio.s
betwee. successive landing aircraft are required, and h••ce rather low
landing rates must be expected A Budd •• deterioratioa of the weather
at a busy airport is likely to create a situatioR in which most ar.riviBg
aircraft must suffer some delays in a holding pattern before they can
proceed to a landiRg o
0'

0..

As an aid to navigation. near the airports 9
leg or course of
the adjacent radio range is gen~rally positioae,d so as to lie acrQSs th.'
airport. U.der approach control procedures the air rout. traffic control
ee.ter clears aircraft to a radio marker situatGd OD. on. leg of the radio
range. Holding patterns ,are stacked above this :ru.rker at lOOO-toot, ilI.tervals, the patterns being of a. race-track shape with one leg exteRdi."
along the radio raage fro. the markero These holdi.g patter.as are uader
the supervision of the approach controller who informs the control ce.ters of the free altitudes at which aircraft ~ be introduced i.to th.
patterru!lJ UpOIl reachimg the holding marker at the proper altitude I the
pilot establishes contact with the approaeh controller who the. a.suae.
complete control of the airerart o
H~

Limitations of the

Navigatioa

P~esent-Da.y

Syste.

GeReral~ speaking» navigational aids presently in use do Dot
give pilots or controllers continuous or suffiCiently accurate positiouJ.
information, Ror do the facilities provide a satisfactory number of
flight pathso

During poor visibility conditions, pilots know their positions
along the airways only at specific poiats -- these points being the
radio markers, cones of silenc.~ or ifttersectiolls of course. of adjace.t,
range statioRS~ This is a distinct disadvantage g traffic is Dot o~
restricted by the single lane of traffic and lim:i.t.d number of altitudes
which can be flown (bounded above by'coRaideratio•• of cabiR pressurization and below by the height of the terraiR)>> but must b•. further curtailed since the limited Rumber of reporting poin\s requires large lORl:1tudinal separatioas o
The preselllt-day airways are restricted. to t~. fixed, widely'
separated courses ot the radio rang.s o .The positioJl of t.~. course.
usually creates flight paths of uu..cessary le.ph, shce (lijs. fixed
paths between range stations rather than direct straight-liBe rou\ •• must
be followed it the navigational facilities are to be used o 'These paths
a.re immobile and cannot be altered to take advantage of weather cORdi\1o".o
1m particular"the radio'ranges. are less reliable in poor weather wh•• t ••y
a.,re m.ost urgently •••deCI. o

Report R-203
2.

Commuaicatioa

The overloading of available eommuaicatio. ehaRa.Is and the .tfects of static and interference are two aspects of the air-grouad co~
mURieation problem o Of greated im.portance is a considerable time lag ia
handling messages as they are routed to and from the' control c.at.re o
The m.ultiple. handling of messagesji the Recessity for frequ••t rep.titl0••
to avoid errors and misu.derstandings, and the overloading of availabl.
circuits are problems presently encountered both in commun1catioa b.tw•••
ground points and betwee. pilots and ground statioRs o Pilots and controllers a,like spend a l,t,sproportionat'e amouRt of their tim. h preparing
and transmitting messages~\ pilots must do this while busily engaged i .
flying; controllers must interrupt their control procedur,es to estab~.m
radio or telephone contact and to deliver (and possible repeat) the m.ssages o
\ '

Traffic CORtrol

The present system of contro19 based as it is upon the priacipl.
anticipatioR, is handicapped by a lack of accurate and complete taformation of preseD\ and future tr~fic conditions along the airwayso
This situation arises from several causes~

The lack of an adequate knowledge of the winds gen.r~
ia earlier or later t~e5
arrival at reporting poimts than have beeR previou5~ predictedo

re~ts

Position reports are likely to be delayed, incorrect, or
inaccurate
0

Aircraft flight plans are handled f8first"'coll.ejl first . . .
serve~n without regard to published schedules, uas ....
much as these schedules are susceptible to rapid chang••
and frequent cancellation during IFR cOllditio•• o Because of this fact» because of the randoa aature of militar,y and D.oR-scheduled air-carrier traffic» and because
pilots may Rot file an Approval Request until the time of
takeoff, cOJltrollers have oBly a limited idea of future
'
traffic eoaditioBs o

d) , The rapid chamge of weather conditions may result ia
numerous en-route aircraft without clearances sudd.~
applying for !FR clearaRces o

It i3 chiefly for these reasoms that aircraft caD be cleared for
separation through only one or two sectors at a tims~ Such a procedure;
as Roted, results in a coordination problems with an accompamying decrease ia overall efficiency and speed of operatioRo
A. exhaustive study has beea made of the conditioRs UDder which

Report R-203

the present-day air traffic control methods and procedures might, break
down under the weight of am increase in traffic den$ities o 1 It was eo.eluded that with high traffic densities over a proloDged period the following limitations of the controllers would become appar ••t~
a)
b)

an inability to properly visualize the traffic pat~.r••
beyoRd a certain degree of coaplexi,yo
an inability to resolve complexities aad de,eraiBe the
appropriate instruo~io.s with suffiei••t speed;
failure to be able to make aeeurate and curre.t tabulations'of the :necessary data in the time available.

The routine, mechanical task of transmitting, accumulatiagg
sorting, classifying, recording, and distributing the informatioa Deeded
to pe~+ol"'Dl the cORtrol functioRs is tremendousj) and probabl,. CUt only be
fully a'tspreciated when ORe has seen an air rout. traffic eo:ntrol e.n~.Jr
in full operation during IFR operatio.s.
1. Reference 11

Report R-20.3

-26CHAPTER III
The CollPU'ier

This chapter is illtel\ded as a brief ilttroductioJl to the g.n.eral
subject of automatic digital computing machiRee, al\d ia particular to the
Whirlwind Computer.
The first seetioR of the chapter is devoted to a diseussioa of
a sim.ple computing system and to a.n iden.tif'ieatioR tilereiR of the basic
elements of more sophisticated and completely automatic syste.s o The
discussion is focussed on digital computers, of which Whirlwind is a.
example This digital character evolves from the fact that the computer
handles aDd stores only informatioa whioh has b... properly expressed 1a
qwultti~.d or digital form -- that is to saY', in.formatioR which is completely expressible by a set of Rumerieal- digits o ,±his is to be 00.,trasteasic
Ratio.s dealing with the use of a computer in at. variety of probl.a o
Specia.1 atteJ'lt.ion. is give. the applicatioD. of a. computer to physical co.~
trol systems.
The secoRd part of this chapter discusses the specific characteristics of the Whirlwind Computero A number of quaDtita~iv. figure.
are i.troduced so as to provide the backgrowad for diseussioRs of suee ••diag chapters o No effort has been made to make the descriptio. of the
computer complete; additioRal details aDd descriptioRs of operation may
be found i . the literature listed in t.he Bibliography.
Ao

Gelleral Philosophy of Use of A Computer

Elements of a Computing Slst_

A familiar example of a computiltg system. is that of a h'UDlaR op-.
erator supplied with a desk calculating machine, a set of iastructioRS g
a peRcil, aDd a block of paper /) Despite the apparellt simplicity of this
example, it graphically demollstra,tes the basic ele...ts of the more complex od automatic syst_s which are typified by the WhirlwiRd Goapu'ier.

OD. of the basic elements i . this s.ystem of manual eoapu,aiioa
is the desk calculator -- this, ia fact, fulfill a the requir..eRts ot th.

Report R-20.3

-27arithmetic" element in ageneral:lzed -syst-em o Theebaraeteris tics of this
particulaT·arltmneticelement"·'are'ttmt·when- it is ' supplied' with one or
morenumbers'andwhenal'aritettiar'-modeof operation --add!tion, subtraetion,multiplieation-, ordivisi'On ,-- is-spee'ified 'by-s.'Ctivating the
proper lever, button, -or'switch, the--des-ired 'operation is -~'t1tomatically
carried.' out
Theeomput-ed 'resul-ti-s' -gene'rallyf'~und in the accumuJ,a. tor ~
one of the several registers used in the performance of arithmetic'operations.
0

The-numbenr·to-"be··fed' into tlte-"~rithmetieelement and the operations to beperformed-'are-usuaily -spee:i;.fted·in the overall inst:t'uc~
tionS whieh-mtlSt"be-'g:brentothenaehine-6perator 0 .... Consi1ier- these instructions as being .listed 'on a set of filing 'cards
SO!g.e 'eards will
hold numbers to 'be used in ·the ealeul8.tions~other -eartls.'will hold written orders 'coneerningthe-opera:tions-tu-be'carried'outus1ng"numbers
appearing un -thef'1rs-ttypeJ)foeards ,and sti11--other'~smight be
blank in anticipation of/the- S·torageOfHComputed -results 0 We shall corl.sider -these -cardsas-bein:g-'arranged-1-n'a'logical- sequence 9' -w:1t h each
caro-being-numbered'C-onSeeutively
'I'hetota.iityof . thes~eards~;"" consisting of orders, numbers 9 and space for results -- make up the storage element o
0

A third component of oursimplesystem-i-sthe~n operator
who effectively a:etsasthe c"ontrol element. Thegem3ral fUnction of

the control element is to inspect the cards of the storage element consecutively~ carrying out the orders specified thereon~
By restricting
the complexity and standardizing the form of these orders the intel-'
lectual requirements of the 'coiltrQl:~lelhent'-can;;oe made rather small;
the basic requirements are that it inspect the cards in sequence~ understand the written orders thereon,9 and be able to carry out the simple
physical operations needed to implement these orders~
If we consider the controlJ? storage :;and arithmetic elements
as being physically isolated ina room,9 then a means of cozqrnunication is
required between the room and the outside worldo If all of the cards
of the s tora,@' element were initially b1ank9 then an obvious need is a
means- 'of transmitt-ing- into ,the sterage-'" element ,the required· contents of
these-cards -- the-orders and numbers need-edfor ·theeal-e¥ation •. Furthermore', when the calculation-- has- 'been-C'ODlp'letect, a -means, of transmi tting resul tsfront the-room"llre-'1leetJs·-sary~ Tbeapparatue -necessary
for the purposes of-eormaunieation'Cau'be'generalized into two types-input and outputo An input element and 'an output eleme~ constitute
tlroadditional-needs ·of a general-eomputing-systemo
A schematic representation of the comparison between the elements of manual computation and the elements of a eomp~ter such as

Report

R~203

-2$Wh1rlw1Rd is given in Figure 2. Without loss of geRerali~y, the storage
elemen\ in the human system is represented in this figure by a Bot.book
rather than a set of cards o The analogies between the individual basic
elemen.ts are given in Figures 3, 4, 5 and 6/)'
20

Handling a Simple Problea

As an illustratiOR of how the five elem.ate discussed in the
previouS seetio. can be combined to afford automatic calculatioR, a siaple example will now be cORsiderec:io In the previous section. th.idea of
restricting the RUMber and standardizing the fora of the computer 1&struetioRs was mentioned; this idea will now be utilized as am attempt
is'made to devise a prograa tor the foraation of the expres5i~!! ~
~9 where~, ~9 ~ and ~ have beeR assigaed specific numerical values.
For convenienee assume that a is written or stored on the card Dumbered
9, 12. on card 10, So on card ii, and s! on card 12, while card 13 is assumed
to be available for the temporary storage of first a partial result aDd
theA the final result.
In foming the expressiolll !:11,...f-e...9., the first step Jlight. be to
This is simply accomplished by the multiplication of ~ by ~o
Thus our first instruction -- the iRstructioD on card 1 if we arbitrarily
decide to start there -- would be to the effeet~ clear the accumulator
of the arithmetic element ',make it cOJltail\ zero) and then add into the
accumulator what is written on (the conten\s of) card 96 This iDstruction can be abbreviated &5:
form~.

ca 9 (clear accumulator and add into
it the cORtellts of card 9)
0

As a result of this order 9 ~ is placed in the accumulator, and the next
step would be to multiply it by.!?,o Thus the instruc-&ion Oil card 2 'Would
be to the effect: Multiply the contents of the accumulator by the COBteats of card 10, leaving the product in the aecumulatoro This order :rDi.gb:i
be abbreviated asg
mu 100

The next step in the co:m.putatiom would be to remove the product .!2 from.
the accumulator aad store it tell.por~ily while the product cd isformed.
The temporary storage ,of .!!. eu be accom.plished by clearing erasing)
card 13, and then iramsferring the contents of the accumulator to this
card of the storage elemen~o The order for such an operation 'Would be

-r

a=t

ts 13.
On cards 4 and 5 would be stored the instructions for finding the product -.......
cd~
ca 11
au 12(1

A-35676
F-I028
SC-207

BASIC COMPUTER ELEMENTS
comparison petween manual computation

r'~
~
NOTEBOOK c;;;;;z::::z-''t::1 =bJ!
OPERATOR

DESK
CALCULATING
MACHINE

incomi,nq
problem

outqoinq
results

I~----------------------and--------------------~
WHIRLWIND I computation
STORAGE

problems
to be done

.......-.....

CONTROL '

li--..-.. ARITHMETIC
ELEMENT

f7.,esults in
deSjred form

A-36233
FI034
5C-2.67

arithmetic
element

bDSic computer elemehts

ARITHMETIC ELEMENT
of WHIRLWIND I compared
with a desk calculator

----

~cr

RtCClvec; number from the main busHolds addend, s.Jbtrohend, rrlultiplicon:i
ord ivisor

The adding unit - Holds the result of an

addition t multiplication or subtraction
.and the dividend in division
Auxiliary register-Holds the multiplier
or quotient

DESK CALCULATOR

ARITHMETIC ELEMENT
PERFORMS ACTUAL ARITHMETIC OPERATIONS
SUCH A S -

add • subtract • multiply • divide

A-3623'
IFI032
\5C266

basic computer elements
control

CONTROL
Just 8S in manuel computation
the operator controls the
steps to be performed. so in
WHIRLWIND I computetion

CONTROL
tokes next instruction in sequence from
storoge • examines it and sends pulses
Bt the proper times to the v~rious parts ~
of the computer to perform the n. ecessery
."<~~
PROCESSES \~'\
! '
\
.
/,1 ,
'\
instructs'"
1/
\\\
ARITHMETIC ELEMENT to
instructs STORAGE to
,.
/,,/
\
CLEAR REGISTERS·
SELECT A STORAGE REGISTER
\\

1/'

AIID READ III OR READ OIlT
EITHER III1MBERS OR INSTRIICTIONS ~

\ \

ADD· SUBTRACT· MULTIPLY
DIVIDE· COMPARE

oU-rpu-r to

Instructs JJ\JPU-r to

instructs

SELECT EXTERNAL DEVICE

START -

READ

STOP

RECORD -

STOP

basic computer elements

storage

STORAG E
the memory for initial
data, instructions, and
intermediate results

•

NOTEBOOK STORAGE
in
manual computation

•

MAGNETIC-CORE STORAGE
in
WHIRLWIND I

computation

BINARY DIGITS ARE WRITTEN
AS ONE OF TWO DIRECTIONS
OF FLUX' IN FERROMAGNETIC
CORES

Report R-203

-29To the product gg,
stored on card 130

This

nqw

added the previously obtained product ab
is obtained with the ord~r
ad 130

~dition

With the computationcomR1ete g the result can be stored on eard 13 b,y
another application or tije !! order, this time as
ts 13
Havingeomp-1Etted' the 'eva1Wttioo' 1)f .the- ~res'ftion'~ +~,; one' might
then wish ·to·proeeed '.' toaftOther-'eomptt'ba·t·i:on,·asf'-or "example' oae whose
first 'order was"to ·be·· found· .on' card· 500 Inasmtteb· as-t'he 'eon~ol element "ha-$' ··b$en-··automatieallY·pl"oeeeding· . ·f'romorder·to .-order" (eard to card)
in--sequenee 9 the eontrol eiement" -JlUSt now be in8-truet-ed to $top and th~n
recommence its actj.,onateard '500 This can be done by a special order i
for this purpose, which qM11 be abbreviated as
sp 50
The cards and their re1uired contents for the program of calculatIon' '-outi1nett ·-on·-·the-· pI e viOlISpages---are--:-'"
~.

1
2

Contents.
-ea9
mu 10

ts 13
ea 11

4
;

7
8

ts 13

sp 50

. _Action
~ in accUmulator
~ i~ accumulator

''!R--or1 card 13

ein acC'UDlllla tor

-ed--ill a~tor

~+ ed' i'nti\ccumula tor
-ab +'ed--on ~ard 13
control element reverts
to card 50 .

a
b
c

12
13

d
.
Initial contents immaterial

•
o

start- of ne:ztt-

pro~~m

of calculation .

Report R-203

-30might be used to ~- i-ll aivalue of !. from teletype ta pe and store it
card 9, while the order
re 13

might 'bg'used"w'1:e£Ord-frOm;-eard--13 -onta·t-el-et,peoIna~ as there
~ight'be1Jeveral

types ·.of-tnput··ai1d·output··'-eq'd.pment·"8"Iafi~ble, prior to
the use of either an m.or
order one would 'Use an additional order to
sp~eify the ~rticular equipment to be used.
As an example, the order
eu 29

.ight be-used to prepare for the
bereti-29o

subsequent~se

of an

~xternal

ynit num-

0om!nglltS'on-C-omputer-'Orders
--

The 'orders- ·described· 'i.u·-the pl eviOttS·-sectfon-are-·-llOt ,of a parttcuia-rly--cromplex·'.na tUl e·g--inIt-mere.iy""·requtre 'Of"'~the"eolltrol'~leJJlent' th8. t
a., bility,-to--perf'~'''B~l-~mp-le--taSks-o . In·-raet-·these-orde~~. Jare Qf such
a 'si-mpie- 'nature ·that ·a· '-dull .or"'l1lrl:ma:ginative' pel SO!l '-would --s~:f!e'e as the

eontrol "'elemento

As-mtght·:bEt:"expeeted :rbecause····-ofL·t~'~stalldti;l'diza tion and

simpltc1vof --the -orders ·tt·is·-pessitrle,·-tQ'· s;y nthes isea satfs'ta'c tory contrOl' element: ~ly;"frcml""-e-leC-triCal'--or-" meebanicalcomponent" o· It is this
fact ~-e-oup-ledwtth-the--eleetronie'and' meehani~l-reaIi:zatio~ of the other
basie-elementS' .--st-onge, aritbmetie,input,and' Otttpttt..,.- ,the. t JnB.kes
pose-ible theeonstruction 'or-an-a'lltomat:i:e -eomputer'oFolf~venience in
llli.S and .the suec-e. edlng,---sect.
. _:i:onthe~"erl$
..
e-.·ef-.·a.···,-.'ti1rtactqrU)"_ au~o
~tie .system-. ··of- ·tmr·'fivlr'-mrle-''e'lementa-shall'' '-be-' postttlated-wttqout
'pecificatlcnot'Constl'uetton·det;a11s--o 11t""~oo'oldance with-'e~r~ct term.
tnology, we-, shallrefer-·t-os-tgrage ~-t;i.m1' '-or regi-st-ersra'tUler than ca:n:1s,
witb each storage position or registe~ being assigned an address eorres- .
pending-to .the· previO'tUitl:r'""' each of whieheallNtake either a. favorable or
uJtf'a,vorable state, the. there would be 2 stat ••e.ts ud 2N
differ ••t actio.s o I. maay cases 9 a good deal of these co.ditio•• aRd
actio.s will overlap or CoiRcide» but ia a.y eve.t it is obvious that
the probleM amd its solutio. stated i. terMS of all possible ~oDditioB&l
statellellts "ltd a.ctioae would be ~UlI.berso..eo A Beater altd more co.cis.
0

000

00 OJ)

a..

pnrasillg i$

u~ually

possible with a flow diagro o

Report R-203

=35A aillple exaaple of a flow diagraa is give. 1-. Figure 8,p wl1ere
actioR Wl is take. if eoaditioR. A &ad B exist, whil_ actio. W2 is tak ••
if A does aot exist &ad C doe. exieto Aa aoted» the flow diagraa ia 1&cOMplete i. that it does aot i.dicat. what occurs i f B or C do aot «Xisti or what happeas after aotio•• Wl aDd W20 It is se.a that a flow
diagru is aore cOllv••i ••t thaa words ia describiag the actio. to b.
tak.. uadsr stated co.ditio•• o
G~.erally speakiag there are a .UBber of differ••t ways ia
which a probl.. ea. b. haadled ~ aJld the state_at of a flow diagrua. .'Us"
be based oa the assuaptioa of a particular ••thod o The aasuaptioa of a
.ethod. or solutio. aRd the co••truetioa of a flow diagraa auat i .. tura
b. based oa a kaowledge of how the COMput.~ would proceed to ••para'. ..a
distiaguish separate eo.ditio•• aad how it would e,rr,y out the de.ired
actio•• by aeaas of the available ord.rs o

The .ext step ia prepari., a progr~ for the coaputer is the
process of codi., or tru. slat illg the prohl.. !ato th.e ac\ual IUl.chbe
laaguage -~ the epeeifieorders aDd auaberso The.~ result of the cod~
iag process is a coded proer'" for the co.put.ero Fro. this coded proCK>
graa the storage space and oper,atlJl, tia. requir •••tl for \h. procrca. be deteraimed o If it is fouad that the solutio. require. more .tarc
age space or operating tim. thu ca. b. allocated to the probl_g a ,••W
••thOd of approach» a ••w flow diagra.g .ad a .ew coded prograa .u.t be
ad_.

Although the process of codi_, eaa be do•• - i . all degr.es of
COMplexity» less sophisticated prograal are likely to be quite uaecoJ\oaleal i . storage space
It has b ••• fou.d that .a experi••ced progra..er
caR usually effect a 20% saviJlg i . prograas coded by R.weo••raj) th•••
saviRgs beiRg made possible by the kaowledge of special ••thod. a~
tech.iquesQ An experi ••c.d progr~er is usually able to disoera part.
of prograa. which have so.. aaouat or siJIilarity j) alld irt suell. ca... • •.
does aot use separate orders for each of the8e siailar portio•• , but
rather he use(l a subprororU! or group of orders which eall b. put to joiJatJ
use by various seetioa. of the ...ia pro crill, 0
0

I. S'tllIJIJll7' 9 the ,~nl.ral require••• s for usi., a coaputer hI.
problell. a.re i;wo-foldg firstg the probl•• MU.t be allalyzed ..ad a proper
, foraulatio* of the solutio. must be aad. i . teraa of coaditioaal .tat.•••ts or a flow diagr~ -- that
the probl•• aust be reduced to a
logical fOrIl.ulatioll; •• ~oJldly.9 the flow diagraa .ust be ••chuiz~ ill
the laaguage of the coaputer with due regard to the capabilities of the
computer ..lld its associated iltpu\-outpu\ equipa••to The two phases ot
the pl....img for the use of .. eoaputer ••tail eo.siderable aBAlysia aDd
8tudy~ the first to •• sure that all possible coaditio•• are met a~\
treated; the secoBd to obtai. favorable storage amd tia. r.quir....ts
.ad to easure proper fUActioai_, ot. the prograa ia ter.as of the eoaputer order"o

i.,

Report R-20.3

.0
Actio. W2

Illustrative Flow Diagraa
Figure 8

Report R-203
-3713

The Whirlwind Computer

R!Feseatati01'l of hformation

In accordaltce with the discussion in t~e first. part of this
the information stored in theWh1:rl'W~d. Computer relates ta the
ope:rat1als to be pe:rtormed, the.ddressea ofparticula:rstOl'ap positions
used ~ these operations, and to data retain.' in. earl.in storege positions;, Direct digital representatioRs are use.d f-or the acld:resses ad tor
the r.r:ume:ricel aata, whereas a cooed numerie,J..represelltatien is ued for
the va1'iousoperati0l18 which· are otherwise specified by two-letter abbreviatiolls
The computer, can, in etfeet,storeaD.d bendlenoD.-llumerical
qual'lt1ties -- for example, alphabetic iDfor_ticm. -- i t al!lappropriate
coded numerical representation is ttsedo
.

.~pter,

_ The number system .ehosen for the dis;l.tal represeatatie in Whirl-

w:iltd is that with the base (or radix) 2.. A, eo:m.trasted with the more
tamil:i..ar decwl. system.
oo~o

'Whi~h

uses numoers1'tb.ose di{d..ts may be 0, 1.1 2, 3,

8, 'and 9, the binary systea 'employs 'oniy the digits Oau;l.ID.

either system the magilitude a:ad position of • digit spee1tlesapa:rt1cu.lAar
mul:tiple of a power of the -base o For example, the decimal number 203.9
is taterpreted as

The bmary number 1011001 is hterpreted in
1023+ 0022+ 1021

similar fashion as

+1020 +002-1 +102- 2

For purposes of convenience in reference, each piece of intO$-

a:tiOl1 retained in a storage position is termed a yord. As uted in. previou.s a.ections there are two types of words ... - one is 8n order or instruetianWhich consists of a specified operation ead
associated address t the seeond is a :aumber or piece of data
'Words _iJl Whirlwiniare
16 'biRa.J7 digits in length; the assignment of these 16 digits for Ordelrs
and l!l.'WQbers is shan in· Figure 90
' ~

r~presenttng 8

number the first

~it

pOSition is reserved
8S numerical
digits with the binary point at the lett • Thefollowitlg cOl1ventiollS are
elJi'ployed: i f the :nuuiber is positive, the sign digit is ! aad Il"ll other
dig,its eorrespo11.d to the biMry,-dig1ts of the Dumber; i f the llumber is
negative, the sign digit is a 1. and the diS its of the word are the ~o.- .
plements(Q. for !, 1. far 2,) of the posit1ve_pitu.deof thel'lumoer.

for the sign digit, 'While the other 15 digits are cODsidered

This particular represet).tatiom. of n.gative !lumbers ·~8la10WD. ·as the
"nine' a-complement" and is used for eOllveaiaee in ar1thmeticoperation.
It should be noted tutthere are tworepre8ent8t~ons of zero:

+
-

0 : 0000000 000 000 000
0:: 1 III 111 1ll ill 111

A-36237
FI039

WHIRLWIND I

W0 RD5

are
and

held

in storage registers
represent either

NUMBERS

INSTRUCTIONS

number

101110101011111011101010111011111
~,

SIGN

NUMBER
.55112

instruction

I0II I0I0I0II II I0 II I0I010 II I0II ] I I
,

OPERATION CODE
TRANSFER TO STORAGE

STORAGE ADDRESS
REGISTER NO. 1615

CONTROL DETERMINES WHETHER A GIVEN WORD REPRESENTS A NUMBER OR INSTRUCTION -

Report R-203

....38With the represeatatioa used and the positioR of the biaary poi.~ at the
1eft9 the .ua~,rs which ea. be stored are + 0t -0, .ad all iategral .ultiples of +2- , betw••• ..., 1+ 2-15 .ad 1 - 2-1
The choice of tll. po....
sitio. of the binar.y poiRt at the left .ea•• that .ultiplicative CORst. .,.
or seale factors auet be associated with the storage of quaatitie. ou\side of the ra.ge "'" 1 2-15 to 1 .... 2-15 For exaaple if a auaber of'
a.glee raRgi., b.tweea 00 aRd 3600 were to be stored they could be stored
as e/720 o by the use of the seale factor 1/720 0

Th.. r.pres ••t~tioR of orders or iRstructio.. is show. i . t ••
lower part ot Figure 90 The first five or left-haRd digits of aa order
,a,re used im the biRary=eoded represeJltatioJl of aJl operatioao The.e five
digits suffice for the defiaitioa of 2; or 32 operatio•• o The r ••ai.i-c
11 digits a,t the right"",hamd eRd of the-.word permit th.. specifieatioll of
211 or 2048 addrees85 of storage positio•• o

RefereRce to Figure 9 shows tha.t orders have the Baae appear=
each beiJlg a 16-digit biAary wordo Give. a
word frOM the i.terRal storage of the co.puter, 0•• could Rot ideatify it
u.iquely as ..a order or a auaber; it word obtai•• sig.ificanee as aa order
or Ruaber dep.RdiRg UpOR its positioR aad use ia a co.put~ prograa o Aa
.iaportaat result of this fact is that arithmetic operatioBS CaR b. carried
out upoa orders ae,well a5 .UMbers aakiag it possible to chaRge the address seetioRs of orders .ad heaee obtai. greater versatility ia the U••
of these ordereo
aRce as aUlilbers or datit9

ElectroRic Aspects ..ad the Storage El••••t

The tuadaaeRtal reaeOR for the us. of the biaary sy.te. of .otatioR is its great eo_veRi••ee ia electro.io ...ife.tatioJlo The fa~t
tha.t a,lIllY digit ill the binary systeM ~a.. oBly be a Q or a 1. e.ables o.e '\0
repres ••t a.digit by a deviee or .ea•• capable of two di.ti.c~ aad dieti_guishabl. modea of operatio. o The static repre.eatatioa or storac. ot
a digit i. possible with a flip=tlop» a bi-stable eircuit of the Eecle.JordaJl typeo Stora.ge of a digit is also po.sible by the pre.e.ce or
abee.ce of a charge oa dielectric surface TAe ~aaic r.pr ••••tatio.,
or tra.e!er of a digit is possible by OR. or two COBditio•• o. a traasllissioll lilte -= either the preeeaoe or abe ••ee of a pulse, or the pres.ace of 0 •• of two differe:at voltage eo.ditio.s o Through tn. use ot
electro.i~ ele.e.ts im ess.atiallyeither a. 2! or aa 2£! state, lar,.
sip&l-to.,;,....oise ratios a.re obta.i••d with a C orr.,polldi.., illereas. ill reliabilityo
0

The i.ter.al operatio. of the COMputer is orgaaized aad carried
out by th.e traasferj) routia" ..)ltd storage of p'ulses or baary digitso Ia
pertor_illg the lU.jority of the computer operatioRs, a .umber of associated
pulses must be dealt with =~ arithmetic operatio•• requiring KaRipulatio.
with 16 di~itso aae of the factors beb.ud the high· speeds obtai.able h
Whirlwiad lsee page 11) is the si.ulta.eoua tra.saissioa aRd operatio.
upo. digits by parallel ide.tical cha••elso This is to be contrasted to

Report R"",203
....39...
a possible serial

lI.anipul~,tion

using but a simgle chanel.

The arithmetic and control elements of Whirlwind are compounded
from flip-flope a,nd other st~nda.rd pulse-technique circuits. storage reg...isters built of flip-flops are used in both of these elements; however,
the cost of such registers precludes their use on any other than Q sm~l
sC::lJ.e!) The 12rge-sca,le inter};u~,l storage mediurtt in \r~hirlwind is composed
of ~pec~a.lly-desi,~?led electrostAtic stor~lge tubes ~ Bina~y digit inform.?tlOn. lS f;tored ~:n these tubes ~a.s charges on £t dlelectrlc surfc"ice.
The design specifica.tions of the storage tubes call for a tube
cc?pa.city of 1024 (210 ) binary digits, A parallel stora.ge system is used
whereby e&ch digit of a. word is stored in a similar position in each of
16 tUbes" In this fashion, the deflection pla.tes of all 16 tubes are
operR.ted in para.llel and a complete word can be re;'ld into or out of stor~,ge in the sarne time that is required for the stor.1lge or recovery of &!.
single digit~ IBy using two blanks of 16 storage tubes each, a total capa.city of 2 x 2 0 or 2048 words can be achieved/} This corresponds to the
existence of 2048 sixteen-digit registers
o'

In their final form it is expected th~t the storage tubes will
the storel,ge or extr~,ction of Ql. word in 6 ).lsecol'lds. At the present
til.le neither this projected storage access time nor the plarln.ed storage
c~1,pacity hcls been achieved, although progress is being rna.de towards these
goa.ls o The storage tubes now permit access times of about 20 - 30 )lseconds,
while the opera.ting ca.pacities of the tubes are 256 digits each, \"Jith
only one bank of tubes installed, the existing capacity of the storag.e
element is 256 words~ ~?he cornputer has been operating quite satisf,lJ.ctorily for a period of six months at this capa.city, €I.nd present plans are to
double this cCl.pacity within a year~
p~;rr~1it

The CORtrol ElemeJlt

The genera.l functioning of the computer is best described as a
continued repetitio:tl of two basic steps:

obtainiJlg the next order (opera.tioR a.nd address) to be
perform.ed

carryiRg out the operation specified i. that order.

and

The control elememt of the comp~ter 5upplies the .ecessar.y
pulses for both of" these steps 9 and ca.rries out step a) itself. Step a)
is ;,lways the
e ~ a.•d relates to fiBding out where the Rext order is
5tored~ obta.iBiRg this order from electrostatic storage, and prepariRi
the machiae to car~y out that order~ The second step, that of c.rryiac

sa..

lei Reference 28

Report R-.203

out the operation, varies according to the nature of the operation and is
carried out by the control element in conjuetion with the arithmetic" storage, input. or output elementsoUe~

StE?P a) above is per.formeCi in 8 useconds plus the time for one
storage accEJss; Step b) usu.aJ.ly ~olveB at least one storage access
plus anytlhere up to about

tlSeeo¢s for the longer arithmetic operations,

A:t the preSEmt time the average colUPuter order is performed in between 50
and lOOusecondsl :in the futu;re
20 and 30 U$econds

ttJ-is

time should

be reduced to between

The Arithmetic Element

The chief components of the arithmetic element are three· 3.6f'J..tt.p-flop registers. Theser,gisters - The A register (AR)p B
Register {BR), and Accumulator (AC) -- have a str:iki.ng eorresponde.nee to
dgit

the keyboarq and counters of a desk ·calcuJ.ator, The AR is butfer register- by which information is fed into
aritbmetie element; while the
AC performs the bulle of the arithmetic work. The BR ae~ las a part of
the AC and i,sused chiefly for mttltipl1cation and division.,

the

''!he basic arithlootie abilities of the· arithmetic element are
addi tion; subtraction, mutiplication; di,vision, shifting (multiplication
by, powers '. of 2), and point-off (finding the eharacteastie of log2X. The
aritfmtetic orders of the, machine; - and it has alr,ady been po ted that
the· choice of the orders is based chiefly on oonVeiience and flex:tbili ty
--~so have certain added features dealing
round-off$magiUtudes;

v.th

etc.

There is a considerable amount of fine detail regarding t1le .
effects of the ari tbmetic orders and other orders used in the maehine9
these affects relating to :matter of electrOniCs, lpgiealnecess_J) and
convenience; for reference these ordeTSare summarized in Appendi;x 1.
Literature dealing with the use of these orders is listed j.nthe BibliegTaphy~'I.

In:eu~tput

Information is introduced to and extracted from whirlwind primarily as electrical signals. At the present time the input and output
equipments are fed and receive their signals from several flip-f;Lap reg'isters, this arrangement is omy tempory and in its final,for the
link between the conputer and the outside 'World will be a:;general-purpose
input-output register; together ldth and in-out,mdtch for selecting various pieces of terminal equipment.
A large variety of input and output eqUipment has been considered for use with' Whirlwind (see Figure 6)~' Up to quite recently the
effor~ have been directed towards the use of photographic fil.:m and
p1mched paper tape6 Extensive use has also been made of cathode ray

1. References

25,

Report R=203

-41tube displays aBd a typewriter-printer as output meanso The paper tape
equipment used is similar to that employed in -standard teletype app1ica-tions~ and the procedures and conventions employed are such that ~ect
input from or output to teletype equipment could be easily obtained. The
flexibility of the computer makes it readily adaptable for use with other
types of equipment $ the choice of external equipment depending chiefly Oil
the applications for which the computer is to be usedo For use of the
machine in general non-real time applioations, magnetic tape units are
now being prepared for use with the computer; the use of the machine for
real-time applications will probably utilize a magnetic drum either for
additional external storage or as a buffer storage for information from
asynchronous sources o
The punched paper tapes are used both for the supplying of the
initial contents ot storage -- orders and data -- as well as for data
needed later in calculations o The slow, mechanical paper tape units permit special teChniques in the reading-in and reading-out of date, and also
allow rather thorough checking of the input data o Further conunants on
this subject are made in the succeeding chapter where -the input-output
problem is discussed in relationship to the air traffic control applicatiolTho

Report R=203

-42CHAPTER IV
Input=Dutput Considerations for the Computer-Controlled System
Although the general purpose of this study is to consider the
mechanization of the present-day system without any ess:ential changes in
the techniques and methods currently employed, one obvious change which
must beoonsidered' is the means of introducing the information concerning
aircraft movements and weather oonditions into the new controlling element of the system--the computer" The problem i l essentially one of
handling and processing large quantities of information A large amount
of information and data must be supplied to the computer in a never-ceasing flow,$) and the computer in turn must provide information. in. the form
of instructions for the control air traffic o The information-handling
problem must be satisfactorily resolved, or it stands as a possible barrier to the success of the computer application~ if information cannot
be supplied to and withdrawn from the computer in the proper manner--with
due respect to speed and accura,cy.... -a computer=controlled system is not
feasible 0
0

This chapter first discusses the data-handling presently involved in the en...,route phases of air traffic controlo A second section
outlines in fairly general terms a system which will meet the existing
requirements~ while a third section discusses certain aspects of the
use of the proposed input =output systemo
Ao

Information-Handling in the Present-Day System

The controlling element in any scheme ~f air traffic control
must have (a) ready access to current information regarding both the
flights in progress and those being planned~ (b) as extensive a knowledge
as possible of existing and expected weather conditions, and (0) complete information regarding conditions at and about the airportso Under
the pre sent..., day system this information comes to the controller from a
variety of" sources~ and in a number of different wayso
Information regarding commercial and military flights is obtained by the controllers over an interphone system from operations
offices established by the commercial ca.rriers and the military services.
These operations offices are in direct radio contact with their aircraft;
they file the initial requests for a clearance (Approval Requests), and
relay flight progress reports from aircraft in the airo Private, civil,
and non-scheduled flights ....""'this class of traffic being termed UitinercantU by the controllers--either subscribe to privately-operated comm.unications stations or use the Flight Advisory Servicel rendered itinerant
traffic by the Civil Aeronautics Administration and the Wea.ther Bureau.
Reports concerning certain a ircra.ft are also received by the center coatrollers via. interphone from ,centers in the neighboring control areas.
10

The !~ight Advisory Service perfor.ms for the itiRerant traffic the
general fumctions of an oper'ations office" '"

Report R=203

-43The last=minute requests for takeoff clearance for aircraft
(Clearance Requests) are received via interphone from the tower operators
when the aircraft are ready to departo The center controllers are supplied with up-to-the-minute information regarding acceptance rates and
s·tack heights at the airports by approach controllers in the airport
towers Quite recently the use of VHF radio has permitted the center controllers at Boston to establish limited but direct contact with pilots
as an aid in sequencing for landingo This innovation has not yet been
made a. national practiceo
0

Weather reports are normally received at the control-centers
by members of the Weather Bureau who collect reports from regular observa-

tion points \rla teletypej) analyze these reportsjI and then distribute the
info!~ation to the controllerso
Further information regarding weather
conditions is obtained directly by the controllers via interphone, both
from airport towers and from operations offices which relay observations
from the pilotso
'
The flow of information also proceeds in
ion 0 Instructions p confirmations, and informatioa
trol of en=route traffic leave the control centers
way to operations orfices~ airports~ or centers in

the opposite directdestined for the eo~
over interphone on the
adjacent areaS G

The amount of information handled by controllers can be quite
depending of course upon the weather conditions and the density ot
traffic
During a libusy session!! at the Boston Center it is not 'Ullusual
to see ten controllers and their assista.nts continuously accepting and
dispatching messages over the interphoneso The situation is further complicated because the information arrives from a large number of independent sources Each controller has access to 10 - 20 circuits, each circuit having u.p to five partieso There is no switchboard to give any or....
derly flow to calls~ and at peak loads a controller usually has a backlog
of unanswered calls awaiting his attention 0
large~

An Outline of an Input =Out put System

The computer can only utilize information in a binary form, and
can generally only initially accept this information as binary....type electrics.l signals o One of the first problems in the input ....output arrangements~ then~ is to convert the information which originally exists in
written or oral form to the required binary signals o Of similar ~port
ance is the necessity to convert the binary-type electrical information
produced by the computer to a form recognizable and usable by humane o
The conversion of information to binary electrical signals and
the transmission of this information to the computer site can probably be
best accomplished with a minimum of expense through the use of existimg
t~letype facilities and techniqueso 1
10 Reference 29

Report R-203

=44=
The introduct,ion of information to the system would be by means of teletype typewriter units situated at the locations present~ connected by
interphone to the control centers-~that is, at the airport towers, operation~ offices, weather observation stations 9 etco;
the present system of radio communication between the aircraft and ground installatioa&
would be retained o Messages introduced into these iBput typewriter units
would probably not be directly transmitted over the teletype system, but
would be temporarily stored on punched paper tapes (see page 47)0 The
informa.tion on these paper tapes could then be read to the teletype system by means of mechanical or photoelectrio tape readers o
Considering first an individual input unit (typewriter) liBked
directly to the computer, the factor of speed evidently is of major importanc9 o At one end·of the link is a human operator capable of typing
only a limited speed=-the limitation either being of the human or of the
input unit itself-=and at the other end is a high-speed computer which
operates most efficiently wheft accepting data at a rapid rateo Some
speed-up of the input can be obtained through the use of paper tapes
and photoelectric readers, but the speedsor these readers as we1l as
bandwidth considerations in a low-cost communication link wi1l Dot permit the improvement necessar,y for efficient utilization of the·informatio~ with high computer speed8~ and it would appear that a buffer storage
device must be usedo Simila,rly~ in satisfying the requirements of handling information flowing away from the computer9 a storage medium is
needed as a buffer between the output of the computer and the relatively
slow output typewriter unitso
The buffer storage medium which is to provide the necessary
flexibility in speed should also be capable of providing a. mems of
channeling and sequencing the flow of information arriving from a large
number of independent input units o Although on the input end there is
no question of the destination of the messages--all messages going oBlY.
to the computer==it does become necessary to provide a means of routing
the outgoing messages from the computer to" the individual destinatioRj
this function could be handled by the buffer storage unit 0
The most suitable buffer storage device, from the viewpoint of
availability, relia:bilitY5> flexibility.9 and cost, would seem to be a
magnetic drum on which information is stored by magnetizing small portions of the drum sur.face" A drum presently being marketed by Engineerag
Research Associates has the following characteristicsg
a)
b)
c)

speed of rotation of 3600 rpm
2048 angular storage positions aroURd. the drum.
80 channels along the length of the drum, each
channel being provided with eli. reading writ·ing
head o 1
'

Report R=203

=45=
These 80 channels permit a capacity of soma 160 9 000 binary digit~ increaan
capacities are easily obtained by using longer drums o The drum can also
be efficiently broken down into smaller units by the appropriate grouping
of indiyidual channels into fields o
Informatio:n. can be stored on. the drum in a number of ways; iRformation can be introduced in a parallel manner with the digits of each
piece of information each being stored in a different channel but all at
the sruma angular positions» or information can be stored in a serial man~er with each digit using a successive angular position in the same channelo Techniques such as these would be employed to handle the messages
arriving from the individual input units, in particular it might be desirable to assign a field on the drum to each input unito
With the speed of 3600 rpm» information can be introduced to or
read out from. successive registers or angular positions on the drum every
8~secoftds» a speed even faster than that at which the information CaR
presently be inserted or withdrawn from Whirlwindo However, by m.aldng use
of nom=consecutive registers and by utilizing some of the drum channels
for control informatioB» almost any desired reading and writing speeds CaB
be obtained o For example£> pieces of information c,ould be introduced into
©onse©utive registers once each drum rotation (16 milliseconds) and could
then be read out during one rotation of the drum at 8 )lseconds btervaJ.s o
It is this type of operation which ©ould be used to permit the desired
speed increase in information coming from the input units o
A ~chematic representation of the major components of the projected input system is shown in Figure 10 Informatioa introduced iRto
the system would be processed onto paper tape and would then be read to
the teletype system and transmitted to the computer siteo The Drum Coatrol would supervise the reading of the information to the appropriate
fields and registers of the drum» and indications would be given to the
computer by the Drum Control of the times at which the computer may pro'ceed to withdraw i:nf'ormation from a selected portion of the drum. o The
automatic and direct insertion ot the information to the computer through
an input=output register would be under the joint control of the computer
0

and the Drum Controlo

The read=in of information to the computer would be done periodica.lly at about one-minute intervals This periodic read-in would provide
that no message would suffer too long a delay on the drum On the other
hand~ in order to handle expected variations in message traffic the computer would also be programmed to execute a read-in whenever a sufficient
amount of informatioR==usually a sufficiently large block=~had beeD accumulated 0 Signals indicating the presence of such a condition would be
provided to the computer by the Drum Control o
0

As noted in Figure 10 there will be a number of remote loca.tions
each supplied with input units o As mentioned in Chapter 19 it will probably be necessary for a loca1 operator to insert control instructions and

input
units

paper
tape
readers

teletype
conununication
link

Whirlwind.

At Computer

At Remote Locations

Figure 10.

~put
II

System

Sit~

Report R=203

-47information into the eomputereThi" wou;Ld. pro\labiy' be go~e 91 a special
input unit at the computer 51te It_ will probably also be -'necessary for
there to be a speoial channel or means ot eo:nmnmieating, with the computer for high-priority em~~g~ncy mess,~ge~.
0

The arrangements for the o~tput ot information from the computer would be quite simUariio those indicated_.£or·the--illput,·--rnd the same
equipments would be used. ' The computer would aCctmmlat'6 a--~ta1n amount
of inf'ormat1.on, and then would perform ahigh-speed-:'readii.. '" ." '" ~ . ..... H

Ch+q!til'..

As noted in' Chapter I, ':·!.n;·cmter·that thelIS'e of'acoDlputer as
the central elemellt.··o~:a1'1'·:a;1~ ~~~~~:~~,.Stemreach··arrr·de-~. of feasibili ty, the rel1ab11ittof'''·'iH,'o~empl1ter·-·must 'be . extremelr--hi$h. It is
similarly imperative- ·that:'·:\he:,:·\;i..t~output-~-system pOS'S'dS a" 19-igh degree
or reliabili ty ~ and ·:turthe~..,\*,!t:~"1t:·:"bEr"eperated in an error-tree manner
with great eare being' taken to'8;wid" those-errors due tohumaD operatiOD

as well as errQrs arising trom deterioration or failure of mechanical or
electrical cemponents.
I.. large-~'-of':~~()ds"imvebeen '-developed to improve the
reliability ot eODnmmteatiomJo,~tnciitde the retransmission of the
informs. tion over- the·sa_:oxr·"~ltea-t.taeilitfes-andthe tra:q.smission ot
the information-in'-'eodedf~'iw1rl-ch-are"'htgh1T' sus·eeptible to checking.
The problems which- are-:'encQuntend-a-re -quitesimilar to those which arise
in conneotion with the iriterna'l reliability' or computers, and similar
steps and techniques will have to be employed to obtain satisfactory operationo
1

Cer~in precautions MUst also be taken with regard to the human
element associated with the terminal· equipment. Extreme care must be taken that the messages be correct ineontent; and that they be properly introduced into the systemo Such precautions would probably ta~e the form
of repeated handling of the messages during their preparatioD"and typing.
This handling would probably entail duplicate or check typing of the messages; a first typing would be Used to prepare a tape against ,which a
second or duplicate typing could be eomparedo This increased message
handling should in itself greatly reduce the number of message errors and
prevent improper preparation or encoding of the messages.

It is always possible,hewever, that compensating errors in the

Report R-203
transmission of the message will permit an error to enter the computer uadetected, and it is similarly possible that the original message upon
which the first and second typing were based might hav.an uadetected
error
In either case~ the resulting errors in the messages mal be of
an obvious kind or may be ra.ther subtle and difficult to detect. Aa
example of an obvious error would be an air speed of 800 mph rather thaa
300 mph; an. example of a. more subtle error is the filing of a requested
altitude of 7000 feet instead of 9000 feeto
0

The obvious type of error can be detected by an alert human operator if it arises in the initial m.essage,; similarly; the computer may
be programmed to spot such errors which arise mechanically or electrically
and which escape the measures used to check the transmission. The subtle
types of errors provide a great deal more difficulty in their detectio••
Such errors are kept to a minimum in the present-day system by a large
amoumt of repetition in successive messages pertaining to a single flight.
Th~.t is to say, each message not only contains the pertinent informatioa,
but is a.lso likely to i1\01OO8 a certain amount of informatioa which is
redundant A particular example of this is the repeat-back by the eontroller of all the details of a flight plan even when the flight plan .as
heeR approved without change
If this type of operation proves to be
rlteCe5sary in the computer-controlled system. it can easily be . . in.corporated..
0

Conversion.

The input units mentioned earlier in this chapter permit a direct translation from oral or writteR iDformatioJ). to electrical sig.ale of
a binary Rature o It is necessary, however, to specify the relatioaship
between the oral or written iYlformatio., the binary informatioa produced
by these unit5~ and the 16-digit binar.r words stored by the computer.
Most of the information in air traffic control communicatioa can
be rather easily abbreviated iDto groups of three characters, each character being either a letter or a decimal digitI) Upon this ba.sis, input
messages would be made up of sequences of three-character groups of words.
Each of the characters used would be translated by the input unit into a
combination of binary digits; if the direct binary represemtation were
used for each decimal digit~ then an airspeed of 185 m.ph would be typed
into the unit as 185 and would result in the flowing sequeDce of biaar.y
informatiolmg
0001 (=1)
1000 (=8)
0101

URfortunately there are 26 English letters and 10 decimal digits,
and this total of 36 eharacter~ would necessitate the use of six binar.y
digits i . the representation of these characters; a group of three characters~ then~ would require 3x6=18 binary digits and hence could Dot be
conveniently stored in a Whirlwind register. The most convenient arr~e
meat for the internal storage of the three-character words in Whirlwind

Report R=203

-49would be to have but five biaary digits per charactero This, of course,
would create 15=digit words g although at the expense of there being o~
32 different characters which could be usedo Fortunately, it appears
that a limitation to 32 characters would not be particularly restrictive
inasmuch as all of the alphabetic characters are not used iR traffieooRtrol m.essages o
Since the ,restriction to five-digit binary-coded peatads is
made necessary only because of the existing register length of Whir1wiDd,
the proplem is probably not of sufficient geaerality to merit detailed
consideratioB o It might also be necessary that the transmissioa of the
m.essages in the teletype system include the transmissioR of extra digits
with each pentad for checking purposes o For these reasons, it will merely be assumed that the binary-coded pentads can be formed aJld transmitted
to the computero This doe.s not necessarily place any particular restrictiOlllLS on the input units or other elements of the input-output system.; i f
desirable, the input-output system may operate with ordinar.y teletype
techniques and the necessary conversion to the proper pented form can be
carried out just prior to the storage of 1.tormatioa on the drum o
It should be realized that the three·c~aracter words made up
of binary=coded pentads represent meaningful information, as do pure
bi:t\a.ry numbers...,=the chier difference being that the computer operations
of additioR9 subtraction, multiplication, .teo~ are meaaingful oBly with
pure biRary n.umbers o One,.. fU"jor reason for a further interaal coltVerSiOll
from binary=coded p~nt.d 10 pure binary form, then, is to permit ease in
arithmetic manipulatioRe
There are several other important reasons, howevaro On. of tn-.
is the saving in storage space o For example, the binary-coded pentad
represelltation for the altitude 18,500 feet, abbreviated to tbreecbaracters as 185» would appear as

o 00001

01000 00101

and in this form it fills a register o
could be represented &5 185 x 2-15 or

5
In the direct binar.y fora this

0000000010111001

In ..ir traffic co.trol work it i~ Dot Jleeessary to specify altitude.
than 500 feet~ and in general there would be no need to go above
a maximum. a.ltitucterof 63»500 fe~to· With these provisioRs, then, a. least,
~ignificaRt biBary digit can be tg~~o8t of as~ repres.enting.... ?OO-foot
~ncrem.el\to There are but 128 (=
+:.) possible altitude levels and
these call be represented by seven. ~~gary digitso If these seven digit,
are considered at the right hand end of the register, than altitud •• are
are represeJlted as follow58
clo5e~

Report R=203
~

o
500

1000
4000
32000

63500

UJlused Digits
000000000

Used .Digitt'.,
0000000
0000001
0000010
0001000
1000000
1111111

am altitude A measured i . feet is stored in the computer as
the binary form of ~ x2-15 o With such a representatioD there are aiR.
URus~d digit3 ia a register which CaR be used for the storage of other
qua.titie~o
Further comments Oil the mechaniza.ti.onof the storage of two
or more quantities in a siagle register are reserved until the.ext chapThu~

tero
The problem of cORversion is rather closely associa.ted 'with
matters of checking ud storage allocatioB0 This fact caB be better appreciated if two aspects of a conversion or traaslatioa proeessare •••tiolted o The first is the decision as to the end result of the cORverslo.
-~that is~ to what pure binary form the oinar,r-eoded pentads should'~.
~oRvertedo
IR the case of numerical iRformation the cORversioa will be
dil"'ectly to pure bilIAry form.~ suitably scale factored o For facility ia
the utilization of the •.vailable stora,ge of the com.puter s however, it become~ desirable to establish a rather arbitrary cORver~ed form for certaia
.uerical and alphabetical quallltities.. For example, the three-cn.aract.r
abbreviated form for the Boston reportiag point might be BOS, and for
cOJ!lveD.ience in determinil\g the sequence of registers containing data. 'pertaining to the Boston reporting point it will be necessary to establish a
particular pure binary equivalent for the three-character group BOS o
{The details of this scheme are mentioned in the following chapter)o For
rel-SORS such as th.ae$) the cORversion .from. the b iaary'-c oded pentads to
the pure binary form is m.ot so much all a.dditioaal, piece of work as u aid
in the programming for the computero
The second aspect of the conversion or tra.slatiOD is tn•. me..8
by which it i8 carried outo If there 18 a d.fimite arithmetic re1atiQaship or correspondence betw~e. the original and the converted fora, the.
it is possible to perform the conversion by arithmetic operatioRs o Ia
other eases,where the correspoDdeDce is not direct but may, actually be
rather arbitrary» a stored table .of values m.ust be used forth. cOllve;rsioa
proces5 o With the table-type conversioR 9 i~asmuch a8 all the URcoaverted
for.ms must be stored i . order to obtain o••-to=o.e corresponde.ce b.tw•••
the two forms 9 it is an easy matter to discover if a particular unconverted form does not exist among the stored values o If such i5 tn. eas.~
~h.R it is reasonable to expect tha.t the URconverted form was a.rror.
~imilarly~ with the arithmetic type of cOllversioll the computer can be
programmed to investigate if a converted aumber fa.lls within. a reasonable
range of values o Examples of the use of such checking are give. ia the
follo'Wing cha.pter 0'

Report R=203

30 Message Form and Standardizatio.
In order to be able to properly convertg check, and th~n act
UpOR the contents of any message the computer must be able to det.r~
miRe what each word (three ... character group) of the m.essage repres.nts~
whether it is an altitude~ speed a reporting point, .te o Humaa co.trollers are able to identify various quantities because of their size»
unitss> or-position in messages o Although similar techniques could be made
to apply to the uee of the computerg it is far more practical to use a
method by which informatioa is identified as to its type only by its relative position in a particular message o
j

At the present time the CAA has no rigid regulations coveriRg
the contents and forms of a'ir traffic control messages o CAA manuals do
specify the llecessary minimum. contelllts of messages, however g the flexibility of voice communication and the ingenuity of the human controllers
have combined in such a way that tbe contents and sequence of iRformatioa
in most messages is varied from controller to controller~ as well as being varied to meet existing conditioRs~ This is especially true of the
addition of further information or comments at the end ofm.essages o
An example of the lack of standardization of message forms is
the specifieatioR of the flight patb'of an airerafto The various reporting points across the eOUl\try have been give. thr......lett.r abbreviatiOlu.sg
BostOll-::BOSj) Chicago=CHlj) etc o . In addition)) a large' Rumber of interconnecting airways throughout the country have been give. rout. designa'tions)) examples being A5» B3)) R120 At the presemt tim. the patlof u.
aircraft a.long the civil airways is indicated by OBe of the following
methoosg
/
a)

specifying the point of departure g tne airways

to be followed g and the point of la.dingo
STL
A;
B6
R12
CHI
b)

Exampleg

specifying the point of departure)) all interveniRg reporting pointsj) and the point of landimg
Exam.ple~

STL
SPI
PIA
BDF
by a combination of a) am~ b)o
STL

R12

CHI

Exampleg
CHI

NBU

WDF

The specificatioB of a path by one of the above methods permits
the.tra.ffic comtrollers to determin.-.... eith..r from memory;? by refereace to
tables ~r mapss or directly in the case of b) above-- all of the report~
iag poiits over which the aircra.ft will pass en.-route o This is a colldi... ,
tiOR which must also be met by the computero Inasmuch as a route desip.a.tioR such as A5 implies a ltUJllber of reportiRg poiltte along that route»
it would be possible to program. the computer to con.vert a path desipated
in form a) or e) above to the series of desired reporting poiRts as giv••

Report R-203

-52in b)o

The storage requir..ents for such a conversion, however, are like-

ly to be exeessive g aad it does Dot appear to be too restrictive to assume
that 3. pa.th will be specified to the computer ollly as a. list of successive
reportillg poi.t·s~ Similarly, the form of all m.e.ssages i.coming or outgo~

must be standardized -- each piece of informatioR having a designated positioa ia each type of message, with particular type of ~ message being
designated by a key word early in the message o

Computer Operation As Regards IRput-Outnut

The present facilities and computer orders for the iaput and ou~
put of iaformatioR from Whirlwind were designed chiefly tor specific application with photographic film units, and as such are not applicable to the
input-output problem under consideration o
.
The input and output of information from paper tape uses rather
makeshift arraRgements as a temporary expedient until general purpose input-output equipment can be designed and installedo The handling of this
problem has been such as to put the b~rden Oft computer programs, with the
computer executing a number of orders betweeR successive read-ins or readouts, these programs process and store the words during a read-in, prepare and transmit the words during a read-out o Such a technique is feasible only under the r~ther special conditions and slow speeds which exist
with the paper tape equipmeRtj since the reading and writing is done at
such a slow rate 9 there is not lllUch of a decrease in the i.put or output
rate due to operatio. of the computer between the handling of. successive
lines OR tape o

With the use of a drum as a buffer storage, informatioa CaR be
made available at a rate which is commensurate with the speed of the computer itselfo In order to meet such a challenge 9 it is felt that there
must be no intermediate action by a program» with its resulting delays9
in the following respectsg
a)

taking the pentads from the drums

performing ~ necessary checking operations on the traas=
m.issioR
assembling three pentads to give a complete word
storing the assembled words in the registers of WhirlwiDdo

c)
d)

Special orders would probably be Beeessary to perform these fURctions most
efficieBtlyo Similar orders would be Deeded to perform similar functioJla
ia the output of information from the computero
The general scheme of operatio. would be to have a 16~digit iB~
put-output register connected directly to the 16 digit columns of Whirlwtrtd 0 The binary-coded peatads would be introduced in sequeJlc. iato this
register which would then perform the necessary shifting and checking opera.tio•• to assemble threGl pelltads iato a computer wordo While the COll.poJleRts

Report R....203

-53of ODe word were beiag read in from the drum &ad assembled ia the iaputoutput registerj) the previously-formed word would be stored ill the ill'\'"
terRal storage of the machi•• o As aa aid ift handling the •• seages l
special key words .would probably be used to indicate the ead of a .eseage o
Noraally» an ackRowledgaeat of receipt of the message would, ia timeg b.
seRt to the message originatoro If an error were detected in the iaput
message the computer would .take steps to sead a message to the origiaator
of the message to perforM a retransmissioa, aa indicatioR would also be
given to a human operator that an error had occurred so that appropriate
mai.taRance action might be initiated o
There are two special cases of errors in the iaput whick
should be mentioned o In ODe case the error might happen to be in the
word which desig.ated the originator of the messageo To guard against
this eventuality it would proba,bly be :necessary to transmit the word
designating the originator both at the beginning and end of a message o
Equally serious might be an error which lost part of a word or which
caus~d a change such that in reassembling the message in the computer
a staggered version of the original message was obta.i_edo Such at co.....
ditio. might Rot be noticed immediatelY9 and evea worse i the key word
iRdicating the end of the message might not act proper~o Here again,
the repetitioB of certain words and a special form and meaDS of dist~~
guishing the key word would be Decessaryo
Somewhat similar considerations to those mentio.ed above apply
to the read-out of informatioBo The words of the message, three pentads
per register, would be prepared by the computer A special order would
read the words from storage into the output register where the shiftiag
and tra.n.smission. to the drum would be performed o From the drUJI the
messages would be transmitted to their destinationso Messages on the
drum would not be oblltera,ted until an ackJlowledgeaeRt had beeJ1 received
indicating that the message had arrived safely at its destiDatioa o
0

Report R-203

-54CHAPTER V

Storage
The operation of the eomputer requires ~he storageot both the
necessary· machine orders and- the"- iftformation 8. nd-data upon which these orders are to operateo Inasnmc-h as the inrormatfon~baridling- a-spect of' ~he
air traffic control problem is-of the utmost importance, it is. necessary
to consider -the means and methods of storing the large quantity ot infornation and data to be used by the computero
This-chapter firSt considers the nature and. ebaraet~ization
of- the stored data, 'and then discusaes techniques to be used in its stqrage and utilization; espee-±ally in light ofa general need to economize 9n
storage space
Several flowdia-gratns, are il}cl1ided as a means of- illus";' ,
trating particular storage progTamso A final section of the chaptermen~
~ions several possibilities offered by the availability of an external
storage mediUID:o
0

Data Storage in the Present-Day System

The-opera.tion of the--present system of traffic control requires
the storagea.nd utilization of 'a large tt11l01l1'it of' inf'-ormatiOnarid data.o
To a large extent this consists of data which is made a.vailable to the
controllerS in the fOrinof mes-sagefh This, includesinforina tion concernirig currerit and projected flights, traffic- 'coridftirins--nearaiI'ports and
in adjacent areas 9 and information regardin€(weather" conditions. This
general type of'ihf'ormtion is of thetrarisierit'type; it describes th~
present c6nditionsorthe variables in the system, it has buts. limited
time of'application, and it must be renewed or replaced at frequent intervals o
Except for weather reports~ the bulk of this transient information relates to altitudes, sp-eeds; and times of arrival ofaircratt at
the different reporting points
As has been previously mentioned~ eaoh
reporting point-isalloted a space on the flight progress boards under
which the flight progress strips concerriingtraffic at tba t point can be',
placedo These flight 'progress strips are made out from the flight·plans,
each flight progress strip· essentiall~Carrying the full information concerning the flighto In a position of prominence on- each- strip is placed-,
the time and altitude at which the aircraft'Wil1 pass the reporting poi~~
'The flight progress strips are made out in pericil~ permitting easy alt~ra tion or correction9 These- stripsar-e usuallY s-equenc-ed in position Under each reporting point so as to give the proper time sequence ot:arrivals
tha. t' point'4
'0

-over

Other types of V'arlableda ta ..:- weather illformat'ion, , traffic
conditions-··a.t the~l'frports9-··etcOf} -~ are dfstribut~ {)ft.. paper -to· the controllers ~ orelie- 'are 'displayed on blockboards where the information is
visible to all concernedo

Report R...,203

-551ft addition to the data describing the variables in the syst ••,
there must be available for contiauous use a large amount of iBformatioa
describing the 5.1stem constants o Examples are the geographical 10catio_s
of the reporting points, the separatio. betweea reporting points, the
minimum altitudes whieh ean be flown along a certain route, types of
radio aids, ete o The cODtrollers m.ust also be familiar with the perf'oraance eharacteristics of all types of aircraft: cruising speeds, maxiaua
operating altitude 9 rates of ascent, rates of descent, etc o Although the
amount of this "collstat U informatioR is tremendous, in order to e..sure
rapid and efficient operatiOR of the systea it is all committed to memory
by the controllers Even more startling is the fact that each controller
is required to be able to control traffic in any of the several sectors
in the control area. requiring that the controllers memorize these system
and aircraft parameters for the complete area o The magnitude of the task
of memorizing the information as well as kRowing the general procedures
for the eORtrol of air traffic is.videnced by the two-year period neces....
sary for the training of a controllero
0

A General Relationship Between Storage Capacity and Operating Time

It has been the common experience of people who have considered
the solution of a number of-varied problems on a digital computer tha~
there usua.lly exist a wide selection of methods which c an be successfully
employed for any single problem o Usually these methods vary between
those which on one hand require a large amount of internal storage (orders
and data) but permit rapid times of solution, aDd those methods at the
opposite extreme which require a small amount of storage 9 yet involve a
lomg solution time o Betwe •• these two extremes there are methods which
permit compromises in storage space or operating time at the expense of
the other variable o
As an example of the variation in storage and time requir~eDt8
betweeft different methods, one might consider the problem of evaluating
a trigonometric function such as the siRe for a particular value of the
argument One approach is to store a large table of values of the sine
for various values of the argument The siRe of any particular value of
the argument can then be quickly found by inspection of the table and by
the use of a.n interpolation formula o This method is quick, but does require the storage of a table of valueso On the other extreme ODe ca.
use a series approximation to find the ~iReo Here the storage requirements are likely to be small, yet a large Dumber of computer opera.tions
are necessary to find the desired results o Another commonly occurriRg
example is that of the longer problems where the results of calculations
a.t an intermediate stage are agaiR used at a later stage of the problemo
In such a case it is a question of whether these intermediate results
should be stored for later use 9 or. whether they should be recalculated at
the later time when they a.re again needed.
0

There is~ of course 9 no direct relationship betwee. the storage
and opera.ting tim.e required for the solution of a problem. by various

Report R-203

-56methods of attacko To a first approximation, however» such a relationship might be of the form .
e ,-- soT
(1)
where!. is a measure of thecomplexity-o.f'-theproblem, l! is the storage
space needed 9 and I is the total operating time. Since! results from a
number of single computer operations 9 this could a180 be written as
o ~ s
not
(2)
where .i is the number of operations and ! Is "the average time per single
operationo Equation (2) is used-only to demonstrate the generally observed fact that with a fixed time per unit operation» there -remains some
latitude as to the choice of ~ and S which will give a satisfactory solution.
0

Of course it should be noted that there are certain limitations
which must usually be placed on!,' an~n. ,In non-real time, applica tiona
the restriction- is genera11i on"19 this,-being liinited by the storage s~e
of the computer In"real-t:ime appltc'ationsthere also exists a limit on
n9 since nt or I must be equal or less than the permissible time per solutionpermitted"by,thephysical system.~
',-

Storage-Time Considerations in .~he·Pr-esfl!t-Day System

As previously pointed otit9th~j:iresellt. method of storing flight
data involves a duplication Qf the informatton;on all of the.flight pro,grese strips corresponding to tpatfii:Lg}lto 'Sinee-th-ef'light progress
strips are 'stored together b.r repor~ing points~ this technique might be
termed "duplica te __ storage.' by repo~t!ng"":ptl"illib-so
_'
,
!

•

The cost of additional data,storage'space-is very small in the
,pr~sent-day systemal and thi~ method of ,lIduplica te storage py reporting
p01nts~ produces no lim1tat1o~s fro~ t~t ppint of vlew o
Despite its
seeming use1essness al in general agreement with the relatiol).ship of Equation (2) thisduplioation -ofstoragerdoes'·permit"a-decreas.-:i.n the amount
of work or number of operatio~s
Tll~s_ ~is ~prima.rilydue to tI,.e fact- that
instead of having to refer to a numberofrlight plans in order to determine what time relationships exist between several aircraft at a certain reporting point9 the information,i~ alTea-dy available 9 sorted outby'
reporting pointso Further, when by reference to a particular flight
progress strip one has become interested;-m'aparti-cmlar f'light9 there is
no need to refer to a master fli-ght plan to d-et-erm:ille other oharacteristics of the future or past of that flight inasmuch as all of the pertinent information is reproduced-on- that . fliight pr.o~ess stripo
Any other scheme--than that presentliusedwould·be prohibitive
from considerations of operating timeo The'lar-ge- amoimt of eross-referencing and comparisons of data which would be necessary in some other
method would too severely tax the human controllerso It is only by
0

Report R-203

-57capitalizing on 'the large redundancy-lil-stored data that satisfactory operating times are' achieved with-the present ,methods, and-human operatorso
20 Storage-Times Considerations-in a,Computer~Centrolled-Syst.m
It has been pointed- out--tba:t,lnreal':'time application both the
.§. and the! of Eq'U8.tioll (1) assUDle major iiuporU1nceo
In the present example! is limited by the rate at' 'which information arrives at the computor and the rate at which it must be processed and instructions determined for the control of aircraft o The storage limitg of course» is the
physical capacity of the computero "._
'"

At the outset,9 and before any programming has been done on a
problem g i,t is impossible to 'determine ,the true complexity of the problemo
In terms of Equation' (1) fI the 2, is not -k~pwn,' and hence it is difficulttc
say whether any chosen method of approach will produce values of .! or 1:
which lie outside of the acceptable iimitSoln'view of the discussion
above,9 if ei:ther of the limits is 'exe$Eldadit.islikelythat an adjus1;':"'ment of the other variable willpermitooth limits to be' met '-This fact
can only be confirmed in a particUiarcase,9 however 9 by the actual attempt at solution with a new approa~o
0

As a starting'point in'the"J cons ,lderat,ion o'fthe ~fr traffic control 'problemas applied to the'Whi:rilrir;t.d,.Colllputer-j-one 1.6 inc;jlined to believe tba t in view" of the limited amount of'storage'spacetbe emphasis
should be made at conservation Of stor8.g~,'at. the expense of cQmputer operations g particularly in viewd tbe·· . ·rektiH1Tlov timespeio unit operationwiththis machineo This isthe~·&~t.itOOe'· whieh'is adopted in this
ch8.pterand in those that follovo This":'8;,tti:tude·~ however 9 is not blindly
applied to all proble11l$ which arise,9 "especially those in which it is
rather evident that the duplication of a small amount of data will save
both in orders and opera tingtime 0

The foremost application ,of ,t-be.'prineiple of stre$sing storage
economy is f'Qttndinthe'lriethoo'wiU-lr:rill,;b&··e(i-opted for' th~ storage of
flight ~plan .data
Rather than'storing ;th'is-data b,--repox1;ing points as
is done at present, thereby-- requtrmg-'~alarge,amount '-of ~lieat. storage,9 t he method to be usedis-that"of'::rt~ng'"811-data eone~rning one
flight in a saparate group of re-g:l.sterseorrespondingtothat aircraft;.....
that is,theflightdata' will be !ts-tored~il1"-'ailtCI"a-fto· This means that
in comparing ,two flights, the data will not -be-pre-sortedas to reporting
points and t~e computer will have to search and hunt for the data among
the stored-flight plans o
0

It is possible to employ-'stOrage s6-hemes which involve compromises between the st'orage of flight"phndata'pure17-byalrcraft or by
reporting point9 however the storage requiTements as well'as the ease in
u~ilizing the data seem to point towards the method selectedo
Further
r~asons which consolidate this attitude shall be made evident in sections
w4ich followo

Report R=203
CI)

General Problems Associated with Storage
Characteristics of the Stored Informatioa
i

For convenieRc. in consideriBg a number of probl... which are
e.countered in the utilization of trecomputer's storage space, it is
desirable to mentioR 80me of the general cnaracteristics of the various
types of information which must be stored o There are two characteristics
which are of particular iRterest~ the first is the p~rman.Rcy with which
the data caR be assigned to a register or group of registers~ the secoRd
is the amount or length of the storage space which must be allocated to
variour! a.ssocia.ted quutitieso

A large amount of the general data which the computer must
operate upon can be assigaed to permanent positions in the storage
This
is true of quantiti~s whi~h must always be available for use by the COJll=
puter~ regardless of whether the values of these qu1fttities are changfd
lOr not to ~orrespond to existing extern~.l coRditiolU5 o The importaJl~
point is that this information will be contiaiU1.ously needed for program
opera,tioR,9 and the permaRent allocation will aid theutilization of this
inf'orma.tion by the progra.mz o
0

A large part of the stored data fits into the "permanentlY''''
Examples are the geogra.phical iRformatioR re . . .
garding areas$> airportp and airwa.ys~ conversion tables; aircraft per...,
formance characteristics~ and weather information at various poiats and
altitudes in the control area o This last example is one in which the
stored quantities must be varied to lJleet existingco:aditiollso
stored lll classificatiol'lo

The npermanently-stored U characteristic does »lot apply to the
flight plan data o Here the data is needed for a limited time only, this
time being equal to the period that the flight is umder control of the
computer
If all aircraft flew accordiDg to schedule 9 it might be advis3.ble to permanently a.llocate cia set of registers to each flight; howeverp such a scheduled flow of traffic is not achieved at the preseAt
time» nor even approached under C),ny conditions o Even under highly optimized conditioRS involving only prescheduled traffic» the efficient use
of the storage space would require a large amount of double or triple
tenancy of this space under some form of time-snaring o In order to
achieve an economy of storage under the BOR=preschsduled conditioRS it
will be necessary to consider oRly a temporary alloeatio. of a set of
registers to an individual flighto As sha~l be described in the next
section.\> this allocation would be made from a common pool of registers
re'served for that purposeo
0

For the most part there will be very little need of storing
a single» una,;ssociated piece of data o Most quantities can be associated
with larger classif'icatioRs such as flight plan datajl aircraft performance characteristics» airways data» reporting point dataj) etc o Under
the major division by classification, there is a further logical divisioa

Report R-203

-59into the individual flights9indlvidual tyPes of aircraft~ individual
airways, etco These may be referred ·to as the m~mber8 of the various
classifications 0 Each of thesemembersot a classification will generalJT
require a small block of registers for the- 'storage of the associated cIa tao
As mentionedjl for each member of·the aircraft performance characteristics
.classification g the cruising speed 9 rate 'c;>f a~cent9 rate of descent~ etc.,
must be storedo

In considering the member blocks of a single classification9
there are two general cases; either these blocks are of a uniform length,
or the block length'variestrom'member to-meinbero, Constant-1ength,blocks
are the usual oecurre~ce, a common,'example .. b~ing··:the 'aircraft performance
characteristics whe:reinthe·'·s.me-numb·er-..-ofc:quantft1es must be storedf'or
each aircraft An 'eXample··o'frion-cODs:tant,..length "blocks is found in the
storage of the flight plan da;ta o In,'this.casee8-ch -menlber -block corres":
ponds to an individ1l!iI' flight 9 'and>since'thie :;Length~ of a flight is fixed
by the number of reporting points over'wh!chltwill pas8 9 the blocks ot
flight data·' will generally vary -considers. bly from flight to flight.
0

Finding Stored Data
As a

general example 9 consider that, a t some stage' in a program

the comput'er 'desiresto deal' with a'eert8J.:n,~mbero·f -a classification.
Assume that'each member in this eiassifieatio·nhas associat-ed with ita

constant..-length block of~·.Dreglstersandtba.:t··~ -addrtgSfJs of the consecutive storage registers in theolock used~.for-the ·1 member are given
as &19 19 a129ooo00oaign. In particul~r the address of the first regist- er of the 'fIrst"block 'will be' al 9 10
-.
The need for 'ref'erringto"a partietiar'·"member of this classification may-arise because thismember---'_8re~·to··orspecif'ied in an
inputmessage; for example',s. flight plan:=m:i:ght'~ specifY that the aircraft under consideration is a DC-3o The three-character binary-pentad
designation of any such member will be designated as~. (Lines above
letters will be used throughout__this chapter when referring to 8 binar.ycoded pentadre~resentation)o Mf. would be the approp;riate pentad designation ot the i h membero
A first problem wich arises is that of finding the address
~191 of the first register of the b1oekcorresponding to it One method
of attack would be to arrange that address!.,.S &1 1 be deri"!!lble by simple

arithmetic operations upon the quantities Hio~ince the M,tu s wi11 be
somewhat arbitrary9 it becomes convenient toeonsider a pure binary number Mr associated with M.to If the Mlrs are appropriately chosen as a sequence of consecutive numbers, 'then in particular it could be set up so
that
. 81 9 1 a: al,l + n
M.t. - no
In this way "the members occupy consecutive blocks OrB registers ~aeho
For example,9 if MJ. =: 19 ~- 29 ~~39 etco g al~l == 100 9 and D= 10,9 then
the first addresses of the member blocks are gi'ven ae8
0

Report R=203
Member
1
2

First Address in Block
100

110
120

The scheme is easily extended to other consecutive values of M, aad ca.
be modified to utilize other aritkaetic relationshipso
As noted~ this scheme requires that the Mis be assigned in a.
orderly fashioR o This functio. is easily performed by an input conversion table
A ne~,t method of sa,ving space is to let the M' s b. arithmeti£Ally related to the addresses of ~he consecutive registers ia which
the MU 5 are stored o Thu5~ given any Mi a search would be made through.
~he table to find the ~ddre5s of the register containing the quamtity
Mia The address is then operated upon arithmetically to give ~ and heRc.
the address ai~lo This scheme shall be termed the coded-address method.
0

In some cases it is not convenient to have a conversioa table
of the type described above o A particular example illustrated later iR
the chapter com"cerlts the storage of data regarding the various airways"
In such a case a dirferer.tt a.pproach can be applied to the problem. of
findil'lg the positioR of a block of storage registers correspondiBg to
a particular membero In this method the member desigl'lator Di (this may
either be a 3-character binary-coded word or aBY arbitrary binary number)
is stored in the register whose address is ai 1 0 Hence to find the
appropriate block of registers used for membe~ Di~ the computer would iR=
vestigate the registers &1 l~ -2 r» a3~19 etc until the quantity Di were
fou.d o (1l'l this connectio~ it·i& convenient to have constant block
lengths~ where 3,2 1 = al~l+- -,9 a391~"}1+~'e1co This method shall be
termed the searcb. 9me thod o
9,
o

A comparison of the above two methods will indioatea certaia
fundamental similarity, in each there is an extra allotmeRt of one register for each member o In the coded-address method this manifests itself as an input conversion table; in the search method this takes the
form of an extra register per member block~ There is a difference ia
application of the two methods o The search technique is useful whem BO
conversion table is otherwise Reeded~ and is better applied to consta.tlength consecutive blocks o The coded~address technique is useful when
combined with an input conversion table~ and is also better desigmed for
use wheli the pure binary desigPl8.tion of the member M must be stored ill
other registers 9 as is the case in handling flight plan data where the
designations of reporting ~oiRts must be stored with each flight plano
In this case~ whereas the MIs will each be a full register in length9
the M's can be shorter binary numbers and as is mentioned im Sectio. C4

Report R-203

-61. .
of' this chapter, they can be stored in a register together with aRother
quantity A further advantage of the coded-address method is pointed out
below ift connection with the handling of non-permaaemtly-stored data o .
0

In using a block of data it is generally necessary to extract
aad use subdivisions of that block o As mentibRed, under the classifioatiOR of aircraft performaBce characteristics there will be stored the
cruising speed" the ra.te of descent, the rate of aseeat aJld the aaxiaum
operational altitude for each member aircraft o In such cases where a
similar storage patterlt is repeated for each member block a.n.d the block.
are consecutive and of the same le.gth, the computer can be programmed to
look for, say, t:p.e cruising speeds in. registers Qi p9 i ~ 1, 2
0 0 •. 0

In the above ease ~ the computer is progra.mnted to know where to
find the various quantities o Such programming is not possible when each
block does not colttai:rt a simila,r storage" patter.
This is the situatioa
eaeouRtered with the storage of flight pla. data, where each flight will
have a different Rumber of reporting poiRta o In order to properly locat.
the subdivi'sioBS of' the data. the number of reporting points in each fligb.t,
plan ca. be counted and stored as aR aid~ or else key words can be stored
to iadicate the eRd of a variable-length sectioa& The latter technique
is probably best suited to the case of the handling of flight data, iaasmuch as the flight progresses some of the stored data caD be eliminatedo
0

30 Handlillg of Variable-Leltgth Da.ta
In considering the problem of handling data which has a permameRt loeation but which is OR nOR-constant length, a block size equal to
the maximum expected length can be allocated, and the coded-address or
search schemes are applicableo Extra registers resulting from any m~ber
using shorter blocks than the maximum can be used for the storage of general constants Reeded by the computer programo

When dealing with flight pIa. data which is ROD-permanently~
stored a.nd is of variable leJllgthg serious questions arise as to tne .eus
of obtaining maximum utllizatioR of storage As an ex&mple of the problem consider that a. group of registers begiMiDg at address!! have be••
alloted to flight data as fol1ows~
0

Flight
I

II
III

Addresses
m to m+- 19
20 to m. + 29
JI +- 30 to m + 59

.. +

NUMber of Registers
20
10
30

A typical situation whiCh might arise is that i . which fliih~
II termiaates SAd its storage space is made free, and short~ afterwards
flights IV and V9 requiring 25 and 15 registers respectively, wish to
joi. the s.ystemo The questioRs which such a situation creates are twofold:
1) How do we determime where free registers are available?

R...203

-622)

How do we best decide which flights to assiga to
free space which is available?

The first question caB be resolved by either storing key words
in each register when it is made free -- such a key word possible beiag
a aegative number -- or by separately storing the pairs of addresses iBdicating the first and last registers of available groupso Ia the above
example the space made available by flight II cannot be used for flight
IV because of its sh.ort length, while if it were used for flight V five
registers would be wasted, especially since this leftover space would
be too short for amy other flighto The problem becomes even more difficult when it is realized that as time passes and the flights progress,
each flight needs less and less space o Unless this space is made avail...
able for use» the efficiency of the use of the stor~g. space of the computer becomes rather low o
One scheme which permits a satisfactory solutio. and which per...
mits ecoRollies in. storage would be to shift all data upwards filling aay
free register which might occuri thus always' moving the available storage
space to the endo Such a scheme is likely to involve considerable time
and somewhat complicated programming o It is felt that an alternative
sehem. presented below is simpler and fastero
The suggested method is oa. which us.. blocks of a fixed l.agtko
This block length would be eORsiderab~ shorter thaa the average auaber
ot registers lleeded for storil'lg the complete data for a. sillgle flight,
and would probably be of a leRgth sufficient to store the flight data
corresponding to ~bout two reporting poi.ta o A sufficient number of
these blocks would be assigmed to each flight so that' all of the data
could be aecomodated o The address of the first reiister of each of the
blocks would be related to another address i. an assignment table, in a
manner similar to the coded-address sehe.e The registers of the assigament talle would initially be negative 9 iadicating free blocks o The storage of the three=character bimar.y~eoded designation of a flight in a~
register or registers of the assigmaent ta.ble would ~dieat, bhec'illoc&tioD of a block or blocks to that fiighto The reporting points in the
flight plan might actually be stored backwa.rds (although in the i.put
message they would be in the correct ,order.) , so that a.s the flight progressed g blocks used to store the" r~por:tirig points could be made free
from the eDd o Fuller details of tne sc~eme are given in BeetioD Eo!
this c h . a p t e r . ·
.'
o

40 Half=Leagth Storage
As Boted ill the previous chapter 9 with suitable scale-factors
and by utilizimgthe fact that oBly certain discrete values of variables
are of interest, it is possible to represent mOlt of tb.. data. stored iB
the computer by oBly seveR or eight binavy, digitso It caa also be Roted
that most of major elassir~catio.s of storage have less thaa 128 me.bers
and hence seven digits suffiee for the member designatio.s after tke

Report R-203

eo.version of the i.puto This beiBg the ease 9 it appears that advantage
e.a be taken of the economies afforded by storiDg two pieees of iatormatis. i . a single registero 10 eomplicatio. arises due to the algebraic
sig•• il'laSntueh as most of the quantities to be dealt .. with are positive"
The present Whirlwind orders are Rot very well suited for thi.
ha.lf-length storage -..,. either in the assembling or separatio:n. of at co.pound word" Although a special order would be useful, there would Bot be
a co.siderable saving in storage iRasauch as the assembly, separatioR 9 or
change of a eompORent part of the compou.d word could be do•• by a few
special subprogrmas which would be usable by all parts of the u.iR programo
Ttie special order would save considerably i . operatiag time, however o

50 Applications of Programs to Various Blocks of Stora.ge
Althoagh each of the com.pollent parts of the overall computer
programs will llecessarily apply at som.e time or an.other to various __bers of a classiricatio.~ the progr~s aswritt.. and stored im the computer aust have defimit. addresses supplied with the orders o A probl..
whieh arises~ theA, is the adaptability of the progra=s to the various
m.mber blocks of storage o
Of course9 OBce it has beeR determiaed what the member is amd
where it is stored 9 tne address seetio.s of the prograa in question ea.
appropriately modified o This tecnmique is rather le.gt~ and wasteful
of storage, especially when it is cORsidered that almost all of the programs must be Modified
A more acceptable technique is that in which
all such prograas would be written with the critical address•• referring
to a fixed set of otherwise unused registers; th•• for proper program
operatioa the pertiaeDt member's block of storage would be bodily transferred to the fixed set of registers o This traRsfer might be best iaple.eRted by OR. or more special computer orders~ particplarly of the 881£=
il'ldexing type, however~ as in the case of half...length storage the end re . . .
sult eould a.lso be accomplished by a. small subprogram which would be i .
eommOR use o This subprogram would either have to be supplied with the
number of the registers to be transferred~ or else a key word could be
employed to stop the transfer im a Dla11.lter similar to that ill which a key
word designa.ted the ead of an input message o

Point amd Path Dati
IJd'ormatioa to be Stored.
For purposes of computation and control the computer must have

av•.ilaole certain operational and geographical i.formatioa regardb.g bota
the various reporting poiate a.nd the airway or paths joining the.e poiBtso
This informa.tioa is most easily cOBsidered as beiag separated i.to two

distinct elassifieationsg

poiRt data aad path data" As noted earlier
rligh~s shall be separate~
stored by flight pla.~ &ad heDee the poimt and path data represent permaReat allocatioas of storage spaee o
i . this chapter, all informatioa regarding

Report R-203

-64=
The following listing in.dica.tea the types of inform.a.tio. which must b.
stored for each of the reporting points in the co.trol area~

iRdieatiOR of whether the poimt in question i. ia=
teraal to the area or whether it is near the bouadaryo

infor:ru.tioll rega.rding the approach control faciliti'.so
This will i~clude the positioD of the stacks~ the
highest stack altitude in use, the time iDtervals for
successive approaches» .te o

IDforaation regarding altitude restrictions or time
limitations OR traffic over the poimto Aa imdicatioa
must be stored as to whether the poiat is suitable for
holdil1lg a.ire:r~.ft en-route ~ this being determined by the
position and geReral traffic density at this pointo

intormatioR regarding visibility and ceiling co.dit~
ions if there is a.n airport at that reportiRg poiRto

The following listing is typical of the ildorm.atioJl which. must
be stored for each of the paths joining the reporting poimtag

the length of the patho

the minimum flight altitude which is permissible aloa, this
patho

we&ther conditions, to include visibility conditio•• ~ cloud
tops, and wind magnitudes and directions at various altitudes"

any radio beacons or markers (non~reporting poiats) whick
can be used by the computer a.s B.n aid in the traffic e01l....,
trol o

The various paths and poiRts in a cORtrol area are not total~
unrelated quan.tities illaslI.uch as the area. possesses a certaiR geometry
as defiaed by the geographical location of the poi.ts and their co...ctillg paths. To be able to properly direct tr~ffic in an area, the computer must be able to reconstruct and utilize this geometry. It ia mece$=
sa.ry to determia. which paths m.eet at a point al\d which points lie a.t the
ends of a particular path. In ths preseRt-day system the co.trollers have
maps available for rerereRce~ although they are usually able to rem••ber
and mentally reco~struct the .ecessary geometryo A computer, lacki_, aa
i-nereRt sense of spatial relationship} must perform the task by aea••
of stored informatioR o

Report R....203

The storage and utilizatioa of the path and point data will r.~·
quire the assigRMeRt of designatio•• to the individual •••bers of the.e
two ela,ssification.s o Consider that these desigaatioll. are P a.nd Q, for
poi.t aDd path respectivelyo 0 •• meaRS of retaining the ~.cessary ,eoaetriea,l i.rormatioJll would be to store with each m.ember P of the PO~Jlt
data the des;LgnatioJls of a,ll the paths Q which meet at that pout,'" a:nd
siJlilarly to store with each member Q of the path data. the two piS at
the eRd points of that pa.tho Such a method is quite w&lsteful of storage
inasmuch as a. good deal of the stored informatioJl is redlmduto
A. alterRate scheme might employ the assignme.t of particular
numerical values to prs and Qts in such a way that the relatioAships b.~
tween adjacent points and paths would be d~~ermiaable from these numerical v8.1ues o This schem.e was rather tttoroughly investigated by the author"
and it wa.s determiaed that such a '·Jlum.b~riJ1g system. was rather difficult
to implement due to difficulties encountered whe. three of four pathsl
joined a.t a pa.rticular poiRto Although a. fea.sible mumbering scheme might
•• sily be ~thesized for a particular area, the method turned out to be
extrem.ely difficult and lengthy of stora,ge spa,ce when applied to an area
with a general geoaetlr,y;;·"
As a point of introductl0. to the scheme which it is thought
presents the best solution to the problem at hand, it is recalled that
all reporting points are to be designated iR messages by three-letter
a.bbrevia,tiolt5 as in. the present practice
These Ul'tcoJ!lverted desipatio••
shall be denoted by Po IR a normal size control area today there are b.~
tweell 25 and 75 reporting poiJl'lt·. o It i8 convenient to consider a maxilaua
of l28~ inlu3l1.uch as pure bi:m.ary designations requiriJlg only seveR binary
digits Gj,re then required o Thes. pure binary member desipatiora.s are
deRoted by P g and could be assigned by means of a coded-address cORver.ioa
0

table

Each path in the area is cOMplete~ defined by the poiatsat
its two ends o If these two points are Pj and Pk9 then the compound 14digit word produced byjqixL:mg Pj 2.Rd Pk ealt serve a,s a designation for to
path This compou-d path designatio. ca. bi,de.oted as Qjki and if Pj aad
Pk are thought of as numbers between 0 x 2and 128 x ~- 5~ Qk P j
x 2'fJ + P~o The word Q.1k would be com.posed such that the first seven aigits coe froll. the P wi£h the smallest mapitude o The storage of the path
data. can the. be doftei:a blocks aceorU~

Report R....203
of progress of the flighto Storing the flight data by reportillg poiats
alone would require the additioRal storage at each reporti.g point of the
IUlll.eS or desigllatiolt of the preceeding and succeeding points of the
flight 0
Fo

Use of Additional External Storage Space

As a contrast to the previous aectioR. of this chapter which
have coacentrated upo. the necessity of eeoBomiziag storage space, it is
iateresting to i.traduce several comments cOReer.iag the possible use of
a. exter.al storage medium with the computero The remarks whick are .ade
below are purely of an illustrative nature, however~ aRd the reaaiader of
this thesis is based upoa the teehaiques and methods already •••tio••d o
The use of am exterJuiI,l storage medium. implies tb.at iJlforaatioa
to be used must be traasferred to .Bd from the external .ediua, this
trti.nsf'er being either of a single word oro.f a. block of registers Both
the reading to and fro. the external medium .ad tne 8ubsequ••t use of the
informatioa requires a certaiR. aJl.OUJlt of time over ad above the tl.ae required if it were iaitially stored iRteraallyo In spite of' tRia Co&,
sideratiOB there are two factors which make the use of such devic.. quite
feasible
In the first place, although a large amouat of data must be
stored~ o~ly a small amou.t of' data -- usually of block size -- is actu;,11y :aeeded at a particular time; in additioJl, the luted for a particular
block or section of the data will generally be known somewhat in adva.ce
of' the time that it cam actually be used a Similar cORsiderations apply
to various programs and subprograms used by the com.puter; ol'lly a. small
.umber of' th~m are Reeded for any particular task being do•• by the computer, amd the meed for these programs is usually kROWR iR advaBceo This
rorewAr~i.g of the use of data and progr~s will become more evideBt i.
suceeediftg chapters when the large amount of bookkeepiag or maaipulative ope!,:atiolts necessary for the execution of th.e various computer programs or subprogr~s is discu$sed o In view of this fact, it would be
possible to arra,Rge the programs and the various sequences of calculatioa.
and operations in such a way that inhereat initial delays in the use of
aa exter.al storage device would offer but a small drawbacko If oae had
a magnetic tape unit which could be started and stopped rapidly, theft i .
the period betwee. the time wheJl it become 3 kliOWB. tba t sOJl.ething is
wa.ted and the time when it is .eeded, the tape eould be moved forward to
the desired iRforaatio. in preparatioa for a read-iR o A second factor
to be reeogmized is that if the exterl'tal storage device is used predoa.."
i.ately for permameatly stored data or prograas, there is RO ••ed for ~
time used ia_reading iRformatiol't back to the storage device a This ail.t
make the use of paper tape aBd a photoelectric reader feasibls o
0

The availability of the exter.al storage would i.troduce large
enaltges into the provisions a..lready ellumerated for the alloeatioJi aJld us.
of the i.ter.al storage space of the eomputero A change to the storiac
of flight informatioD solely by reporting points, as im the present-day

system, Bight be made or some eompromise between the two radically

Report R-203

-75-

differe.t storage schemes adopted; storage achSMes whick ecoRoBiz. o.
tille rather tho storage eould be adopted. TAe lIlec.s.1.ty for coavarsio. might be elillillated if the coaputer were progra_ed to eo_put. i .
teras of the biaar.y-eoded three-character groups. Whereas this chapter
has co.sidered that times &ad altitudes be explicitly stored as aum.rical
quaatities, if a large amouat of additioaal storage space were available
it would be feasible to allocate particular registers to particular tia••
ad particular altitudes. The oeeupaJl.cy of that tiBle or altitude would
be indica.ted by stari., the appropriate flight c!esigJlatioR. ia the corres!".
poadiag register. SUCh a sche~e would save greatly OR the "OURt of tiae
wAich would be required to detect situatic.. ia whick there is i ••utiiei.
separatioas be~wee. aircraft o As shall be JUl\tioaed in. Chapter X, the
ava.ilability of additioaal storage space would aake it possible ill a aodified syste. to haadle amd give priorities to scheduled aircraft.

Report R-203

-76CHAPTER VI

Areas and Airports

Before a study can be made of the computer program, it is necessary
to discuss in fuller detail some of the problems whioh arise in the controlo! airoraft flying between areas or approaching and leaving airports.
This chapter discusses these subjects from two points of view: the necessity of a more formal standardization of control procedures for efficient oomputer mechanizatiQn, and the means of coordinating between the
oomputers oontrolling adjaoent areas or between a computer and the approach controllers in the airport towers.

1. Elimination of Sector Control
In Chapter II mention was made of the general division of authority
first into areas and then
into sectors. As noted, tne division into areas, eachsupe:r:'Vise4 by a
single control center, was made.for reasons of economy, while ·the divi·sion into sectors was neoessitated in view of the limited speed and datahandling capacities of human controllers. In both cases, however, the
division of authority requires that careful attention be given to the
means of ooordinating the individual efforts and actions of the sector
and areas.
in the oontrol of nationwide en-route traffic,

The speed and oapacity of a computer is such that one oan reasonably expect it not only to replace a single controller, but all the controllers in an area For reasons 'somewhat similar to those which justify
the present organization of the control system into areas, it does not
seem either necessary or advantaseons to postulate the use ofa single
computer for nationwide control, even if' a super-computer 'such as would
be needed were available. In fact, a division of control into a number of
area computers would offer oertain advantages as regards mairrt~nance and
overall system reliabilitYi>
0

The use of a computer as the controlling element in a single are~
obviates the need for seotor control and sector ooordination; nevertheless, provisions must be made for the successful coordination of the
individual areas into a single unified system of controlo
20

Area Coordination in the Present-Day System.

Under present methods, centers in.adjacent areas coordinate and
share the oontrol over aircraft flying along airways or over reporting
points which lie near or along the boundaries between these areas~

Report R-203

Both control centers store the flight data regarding those paths and
points 9 and the two centers agree upon a common clearance for the a~
craft involvedo In these cases one of the centers normally has explicit
and final control over an aircraft until it reaches a certain point in
lts flighto The paint where the transfer of control is made may e1thel;"
be a reporting point or may be a point defined as being a eertain:amount
of flying time away from a particular fix or markero The point of transfer of control is fixed by agreement between the two adjacent areas, but
the agreement is usually of such flexibility that variations are per-··
mitted to handle dense traffic conditions, especially in the case of aircraft landing near the boundaryo To quote from the ell regulations:
If weather and/or traffic conditions require ,"'th$cen-

tel" controlling the point of intended landing may request an adjacent area to clear iircraft to a specific
point during a specific periodo
I
The sharing of f,light information is made possible by the
prOVision that approprIate flight plan data and control information
pertinent to an inter-area flight be forwarded from area to area as the
flight progresse6~ This data normally includes&
a)
b)

Flight identification and type of aircraft.
Estimate (time) and altitude over the last fix within the control area and the altitude of entry into
the adjacent center's area i f different from the
altitude over the last f i x o ,
Actual ground speed 9 if determined; or esttmated
ground speed
Point of departure, the remaining portion of the
route of the flight, specified in the original or
amended clearance, and the point of first intended
landing
The estimated time of arrival as specified in the
flight plan (time of departure plus elapsed time)
based on the time zone of the departure pointo
Clearance informationg
1) Clearance limit, if other than the airport of
destinationo
2) Special information, if issued 2
0

The above information~ with the exception of item 0) which
need not be included in the case of scheduled air carrier or military
aircraft 9 is transmitted by the center which has initial control of the
aircraft so as to permit reception of the data by the adjaceni center at
least 30 minutes prior to the time that the flight is estimated to enter
that areao If the point of departure is not at a sufficient distance to
permit the transmission within the specified time~ the coordination is
made prior to the takeoff of the aircrafto
10
20

Reference J.6 ~ page '20
Reference 16 ~ page 15

Report R-203

-78Because of the fact that an overall picture of the traffic cannot be
projected very far into the future, it is impossible for the coordination
between areas to effect very much more than a check between centers regarding traffic at or near the boundaries. The acceptance or denial of a
clearance by an adja~ent center is usually based on immediate conflicts
affecting aircraft arriving at or departing from boundary airports, and
in very few cases is the acceptance or denial of a clearance based on expected traffic conditions at points in the interior of areaso
The only other for.m of coordination between adjacent areas is the
nature of' restrict1ona~ restrictions as to usable altitudes over certain
points or to the number of a1rc:ratt which may be dispatched to an airport
in another area over a period of time& Such restrictions are usually
caused by abno~lly bad landing conditions which have resulted in congestion, high stack altitudes, and long delays in landing& The two e~amples
which are familiar in'the Boston Control Center are time restrictions on
flights into the metropolitan New York area, and altitude restrictions
over Hartford, Connecticut, a crOSSing point for North-South and East~Weat
traffic in New Englando
Although the controllers check all proposed flights for confOrmity
to existing time and altitude restrictions, for the most part they find
little need to enforce such restrictionso This situation arises because
the operations offices (military and commercial) are infor.med of the restrictions and hence plan their flights accordingly Nevertheless, in
order to guide itinerant traffic and check upon the military and commercial operations, the pertinent information must be stored in the control
center
0

Outside of the time and altitude restri~t1ons, whiCh are sparingly
used, there are no attempts made at the present time to control or regulate the overall flow of traffic between areasG As will be mentioned in
Chapter X, such flow control offers a possible chance for improvement in
the present system
0

Criteria for Area Coordination in the Computer-Controlled System

OoQrdination not only involves a sharing of information between adjacent centers, but require~ & procedure for reaChing an agreement between
adjacent centers as to the control of flights near the boundary For'human operators communicating by means of interphone, the compromise and
agreement upon a satis~actory c1earanee and the decision as to who shall
control the flight until a particular point can easily be madeo
0

The computer, on the other, hand, cannot be easily programmed to compromise or bargain satisfactorily, and it is Itluch more desirable to have
a. rather standardized procedure for the coordination
The procedure. which
6lhal,l be adopted for tbis study i~ that of having one of the two involved
{area} computers propose a clearance, and then assume the respons1b1l~
ity for making changes or corrections upon the suggestion and advice
of the second computer
That is to say, the coordination :i.S to be
effected by a procedure of proposal and confirmation
For such a
0

Report R-203

-79-

procedure to be lnechanized, a distinct method is needed for decidingvhich
of the two computers will assUlIe the task of making the proposal and car~
ing out the cont~olo
It does not appear very desirable to have the question ot which
area is to control a flight near the boundary of two areas be eompletel1
dependent~ as it leat the present-time, upon the actual area in which
the aircraft happens to be~ particularly in the case of aircraft ascending from or descending to airports at or near the boundary. As will be
noted in Chapter I1~ it is theQe sections of a flight which offer the
most difficult problems in control$ and it appears to be much wiser and
more practical to have the supervision ora flight be dependent upon both
the position and present plans of the aircraft, that is, the complete
supervision of departing and ascending or arriving and descending aircraft will be allocated to a s~ngle computer, regardless or the actual
position of the aircraft with ~espeet to the boundaries o This procedure
makes it mu~h simpler to check and olear for separations'on ascents"'and
descentso
In order to appreciate what this plan of unique control for ascents
and des~en~" entails 9 it is interesting to note the performance figures
of typi~al aircrafto The rate~ of ascent and descent vary according to
the type of aircrafto General:t,.y, the faster aircraft fly higher, and
because of cabin supercharging are able to desclnd or ascend more rapidly.
Typi~al speed~ for faster commercial aircraft are in exoess of 250 mph,
with altitudes of between 10 9 0QO and 30,000 feet and rates of climb or
descent between 750 and 1500 f~et per minuteo The older, slower and
smaller aircraft generally do ~ot have cabin pressurization and hence are
limited to a maximum altitude ~f lO~OOO feet, with rates of climb and
defscent up to 500 feet per min~te, npresen'tativ.:e air speeds are between
150 and 250 mpho

From the above figures it can be seen that the rate of climb or
expressed in terms of lertical distance R§r horizontal distance
travelled is essentially const~nt, regardless of aircraft typeo The
figures also indicate that the maximum distance required for a full oncourse1 as~ent to a oruising a~tltude or a complete on-course descent
from cruising altitude is about twice the average distance between reporting pointso For purposes of s~andardizing the coordination procedures it
shall then be assumed that any "full ascent or descent will at most cover
two reporting pointso As::is noted in the next ~ectlon9 in the speoial
case of short flights from one area to the next it will be assumed that
a single point asoent or desce~t is sUfficient
des~ent

40 Area Coordination for Computer Control
A representative plan

of two adjacent areas and an interAs shown~ it is assumed that
be placed so as to intersect the airways about midway
pointso A, ll, E9 and Fare internal reporti1lg points,
vi~w

Cilonnecting airway is shown in Figure 150

the boundary w~ll
between report~ng

This is to

be~1stinguisheq

from an ascent made while circlingo

I
B

ARE.t I

AREA II

BO
o.y
LI

Figure

Report R-203

-81-

e and

Dare boundary points
Each area computer will store the data for
and have unique control of flights over or b;etween its internal bounda.ry
points; both computers will store the relevant data concerning the boundary points C and D, and control of flights over and between those points
shall be according to situation
enumerated belowo (The individual
situations are describ~d for traffic flowing from left, to right in Figure
15; by an obvious interchange of letters 'they apply to traffic in the
other direction)
0

Flights originating at or to the left of A or B and teminating at C are to be under the control of Area I. In this
case, the_computer of Area II need not store flight plan
data corresponding to point Co

Flights originating a.t or to the left of A and B and terminating at D are to be controlled by Area I and will receive
clearances coordinated with Area 110 Area II should receive and store the flight plan data for pOints C and D of
these flights at least 30 minutes before ,the aircraft arrives over Co

Flights originating to the left of A or B and terminating
at E, F, or beyond are to be initially controlled by Area
I, control reverting to Area II over point Co In such
cases Area II must receiv~ a 30-minute notification of the
arrival t~e at 0, and it necessary that area may request
the aircraft to be at a specified altitude at that point
0

Flights originating at A or B and landing at E, F, or be yond should be initially cleared and controlled by Area I
upon proper cOQrdination with Area 110 Assumption of control by Area II should be made when the aircraft passes
over pqlnt Do

Flights originating at C and entering Area II should. re~
ceive their . clearances from Area II upon proper coordination with Area 10 Area I should receive a,30-minute notif'i,cation and should store the da.ta for points C and Do

Flights originating at D and proceeding to the right are
solely under the control of Area 110

It will be assumed for convenience that no flights proceed
from A to B via C, or that C is crossed in any other than
an inter-area flight
The boundaries between areas will be
chosen with this fact in mindo
0

The situation of 4) above does not permit a complete coverage by a
single computer of a two-point ascent or descent when an aircraft originates at A or B and terminates at E or Fo For these relatively short
flights; however, an aircraft would not reach or fly at a high cru±si~g

Report R-203
altitude and satisfactory control should be possible by a single-point
descent
0

50 'Transmission of Flight Plans from Area to Area
i

Under the pres$nt ~fr traffic control regulations, pilots tile but
f~ight plan, regaWdless of the length of the flight or whether
it enters one or more other- aresso This fl.ight plan specifies theeomplete route of the flight,and it is passed from center to center at
least 30 minutes in advance of the expected time of entry of the air,craft into the next area
The only ~xception to this rule is in the oase
of aircraft departing from.. airports
or near the boundaries of areas;
in this case the flight plan tafiled with both traffic control centers.

a Single

For lang flights, the specification of the route ~ompJ.etely in
terms of reporting pOints would'be gutte "'lengthy This does not c:re~te
any difficulties fnthe present-day system due to the relative cheapness of storage space; ill addition, tIle problems are lessened by the
relatively small :number of .~ong: flights and because the route of thEp
flight, especially that part outside of the originating area, 1s lls-qally
specified by airways desigpationo
0

•

There are two approaches ,:which may be taken towards this problem'
in usi.ng the computer
The first wQU.ld entail, the storing of all the
0

necessary flight plan data-- and the forwarding of same
Such' a scheme
would be rather expen$ive in storage space and appears to be'less favorable than a procedure under which only the relevant part of a lopg flight
plan would have to be filed a:nd stored by any single computer The.$cheme
would mean that each computer. mus.t notify the next area of the need 1;o~ecure
.~ part of 'the flight plan frQ'll1 an operations office
The mechanization.
'of the plan is'as follows~
0

The Qriginal filing of an extended, inter-area flight
plan 'Would.. include only the points of the original $'rea,
except in the s itu.at ion where the flight covers ()~" a
few points. in that area
In the latter ease the ti~1ng
'Would con:tain all the points of the first two arees.. If'.
olUy the 'P01llts for the original area are tiled, th~l'l
the computer in that area has the responsibility for
notifying the computer in the succeeding area 30 mi&~tes
in advance_ of the arrival of that flight to secure its
portion of the fllght plano If the original filingco:mtamed both areas completely, then the flight plan
should be forwarded by the first area to the second ~ithin
an 'appropr,iate time <> (The previously described pro~- .
edures for area ooord:ination permit a 30-minllte adv4nee
'Warning to be made) <>
0

The same procedure is to be extended for longer flights, eae)! area
either receiving its part of the flight plan from a 'previous area or.as a
result of a request to an opelfat1ons officeo
the latter ease, the
requesting eomputerwould receive either the reporting po1ntsfor its
area alone, or if the flight is of short duration in that area, the

Report R=203
reporting points for it and the succeeding areao
The differences in the methods of forwarding the flight plan data
are necessitated by timing dltfleuIties2) i~ particular the necessity of
having the computer in an adjacent area ..:reoeive the flightplan'before
it receives a notification of the arrival of the flight from ~;,'.:: ';
preceeding areao
B0

The Problem at Airports

Present-Day Coordination at the Airports

As has been mentioned in Chapter II" in order to handle the increased
traffic at airports in recent years, especially during poor weather, it
has been becessary to supplement the normal airport tower control with
special approach control procedures and facilitieso Inasmuch a.s all air~
ports are not supplied with these l1acilities at the presentt~:, ':it is
necessary for the centers to be able to properly coordinate and handle
the arrival and departure of aircraft from airports which mayor may
not have approach control service o
At the airports where only the normal tower control is exercised, the
bulk of the task of handling arrivals and departures is carried out by

the center controllers 0

The center has control over all'departures as
well as over separations between successive departures and between' "
departures and arrivalso This responsibility requires that in congested
conditio:ns the center must specify characteristics of ths :'takeoffs - ...
direction of takeoff s> turn aft.er leaving ground~ climbi:gg instr~ctions,
etco As regards arrivals~ the.-center ha,'s'e.omplete control over the airoraft tmtil they have been specificall,

When centers clear an aircraft to a point prior to the assumption
of responsibility by approa~h control~.the aircraft is given specific
holding instru~tions~ including altitude and the expected time.atwhich
the approach and landing will be madee This function£) as well as the
general coordination of the two traffic control services~ requires the
following exchange of information:

Report R=20.3

-85=
Approa~h

Control to Centers
Highest altitude in use by Approach Control at the
hold;tng pointo

b)
~)

e)
f)

g)
Center~

a)
b)

d)
e)
f)

Aver$ge time interval between successive approaches
as determined by the towero
Revision of the expected approach time issued by
the ~enter when the tower calculation indicates a
variation of 10 minutes or moreo
Arrival times over holding point or statement that
aircraft is under tower control, if released prior
to arrival over holding point.
Departure times of departing aircraft.
Available information relating to overdue or un~eported airorafto
.
Mi~sed approa~hesol

to Apprgach control
Identirication~ type, and point of departure of
arriving aircrafto
Estflpated time and proposed altitude ot arriving
. aircraft over bolding point or a.ctnal. .t'!Dte"''If ~1r.
craft is released to Approach C~ntrol after arrival
over the bolding pointo
Expe,cted approaeh time illsuedo
Statement that,aircra:fttlas been released toApproa~h Control" incltJd&.g the point or time at which
released if other than the clearance limito
Anticipated delay to departing IFR trarr1G~
Identification and destination of proposedIFR
depal'tU'lC'60 2

, !pproa~h control9! it is noted~ handles a good deal or ~hedut1es
regarding arriving and departing s,irorafto Despite the great usflfulness
~tthe, appl';'oach ~ontrol fa©li11ties in the present-day system, tbe problem or controlling traffic in and around approach control airports
still presents a good deal of compli~a~~n tor center controllerso Tbe
problem manifests itself' c.qdefly in the \ha.ndlJJlg.,Qt-.Q.§.p.a.r~ aircraft
and in the avoidance of ~onflicts between departlng aircraft and other
traffic along the airwayso

It should be noted that there is no strict standardi~tion of the
loc;ation of the airport, with respect to the range stations and airway-so
In many ~ases one leg of the range station falls across the airport,
but this 1s not always true and the airport may lie a short distance
from the airways and the range stationo This fact~ coupled with the
necessity of using different runways in different wind conditions and
the existen~e ~r various obstacl~s near the airports, requires that
various flight pathebe used in Joining tlle a;rways after a takeoff.
,;

10
20

....

.. ",.

;;

Reference' 16~ Page
Ref'eren~e l~ Page 18

Report R-203

-86=

these paths must be such as to permit the aircraft to join
the airways at positions or. altitudes at which they will not confliet
with holding traffic. The problem is complicated by the existence of
adjacent and conflicting airportB in metropolitan areas. A further complioation is that of getting aircraft to altitudes above those established minima which can be safely flown along the airways. In some
cases, this latter problem requires that the aircraft climb a~ th&
minimum altitude while shuttling back and ror~h between two points,
usually the airport and the adjacent range station At Burlington, Ver.~
mont l' aireraft climb while flying out and back along an otherwise unused leg of the range station eJEtendins over Lake Champlain
particular~

Requirements for Computer Controlled System
, ,

The preceeding section pointed out the specialized problems whic4
exiet at the airportse The problems are eeen to be greatly eased and ,.
simplified from the point of view ot the air route controller b.f the
establis~nt of approach control servicese
The specialized nature and
the individuality of the approach control p~ocedures at each airport are
such that it is somewhat doubtful if a high-speed digital computer could"
be economically and practically useful in "ihis phase of the overall
traffic problem., At the present time, special purpose equipment l as well
as radar devices 2 are being used to grea.t advantage in the operat. .ion
and improvement of the approach control service, and it would seem that
the particular nature of the approach control service is be,sm served b7
developments along these lines,
From the point of" view of using a digital computer for en-route
traffic controlJl there is little doubt that it is very ~poirtant to
separate the specialized problems of airport traffic control from the
general control of airways traffic, That is to say, it is desirable to
restrict the use of the computer to purely .§:irways control, leaving the
off=airways control to approach controllers'at the individual airports"
For this reasonj) it shall be s.ssmned that all airports in the computercontrolled system will be provided with trained personnel and the neees,..
aary radio equipment such that the approach control service can be
implemented
The standardized coordination procedures that shall then
be assumed are not in strict accordance with current practice, but are
sufficiently similar to make them Fealisti,c.g
0

1) If an aircraft approaches and desires to land at an airport equipped with an approach control marker or specific
holding point~ the pilot will receive center clearance to
this marker where he is to contact approach control for
further instructions·~ When there is no holding traffic,
the pilot will have been instructed by the computer to
approach the marker at ap altitude commensurate with an
immed ia te landing, i f there is hold ing , the cQihputer,·
clearance ~ill instruct the pilot to hold over the marker
at a specified altitude while contacting approach con~ro1o
~----------------------.-

i,> . Reference2"

322

Re:terence; _l~<

Report R-203

..J!ne.
If an aircraft approaches an airport not equipped with a
special apprca~h control marker 9 the pilot will be instructed to contact approach control at a specified flying time awayo When there is no holding trarric~ the
~omputer will have brought him down to an approach altitude by that time, if there is holding -= and the holding pattern in the absence or a speoial marker will be
anchored on the range station -- he will bave been cleared
to hold over the range station at a specified altitude
while awaiting his turn to lando

In performing and coordinating these control functions:1 the -flow of information between the computer. and the approach controllers in the towers
would be the same as Is listed on page a5~
': . .

As noted 9 p~esent-day approach eontrol procedures do not eliminate
the necessity of center controllers handling Ole~tain situations with
regard to departing aircraft9 espe~ially in regard to.getting aircraft
on the airways at proper altitudes o In accordance with the decision to
restrict computer control solely to the airways9 the follOWing aS$um.ptions are mades '
Each airport will have specialized departure procedures and
traffiC] patterns which aircraft must use in reaohing the airways from the airportso 1 The control and separation of all
departing airoraft from the pOint of takeoff to the point at
which they reach the airway~ will be the responsibility of
approach CJontrolo The computer wUl be informed by means of
the Approval Request of the point, positioD9 and altitude and
time at whi ' as 1s then knO'Wl'l) by caneel.latlon of the flight plan, leaving the airways, leaving the area, or landing
In the
case of an aircraft leaving the area, and indication should be given of which of the points
of 1 tem g) above is a. boundary point over which
control will be exercised by another computer.
0

5. Flow Diagrams
The ,generaJ.. flow diagram for the Approval Request Program
is shown. in Figure 17. As noted, five main :functions are blocked
out and numbered in this figure ~ more detailed flow diagrams for three
of these functions are given in Figures 18, 19, and 20. Throughout
the diagrams certain functions are specified which begin with the word
»checko» In all such cases it is implied that if the computer discovers
an error, the program wiU be interrupted and the computer will proceed
to another program used to notify the appropriate authorities of the
error(s)d1scoveredo

~eport

R..,.203

,v
11

Preliminary set up and storage

Perform preliminary operations associa.ted
with irtitiation of control

,I,
Check and store flight plan data

Perform preliminary operations associated
with termination of control

Prepare for succeeding action

General Flow Diagram for Approval Request Program
Figure 17

bport R-203

-99Start

rnsp~t Messa~e rd.lti£ication and

prepare to use it in guiding the course
of action of program.

Find a free register in the flight
data assignment table (see flow diagram
of Figure 14). .Store the Flight Igentification in this regis~er, and prepare
for the storage of the flight p~an data.

Store a void time for Approval Request
in present-stat~s~information register.

t
1

Store Communication Routing
Does message specify a standard type ;
of aircraft?

yes

Convert 'and ~re aircraft
Check proposed airs.eed
performance characteristio,s with standard stored val~.

1·

Store proposed
airspeed.

012/
. Flow Diagram for Block Al of Approval Request Program

Figure 18

Report R-203

-100-

from A2

Prepare to deal with first reporting
point specified in message.

Convert reporting point designation
to pure binary form. (See flow
diagram·ofFigure II.)

Check correspondence with preceding
reporting po~nt (see flow diagram of
Figure 12: this check is not made
when dealing with first reporting
point).

Jl'

Check desired altitude of aircraft
permissible flight
altitude along path between present and
preceding reporting point.

withmin~.um

. . 1

Check for conflict with any known
restrictions at reporting point.
Commute approXimatJtime at which
aircraft will pass reporting point.

po~t
~ints

Store reporting
binary designation
and appropriate time calculated above.
Have all reporting
measage been dealt

given in

wit~?

specif~~ p~t on~-;----,--y-e-s--------~

Is next
which.belongs to section of
flight plan to be relayed?

"no./'

ye~

Prepare to deal with
next reporting point

stbre temporarilY the
part of flight plan to
be relayed. ~
"

to
Flow Diagram for Block

of Ap2roval Request Program
Figure l~

fro~A3

Distinguish method of termination of contr_ol_
Aircraft leaves area

in area
airw

YSo

:Do we have part of flight plan to be relayed

Compute rate of descen per mile 'travelled.

to next area?

yes
Prepare for relay at
appropriate timeo

Prepare to extrapoltte hackwards from airport
of destination~ be.ginn;ng .. at probable- acceptance
altitude.

Prepare for notification
of succeeding area o

• t
ConS1Oder'
preced·1ng repor to1ng p01n

Prepare to extrapolate backWards from last
point under control of computer.

.j,

Consl.!"de'rpreceding

Is this ::J:t at

pamt
0

ai~ort of departure?

CheCK for correspondence
with proposed. altitude at
which, aircraft will enter

yes

Does aircraft pass this
pain t more than 30 minutes before it leaves the

ar~a?'

un~r ~nt.rol of this area?
~yes

Is the. point

area.

Prepare for immediate
coordinationo
/

Have we extrapolateq,

back'far enougb to
reach cruising
altitude?
; /

Store an appropriate indi~
cation inpoint-~9ndition
informatiOll- register.

yes

appropriate.indication
coordination and presentstatus information registers.

~---utore

Flow Diagram for A4 of Appro
Fi~!

Request Program

n~ _ __

a
I

Report R-203
-102Section A2, which has not been diagrammed in detail, will
handle and check the method bY' which initiation .is to occur. The
Approval Request will also be checked for conformance to tbe coordination regulations of page 81., At this time the computer will
make arrangements for relaying the flight plan. to a succeeding area
or for notifying that area to secure its seetion of the flight plan
data The program will make appropriate indications in the presentstatus-information and eoordtaation-infor.mation registers according
to the various conditions which existo
0

Section A5 is used in in!tiating any further action required
after the Approval Request Programo In tlios. eases in' which the aircraft is already in the air and needs immediate action on a clearance,
the computer would proceed to the Clearance Program to be discussed in
Chapter IX. In other eases, the computer would prepare to dispatch an
acknowl.edgmeat message to the orie;il'lator of the Approval Request
0

Report R-203

CRA.PfER VIII

Separation ProE~fore a clearance can be given .for an aircraft to proceed
alOBg a section of its flight, it is necessary that careful checks
be made to detemine 'Whether there is a.ny possibility of a conflict
between that aircraft and any other aircraft flying along the same
section of the airways () This chapter discusses and illustrates by
means of flQW diagrams tb,e Separa.tion Program which Yill 'be Used \in
carrying out these checking operationso

'fh~ Separation Program will actually be an auxiliary part
of the Cleara.:nce Program. of Chapter IXo As will be mentioned in that
chapter, the Clearance Program will have the ··l"esponsib11ity for
formulating the in!tial clearance for aircraft as well as extending
these clearances ate becomes necessary during the progress of the
flights
Clearances will wnutlly extend ahead of the actual position
of the a.ircraft by several reporting points, but the basic .unit of
distance in ~aling with clearances and separations will be the flight
path bet~ren succ~ssive reporting points
The Clearance Program will
supply to th~ &;pa-ration Program the identification of a partieul.ar
aireraft and the two-su~cessive :reporting points between which a. separation ~he~k 1s d~sired, the latter program will then proceed. to investigate the s~parationswh1ch exist or will exist between that aircraft
and all oth~r aircraft whose fl.ight plans have been stored by the computero
0

The end result of the use of the Separation Program will be
an indi~ation either that separation does or does not exist insofar as
the availahle data stored by the computer is complete
In this
sense, the Separation Program serves only to perform a checking
function; the decision as to what e.ction should be taken as a resul.t
of the discovery of the existence or ~ack: of separa.tion will be the
function of the Clearance Program.
As an aid in formu1at1ng the
plan IOf ac'ti,on which must be undertaken, the latte,~program will
also be supplied with a.ll available information regardiztg the conditions whi.~h are discovered to lead to a possible conf1.1cto
0

It is a suprisingly simple task for an experieneed human
controller to scan a number of flight progress strips and decide
whether or not proper separation yill exist between aircraft
The
ease with which this can be done is to a large degree attributa.ble to
a hUman. t ~. ability to visualize a four-dimensional system. -~ geographical
relatioasl1ip of airways, distance along the airways, altitude, and
ti'm.e
This ability is not inherent to the canputer, and i tis. rather
in.teresting to note in this chapter how the action of the computer
in. investigating separationlS resolTes itself into a verita.ble maze
of' comparisons and checks
0

Report B-203
-104=
In accordance with the discussion of Chapter I, it is
not to be 1'mp11ed that the flow diagrams of this chapter hau

1 <>

References 44=49"

Rep@rt R-203
-105-

wnataction
giscovered;20

shoUld be taken- to resolve' or avoid any conflict:fwh1ch- are
further comments on this fact are-made in -the next chapter.

TYpes' otSeparat1on

Aceording",to -the eli,-

regu:lati~
,

'lOngitudinal ,Vert-i-Ml'~ --or' lateral ---separations allall
qe'provided1all airoraft operating on IFr traffic
-clearance-s
•
'...
..
.

"'.,

The longitudinal separation is further defined as ~
,~,

, •. e ' .

The "'long!tUdinal'spae1ttgoftli.rC-raftt\t the same
8,1ti tude by -amini1ilWlt~i-s-tanee - -expre~sed in un!ts
ortme'~ 'so' tbat-:.atter one:~a-ireraf't'~sses o17er"a
sPeCified ~posfi!1on~" -thenixt':~i-ng 'aircraft'"
Will not arrive "'over- 'the'" same position within the
minimttm -rmmber 'Of minutes 0 2
<"

The,'mlnilin.tmt:ii1e -'separations a:re":not ,-: f±xed~' bt1.t'are ' varied'-- to meet" ce~
taineoiidlt:rons--wh16h-"arise 'The-"'detal1s" ijf 'l6rigituainal separation
0

are discus~d ~1'l- de~~l i~ Se~tion ~2.,ofthis _ cha:p~er:~
-Vertfeal. "-(itltftttde}'s-eparatfon

of aIrcraft' is fiXe,d atlOOO

in' 'the~speei8.1-:case'of"·long'transoe-_ni-e fIlgh'6s'--wliere 'S: ._,
r~"-df~fciitl separatfoilis neeessar.f'due'to' the ll:ilnited' "ra~e
o'fcruising_iUtltudes "at whien'- fuel econolDi~s"-ai.Hiieved· and "at Which
reet:-exeep't

e('mrt~s

,there is'-ncr need for

....,.-

•

US$
,1

;",

-of'--oxygeri. e-quipmento
.-'

.,.,

,;,._

La-teral-separation is defined as;
'M'"

•

Thelateral'-spacingOfaireriift at'tfle'-sa~al tItUde"

requiriiig::operatfon": Oix:"afff~t, i-OU:tes' ~"'lil' geographl«!ial"localitles as'deterriliried <= ~ the other must be made with respect to aircraft
cros~ing the paths at the reporting points or B2rcraft holding at or near
these pointso
0

It is possible to perform a check both along the path and at
each of the end=points}however~ such a method of checking becomes redundant
when applied to successive pairs of reporting points due to the .,duplicate
checkirig at the points for crossing or holding aircrafto As a means of
eliminating the double checking at these points9 the Separation Program
will operate so as to check only at the first reporting point and then
along the path to the next reporting point~ The check upon this second
reporting point,will be performed when the program is requested to check
the next pair of pointsc Any danger arising from the fact that the further.
point in each case is not checked immediately is circumvented by always
hating the Clearance Program keep well ahead of the present position of
the aircrafto The fact that most flights will terminate along the airw~s
at a special marker where they will be turned over to Approach Control will,
in general~ bypass the need for a check at a final point of the path.
In checking along the path ,between two ,points , the check 'W':Lll
usually not attempt to determine any lack of separa:tion or conflicts which
Will OCC't.lrjust beyond that path on some adjacent path. This will be left
until-a check is requested along that patho Oneexc,eption to this procedure
willibe necessary when dealing with an aircraft departing from a point
adjacent to a reporting pointe In checking the separations at the various
poin~s9 however, it will, of course~ be necessary to investigate the traffic
alon~ the connecting airwayse

Report R"",203
=108=

!?!,okell Flight Paths

The flight of an aircraft along the path b~tween two points PI and
P2 can be respresented by algebraic expressions containing the variables of
distance~ time~ and altitudeo
The flight may also be pictured graphically as
in Figure 21 where distance and altitude are plotted against time. With these
plots, the checking procedure would be to superimpose one at a time the flights
of all other aircraft travelling between the points PI and P2 and then supply
the separation criteriso
Figure 21 as sumes that the speed and the rate of change of al ti tude
a:c'e constanto The more general case which must also be considered is that of
br~en flight paths
in which aircraft will cruise and then descend or ascend,
or vice=vers8o In these cases the plots of Figure 21 will be composed of broken
lines with each part having different slopes corresponding to variations in
speed or rate of change of altitudeo (It appears that the separation standards
are sufficiently large to consider the changes in slopes as immediate rather
taking place slowly with a rounding of the plot at these pointso) Either the'
aircraft that is to be checked or any aircraft against which it is to be checked
may have such broken flight paths o As a means of convenience in checking such
paths for separation~ the general procedure will be to individually check all
parts of the broken fli ght path of one aircraft against the indi vidual par:lis
of the flight path of the other aircraft o
Co

Details of the Separation Program

general Flow Diagram

The general flow diagram for the Separation Program is given in
Figure 220 The three parts of the flow diagram which are outlined at thebottQm
of the figure are each discussed in more detail in succeeding sections of this
chaptero
The firs't step. in the program is to prepare to carry out the che~k
for separations between reporting points PI and P2 with the toiliarae'lieristics of
the flight o:f the specified aircrafto This initiil step must prepare the program
to handle a broken flight path as well as aircraft departing from airports.
With the initial preparation completed~ the program will proceed to
investigate all of the stored flight plan datao If an aircraft is found which
crosses or holds at, point PI or travels along the path between Pl and P29 that
aircraft is considered as a possible conflict and the p~ogram wiII proceed to
prepare to check against ito Similar steps as above must then be taken to
prepare to check against the possible broken flight path of this aircraft as well
as the possibility that it is a departureo
For the sake of convenience ~ the actual cheeks for separation are
divided into three categories == Longitudinal 9 Altitude~ and Holding -- each
with a separate subprograme The preparation for these checks will be such that
an ~rderly procedure is used in dealing with the various sections of broken
flight paths for each aircrafto As is shown in Figure 22, if no conflict is
found on a passage through one of the three subprograms~ a cyclic process is

DISTANCE

'P2

...

'~'~{, - ' -

-.- -

~....i.__

,I
I
I
I

,PI
.~

_______- "TI'"- _____

~-------a...--~

TIM E

' ,"T2

, (a) TIM~-DISTANCE PLOT

ALTITUDE

_________

___

~--a...

____

TIME

T2
(b) TIME-ALTITUPE P ~OT

GRAPHICAL REPRESENTATION OF PATH OF
AIRCRAFT BETW'EEN TIMES TI AND T2
,

ReportR=203
Input g

=109:Aircraft tQ be checked
for ictpaI'lation between

11 and 12

/1
Prepare for check with this aircraft
Prepare to :l.n*iate search among
st9red flight plan data

-L

1 - - - - -_ _

Search amoni stored flight
plan data

Can a possible conflict be
found

outputg
No conflict

~no

Prepare to continue to
search among stored
flight plan dat~

yes
agai~st

this aircraft

yes
Has check been completed along
broken flight paths for both
aircraft?

Distinguish typt of
check to be mad~

Longitudinal
Separation?

Altitud.
Seplratioll?

Holding
Separation?
no

Output g Conflict discovered

Genel"alFlo'W Diagram For Separation Program.
.Figure 220

yes,

Report R-203

-110=
set up until all sections of the flight paths of both aircraft have been
checked o If no conflict is found anywhere along the paths" the program
will return to continue a search among the stored flight plan data. The
ultimate result in any case~ regardless of intermediate action, will be
ei theT the discovery of a conflict between the input aircraft ani an
aircraft from the stored flight plan data or else the confirmation that
no danger of conflict exists o
2.,

Longitudinal Separation

The regulations governing longitudinal separations specify
the following separation standards~
Aircraft flying on the same or converging courses:
1)

Ten minutes if radio facilities permit frequent
determination of position and speed; otherwise
1.5 minutes
Five minutes if a preceding aircraft has filed
an airspeed at least 2.5 miles greater than that .
of a succeeding aircraft o
0

Aircraft flying on crossing courses:
Ten minutes if radio facilities permit frequent
determination of position and speed; otherwise 15
minuteo 1
A flow diagram for this longitudinal separation check is given
in Figure 23~ The two aircraft to be checked are denoted .by A and B,
where it is assumed that A~ the aircraft being checked for separation,
leaves point PI at time Tl ap,darrives at P2 at time T2 T~eQT' s
indicated in the diagram refer to the time separations of either 10 or 15
minutes as noted above
The letters OK and X are used to indicate
whether separation does or does not existc 0

There are a number of ways' in which the computer could go about
the task of detenning if the separations listed above erlst o Basically
the problem is that of detennining .if the paths of the two aircraft in a
time=distance plot as in Figure 21 cross or are separated by less than
the specified minimumo Several attempts at different methods of programming
seem to ~ndicate that the procedure illustrated in Figure 23 is the
simplesto
3.,

Alti tude. Separation

The general pr~gra:m for checking altitude_ separations is given
in Figure 240 The first steps in this program are to distinguish three
separate cases, crossing paths ~ departures, and aircraft which have been
permitted to make a non... restricted altitude changeo

'Ref erellce 16, page go

Report It=203

-111=

Input g Aircraft A a.nd B with no
lateral or vertical
separation
(See "Figure ~)
0

Is B on

Set up

i cross~~g

~~n~

same or converging course)
Set upt1 T

Does B pass PI between

::1

course?

AT : : s r +A T?

yes

r~e!t~~

B flyin: in same

o
OK

Does B pass P1 between
Tl =

and

Tl+~ T?
no

yesl

Does B

yes
Is B

pa~.s

5 and

P1 between
t1 T?

Do the time separatipns
at PI and P2 alternate

in time?
yes

mph

faster than A?

1
no

Does B pass.rl

lyes

between T1 +

Tr-

yes

and

as B pass P2 between

T2--AT and T2+J T?

yes
X \V

Is B Z5 mph,
I.' slower than .A.~J
~
yes
no
.
01(-

Flow Diagram for Longitudinal Separation Subprogramo
Figure 230

no1
OK

Report R=203
Inputg

...112=
Either or both aircraft
under consider.tion are
changing altitudao

Is the check being made against a
non-restricted aircraft?
nol

Are

\J,£--yes_-4r------t

t::i:

J Dep~tur. Separation

both d.pari:ingy::rcraft?

Is this a question of c~ossing
separation at the firstpoiot?

-l>[ CrossingS.paration

yes

Are altitude separations at both
ends of the path greater than
t

::J

1000

,'tL[__n_o_ _ _ _

1__~_6"Il

to Longitudinal
Separation

Db aircraft cross altitude

{endo>potnts~

betwe .:.
y W~

____

---'l~
if

OK for non-restticted
&1ti t 'UCle change ~

Calculate time at which
altitudes are crossed, To

Calculate position where aircraft
pass
Is this· . a.1ong the path?
0

no~

yei·

Calculate time at which aircraft
pass,

Tpl

Are aircraft flying in s arne direc ....
tion with one aircraft initiallY cruising?

~----~

I Case A.I

no\

I-------Case B

General 'Flow Diagram for Altitude Separation Subprogram.
Figure

24.

R~port
II

R=203

After eliminating the three special cases, the program proceeds to
determine i f there is an actualc..onflict .of altitudese If there is a l.ack.ot
separation at bothend.points~ . then bO.th aircraft are following the .same path
inanaltitude=distance plot and the s1 tuation reverts to one.,of longitudinal
s eparat1 on, if altitude separation. .exists at both end points and the aircraft
cross altitudes in between,9 the time thatthesealtit~des",.are eroseed.;_ Tc is
calculated e A check is then made to see if the aircraft actually pass eac,h
other, and if so" the time at which they pass, Tps> is calculated o . (These,steps
basically consist of determining if paths cross ~n the plots of Figure 210)
The separation standards regarding altitude changes are as
follows:
Alti tude Change?"'~Same_Pil"ection Traffic:
When lateral separation is not provided and an aircraft will
pass through the, altitude of another aircr~t, the following
longi tudinal separation shall be provided:
'
.~,
(1) Five minutes at the time that altitude levels ,"are
'crossed~ and provided that such separation is
authorized only when:
(a) The vertical separation at the time of the
commencement of change is 2000 feet or less; and

(b)

A leading aircraft is peing cleared for descent
through the altitude of a following aircraft, or
a following aircraft is being cleared for cl!mb
through the altitude of a leading aircraft; and

(c)

The al'ti tude change is commenced wi thin ten
minutes after the time the second aircraft has
reported over a reporting pointe

Altitude Change=-Opposite Direction Traffic:
(2) Where lateral separation is not provided, vertical separation
shall be provided for at least ten minutes prior to and after
the time the aircraft are estimatea to have passedo If reports
are received that aircraft have passed each other, this
minimum need not applYol
In view of these provisions, a division has been made into two
one which deals with the more specialized case in which both aircraft
are proceeding in the same direction and in which the first provisions listed
above apply, and the second dealing with the more general case listed immediately above o These cases are noted as Case A and B, respectively, and are
shown in more detail in Figures 28 and 290

checks~

Wbenit has been discovered that the aircraft against which the
check .is to be made is being permitted to make a non-restricted altitude change,
the'Computel" will proceed with the action dia,graromed in Figure 250 In such a
case the aircraft making the non~restricted altitude change ma,y be a~here
10

Reference

16~

Page

Input ~

Aircraft against which
check is to he made is on a
non~restricted altitude change

Is non=restricted aircraft already
on this section of the path?

yes~

Has a recent report been

received of the altitude ot
this aircraft?·

1
no

Prepare to dispatch a
messa~., to aircraft to
determine his altitude.
Set

tl~time-delayed
action

Send message
Prepare return to check against two
imaginary aircraft g ; :one cruising at
initial altitude, one cruising at tinal
altitude.

Prepare for future action that will take
p~ace if conflict is determined with
e~ther of imaginary aircrafto

~in secti~

To
of Separation Program
to determine type of ~ctiono

,Flow Diagram in

~eparation

for Checking Against a Non-Restricted Aircraft.
Figure 250

Report R&.203
betW'~en i taini tial and final al ti tudes
The first step is to deterll1:ne
if the non.,.,restrioted aircraft has begun that portion of.his flight Jif'
so., aneffert is made to. determine his present altitude as a means of
more closely restricting the regien ef al ti tude uncertainty. Tn.e computer
preceeds by making two checks, . one agains.t an. imaginarY aircraft cruising
at the 1nitia.l al ti tude and one against another aircraft cruising at the
final altitudeo If no. cenflict is discovered with eithero£ "the ,two
imaginary aircraft$ it is safe to. assume that there will be no. conflict
with the non=restricted aircrafto
0

The pertinent separation standards with regard to separatien
of departures are as follews:
a)

Fiv9=minute.separatien at the time altitude levels are
cross.edif a departure will be flewn threugh the altitude
level of . a preceding departure and both departures
propose to. follow the sameaourse o Action must be t.aken
to. ensure that the fiveaDminute separation :will be maintained er increased when altitude levels are crossed.

Three~minuteseparatien at the time coursesdiverge.if

aircraft propO,se to follow the s.arne course . . .immediately
after take~eff.and then felloW' different courses,
provided aircraft will follew diverging course.s. n.t.bin
f1 va minutes' .after take~off e. Action must.beund..-taken
to insure thatths_three...minutes.eparationwill be
maintained .or increased during the peried the aircraft
are following the same coursee l
",t
Tbeflow diagr.am fer the ceurse ef action in checking the
separations of departing . aircraft .is shown in. F.igure 260 It. is assumed
that both "aircraft _will j oin_the airways at. a . . point Qnear areperting
pointP, . and that bethaircraftw111pass peint Q at altitudes differ'ing
by less than lOOOfeeto (In some cases P and Q may be .at" the same point. )
As indicated in the now diagram, a check is first made to. seei! the
aircraft are going in different directieI;l.6 atQo If this is not the case,
then it is determined if they are proceeding to. the peintP where they
diverge 0 If the two aircraft diverge~there, then the three-minute
separations are put into. effecto In the other cases where beth aircraft
follew the same courses the separation is handled in the Longitudinal
Sep~ation Subprogram~ er else in a special manner if .the fellowing aircraft flies up through the altitude of the firsto
The manner of handling crossing aircraft where either or both
are changing altitudes is indicated in Figure 270 The chief censideration
here is that if the aircraft paS$ the point in question with less than
IO=minutes difference,jl an altitude separatien of 1000 feet must exist for
at least ten minutes before and after one ef the aircraft cresses the
pointe .

Reference 16~ Page lla.o

Report R=203

..,116-

Input:

Departing aircraft joining
airways at point Q near
repo~ting point Po

Do flights diverge

immediatelY at point Q?

1no

Do ~rcraft pass
point P within
5 minutes?

Is there a t

eeGminute
separation at point Q?

nol

yes

Do the aircraft diverge
at point P? ~

Prepare to

C::~der

next section of flight

path

Does f llowing aircraft
pass 'through altitude of
the first?
.

yes

X no

Is there three=minute

no
Check for five~minute
separations at all
~ltitudes through
~ both pass

::Sthere a three-minute
separation at P?

separation at Q?
To Ungi tUd.inal
Separation Sub...
program

yes

Conflict?

no/
OK

\Yes
X

Flow~agram

tor Departure Separations 0
Figure 26 0

Report R-203
Inputg

-117TWo aircraft whose paths
cross at reporting point Pl.

t .h

Find,the time at wh~O one a1rcra~~
passes overPlo Call the time Tao

t.'

,Find the ~ime at which the second
aircraft passes over Fl. Call the
time Tb o

ITa - T~<
yes

10 minutes?
no

1
OK

Is there a lOOO...foot difference in
altttude at Ta - 10?

yes

n~
I

Is there a lOOo-toot difference in
altitude at Ta + 101
yes

Do the altitudes cross?

Flow Diagram for Crossing Aircraft When Either or Both Are Not Cruisingo
Figure 270.

Report R-203

-118-

The flow di~irams for Cases A and B of the Altitude Separation are
given in Figures 28 and 29. The action in these cases corresponds to the provisions listed on page W.

Holding Separation

The relevant provisions for the separation. of holding aircraft are
given as follows:
When aircraft are being held in flight, the appropriate
vertical separation minimums shall be provided between
holding aircraft and en-route aircraft while such enroute
aircrafts are within, minutes' flying time of the flight
path of holding aircrafto l
The flow diagram fQr the Holding Separation is given in Figure 30 0
This diagr~ considers the action necessary when either of the two aircraft
under investigation will be holdingo In general, an aircraft will hold along
one of the ~rways, with the holding pat tern anchored on a reporting point Pl o
The standarrJ. holding pattern is to fly along the specified airways inbound to
the point" make a 1800 standard rate turn (30 per second) to the right, fly a
parallel straight course outbound for two minutes" .make another 1800 standard
rate turn to the right and aga,in fly towards Pl o As noted, if aircraft are
holding during the d~light hours at ",OO-top" altitudes, there is no need for
any separationo If both aircraft are holding and the looO-foot separation or
"50o-top" altitude conditions do' not apply, there will be a conflict; if only
one aircraft is holding a check must be made for separation at each end of the
holding pattern and special consideration must be made for the case when the
non-holding aircraft passes along the holding pattern or when it only crosses
it at Plo

R,port 11-203

-1l9a

Input g Two aircraft proceeding in
same direction, one aircraft
cruisingo
Is the initial
2

s~paration

OO;e:eet?l

less than

\
Case B

Is' speed of leading aircraft great~r
than that of fo.llowitlga1rcraft?

1,
PO::: 1 ~

yes

Will

alt~Aud.. cild~e

Case B

be colmUinc:ed

within 10 minutes after following
aircraft will pa.ss over reporting

Case J3

Is leading aircraft above and

descending?

1no

yes

1$ following aircraft below

aild aacending?

no~
Is

Case B

Tc -Tpl~ 51

ino ~
I

Set up time~dela.yed
cheek to see if aircraf'
passes over reporting
point in time o

to!{
Flow Diagram for Case A of Altitude Separationo
Figure 280

-Report R-203

Input 8

-120Two aircraft changing a1 ti tudes

altitude~separation

greater than 1000 feet
at Tp?
yes

Is altitude'separation
greater than 1000 feet

at T;e:

Is altitude separation

greater than 1000 feet

atT:.:

Are aircraft flying

in :::osite dire:~ons?
X

Is lateral separation

acc:::able at Tp~
X
OK with lateral

separation

Flow

~agramfor

Case B of Altitude Separation
Figure 290

Deport R-203

-121Two aircraft, either or
both holding.

Inputs

Are both

airC~aft holding?

lyeS .

Set up path covered
b.r holding pattern. .

during

d~tim.?

nol

Is the other aircraft
passing along the same
path?

Is there IOOO-foot
vertical separation?

yes

Are both aircraft ""aO-tops"

no ~

ye~

.~.

-.~"""~---..

Deternnne points &,c.J~ni1!b••
distance away from ends of
holding pat~

Determine points at 5 minutes
distance in either direction
PI-

frIm

Is there lOOo-foot vertical
separation at each point?
yes

Db the aircraft cross

i no

~ltitudes?

Y881
x

Flow Diagram for Holding. Separation
Figure 30.

ReportR-203

-122CHAPTER IX
Clearance Program

An air traffic control clearance is defined as:
~thorization by air traffic control, for
the purpose of preventing collision between
known aircraft, for an aircraft to proceed
under specified traffic conditions within a
control area.l -

This chapter discusses a number of considerations which deal with the
general subject of clearances, and sets forth an outline of the method
by which the computer would proceed to process and issue clearances to
the aircraft under its control. Because it is felt that further and
more exacting study must be given to ways in which a moderate amount of
human intervention may be applied, certain aspects of the Clearance
Program of this chapter are not diagrammed and are limited only to
discussion.

Processing of Clearances

Present-Day Methods

In general, the philosophy of the present system is to permit
the pilots or operations offices to select the times of takeoff and
cruising altitudes for their flights. Within this framework the controllers must provide the necessary control instructions to permit the
aircraft to safely reach these cruising altitudes and later to descend
safely on their approach 'at the airport of destination. Of course, if it
arises that the use of a particular cruising altitude or time of takeoff
will lead to a specific lack of separation with one or more other aircraft
on the airways, the control centers must be able to suggest alternate
flight plans.
Unfortunately, it is impossible under the present system operation to adequately determine at anyone time whether a particular flight
plan will permit safe separations at all points along the remainder of
the flight. As noted, this is primarily due to the fact that the traffic
control centers do not have adequate forewarning of all conditions which
will prevail over the remainder of the flight; this is not only true
of the weather conditions, but the traffic conditions as well. As
regards traffic conditions, it should be noted that under present operation all the traffic information is not concentrated at one point or
with one person, rather the information is spread out among the controllers
for the different sectors and among the centers controlling the different
areas
0

Reference 16, page "5.

Report R=20,3
It should also be realized that it is rather difficult to fly
an aircraft at a constant ground speed~ and hence it is difficult to accurately
estimate times at which aircraft will pass over future reporting pointso Pilots
can fly at a fairly constant speed relative to the air mass, but the winds may
cause rather sizable changes in the true ground speedo Since the knowledge of
the winds aloft is at best onlyhf a sketchy nature, future arrivals over the
reporting points can hardly be estimated with an accuracy of better than a few
minutesj thus making it rather futile to base separations on time estimates of
aircraft at points some distance ahead of their present positionso
Another contributing "factor which effectstne accur.acy of time
estimates of aircraft is the probabl~ lack of synchron~zation between watches of
pilots and the clocks in control centerso For this reason and those above, and
in the general interests of safety, controllers make eve~ attempt to ensure that
separations do not deteriorate to the baremin.imums of the previous chapter;
whenever po~~ible a leeway of several minutes is provided as an operating margin.
In this w~~ safe conditions are possible even if there is a ~light deviation in
the time~ that aircraft pass reporting pointso
As a result of the factors mentioned above, no attempt is now
made to provide or check for a safe channel or path for an aircraft from its
point Qf dep$rture clear through to its point of destination before the aircraft
departs along its flight, rather a flight plan is only carefully checked at one
time for separations over a section of its flight corresponding to . ,.;a few
reporting po~ntso The exact number of points depends upon the geographical
layout of th~ sectors~ traffic densities, weather conditions, etce For the most
part:J thenj) fDY planning is of a short rather than of a long term natureo This
means that c~ntrollers must be able to decide upon and use alternate procedures
at any time ~ traffic is to flow smoothly and safely.
20

Standardtzed Computer Procedures

In using the computer to process clearances, it would be desirable
to give la clearance for the aircraft over a fixed number of reporting points
ahead of its-present positiono It would aleo be desirable to give a single and
complete cle~anceto cover all sections of an~cent or descent, rather than
spli tting up each of these sections of a flight,l the single-clearance handling
of an ascent and descent will permit a complete and ct"'ordinate~. decision to be
made as to tnebest ascent or descent path (see page 137)0 A single-clearance
procedUre i$ made possible through the feature of the '-~rovisions for coordination
betweea ar~all which gi va complete control over an ascent or' descent to either .
one or the other of the two adjacent centers G
As a means of satisfying the requirements ofa standardized
procedure for handling clearances, the computer will always provide an aircraft
with a three=point clearance extending through the succeeding two and up to the
third reporting point aheado As the aircraft passes from one reporting point
to the next9 upon receipt of a progress report over the second point the computer
will proceed to extend the clearancee When the third reporting point ina

10
As used in this chapter~ ascent and descent refer to the altitude change
on departure or arrival~ not altitude changes made en routeo

Report R=203
clearance is on part of the aircraft~s descent, the clearance will be
extended ahead another point~ if necess~ry (see page 79 ), to give
complete coverage of the descent; for an aircraft initially departing
from the ground~ the three=point cle,ranee is sure to cover the complete
ascent and at least a single section of the cruisee
It will be desirable to patte~p the behavior of the computer
after the practices of the human controllers and have the 'clearances
checked for separations somewhat greater than the bare minimumso . This
can be accomplished either by increa~ing the minimums of the. Separation
Program or by checking separations nqt onfY for the true time estimates
of the aircraft but for values increased and decreased by several
minutes When the progress reports from the aircraft indicate that the
aircraft are within a permissable margi~ of safety about the previously
checked estimates, there is no need to recheck separations for the existing
clearance and the computer will proc,ed m~r~ly to extend the clear~ce;
i.f the aircraft. does not pass over tile reporting point wi thin a. specified
margin about the . checked estimate 9 tlle computer must proceed to recheck
and possibly revise the remainder of the previous clearance, as well as
extend the clearance by another pointo
0

It might prove to beadvisable$1 from considerations of safety
and reduction of computer operations" to store with the flight plan,. of
each aircraf'tthe llari.ation in time.permissib:;le at each reporting point
within which the actual time ofpas~age can fall wi thou,\ n.anger Qfconflicte
This is another of those situations in which the e;x:pendi tureof a large
amount of storage space can shorten the overall amount of computer
operationo
Although cle-arances for aircraft flying ,t500...,top:" need'. not be
checked for separations:J it will still .benecessary to proceed in advance
of the aircraft by the three=point clearance distance so that appropriate
action . c.a.nbe . ,taken.to obtain a descent clearance to br..ing the aircraft
down through the .clouds on its approacho It also appears that in the
interests of reducing complication without imposing severe and undue
restrictions on .the system, no flights should be dispatched with a
8!,OO=toptt clearance unless the clearance appears to be valid for the
entire flight on the basis of current weather reportso'

Handlin~Coordinated

Clearances

As discussed in Chapter VI~ it will be necessary for computers
of adjacent areas to coordinate clearances~ this coordination to be
accomplished by having one computer suggest a. clearance and the other
computer accept or reject ito One possible method of implementing this
8.c.tion would be to have the second computer independently devise. and.
process a clearance which could then be checked against the details of
the Q'ther proposed clearance
Such a procedure would be possible inasmuch
as the details of handling inter-area flights require.the storage of the
same data in both computers; it would be necessarYJJ however, that the
standardization of the procedures be such that both computers would be
0

Report &=203
likely to arrive at the same clearance" A somewhat simplerappr.oach would be
that of having the first computer.supplythe details of the.proposed clearance
to the second computer" This second computer would check the time estimates
of this clearance and would then process it in almost the same fasbi.on as it
would a normal clearance in checking for separationo In referring to the flow
diagrams presented later.in this chapter$it will be seen that only_relatively
minor modifications are necessary to perform this coordination action.
Bo

Conflict Elimination

A._ most formidable problem is that whichdeals ..with maldng decisions on how to direct aircraft when it is revealed, as for example by the Separation Program of Chapter Vlll~ that there will baa conflictbetwean two
aircraft A real.ization that a conflict exists will generally be discovered in
the Separati.onProgram when the computer is preparing an original clearance or
when it is attempting to readjust a clearance because_ an aircraft has not met
the condi tiona of a previous clearance
The desired action i.n such s_i tuations
is the determination of anew clearance which will not only provide for safe
separations 9) but which does not introduce long delays or inconveniences in the
flighta .As regards these latter points, it is obvious. that although in most
situations it would be possible to have one of theaircraftinvolved.in a lack
of separation 8ituation hold over a specific point for a certa.in length of time
until the danger of conflict had passed j it would be more desirable to use as
the means of conflict elimination an action such as an altitude change which
would not impose any excessive delays on the flight" It is also clear that in
any system in which the planning and conflict elimination is carried out only
over a limited section of aflight~ there is little guarantee that the measures
taken will be such that they involve a minimum amount of delay with respect to
unknown conditions which may exist later in the flighto
0

As stated in the previous chapter~ the computer will retain and
store a number of flight plans for aircraft which have been successively checked
against each other for separationo The question then arises as to how to
proceed to eliminate the conflict when a new flight plan or revision of a previously accepted flight plan does not give a proper check against the others.
One possible answer would be to consider the accepted and checked flight plans
as being fixedj) with the only variable being a flight plan for the aircraft
which does not check 6 In this case~ this aircraft alone must have its flight
plan altered 80 that it provides no conflicto
This uone=variable" scheme is by far the easiest from the programming and decision-making point of view~ and it is probably quite satisfactory
when traffic oondi tiona are not heavy and there is a. good deal of available
air space" The ideal solution, on the other handi) would be to consider all
aircraft as variables which could have their flight plans altered in accordance
with the desire for proper separation with a minimum of inconvenience and delay
for all aircraft concernedo The mechanization of such a scheme is rather complex,
however~ and it becomes a question as to whether the cost of such mechanization
in terms of computer storage and operating time is worth the end results,
especially when the computer action is balanced against the ability of a human
controller in such a situationa

ReportR=203

-126The difficulty of the mechanization is quite apparent, especially
in view of the moving block system presently employedo With a fixed
block system there area limited number of variations which ,can be made,
in the fli gh't plans of the aircraft; tha.t is 9 there'ar,e a definite number
of blocks and a defini tenumber of aircraft a.nd the problem .inalim!ted
sense is to pick the proper blocks
Such is not the case with the m'oving
block system in which there are essentially an unlimited number of changes
which can be made in'the positions and movements of the blockso
0

A consideration of the use of the fixed-block system as a means
of performing and carrying out scheduling functions for all flights will
mentioned in Chapter Xo In view of the difficulty in the use of the
computer in effecting conflict elimination for a pure moving block system,
it does seem that asamea:p;s of easing the problem it might be possible
to reach a compromise between the two extremes in such cases; one such
compromise might take the form of artificially dividing the airspace into
small volumes which could then be used in de terming what types of changes
a moving block system might employ as a means of conflict eliminationo
Another problem in the mechanization of conflict elimination is
the choice among the rather large number of alternative procedures
which can be employedo As an example of these possibilities~ consider
the following partial list~
a)

Cruising aircraft
i)

Holding
(The aircraft can be held at a particular marker,
although such action should be carefully considered
in light of delays and in possible consequence of
impeding other traffic by holdin~ at a busy reporting
point or intersection of airw~s)

:ii)

Ascent or' Descent
(Care must be taken that any ascent does not set the
aircraft at an altitude inconsistent for approach or
transfer to a holding stack, the aircraft cannot be
sent up to an altitude too high for its operating
conditions
Similar considerations apply to descents:
the aircraft should not be dispatehed to an altitude
below' the stack acceptance level or to an altitude
unsafe due to the terrain conditionso)
0

iii)

Right~Side

Separation

(This is normally only applicable to opposite
direction traffic when the aircraft are properly
situated in respect to distance from tbe range
stationo As noted, it is probably best only to
use this type of separation for experienced pilots).

Report R"",203
b)

=127=
Departing Aircraft
i)
ii)

Holding on the ground
VFR Restrictions

(Thls can be used i f weather condi tiona are
favorable)
iii)

"Stepping"
(Aircraft can be despatched to ascend, cruise, aacend-,
so as to permit safe separationo Care must be
taken to get the aircraft above the terraih mibimums'
etco<~

iv)

Non=Restricted Ascent
('rhese procedures can be used in full or the aircraft
can be requested merely to reach an altitude at a
specific time or to pass a point at a specific altitude)
j

Right~Side

Separation

Approaching Aircraft
i)

ii)
iii)

'Holding
Right=Side Separations
~Steppingn

(Certain procedures of cruise~ descend~ cruise, descent,
etc, can be utilizedo Care must be taken so as not to
place the aircraft too low for the holding stack, if
any exists)
As~lready noted~ it is not unusual at the present time for
several controllers to arrive at different solutiona when faced with the same
conflict problemo An important aspect is the previously stated fact that the
~hoice of variation or procedure must be consistent with convenience and a
mirdmum amount of delayo Hence j in factj while controllers will put certain
procedures into effect with only the desire to effect safety over a limited
section of the flightj they will choose procedures which offer a reasonable
assurance that the flight is not put into a condition where it will be handi-·
capped later as for example by being at an inconveni~nt altitude for transfer
to approach control or for landinge
j

A true appreciation of the flexibility and comprehensive nature
by which the human mind operates in a decision=making situation such as conflict

elimination can probably only be gained when one has tried to outline in a flow
diagram the mental processes which must be carried outo The author has spent
a good deal of effort in consideri.ng the most simple form of the problem in

Report R=203
which the flight plan for only one aircraft was considered as variableo
The resulting flow diagrams~ even with the most stringent of conditions
and a strict standardization of procedures, were highly complex and
lengthy
In view of this result, it is the opinion of the author that the
general problem of conflict elimination is one best suited for a certain
degree of human intervention and controlo
0

There are . 8. number of forms that the introduction of the human
controller might take o He might be presented with all the relevant
informati.on and then be permitted to propose aprocedur~ which the .comPlJt~r would then carefully check for accuracy and lack of conflicto
The
Approval Request Program calculated time estimates for all of the
reporting points over which a flight will pass; these tentative estimates
should provide. to a human a fairly good idea of the expected traffic
conditions and should be useful in guiding his deeisiono As an alternative,
the human might be consulted by the machine in only the most difficult
of caseso The author he:sitates to decide the extent to which the human
element should be introduced; in particular it is felt that a good deal
more experience in the actual operation of the system and a study of
where standardizations could be made without undue restrictiveness would
be necessary before a proper decision could be reachedo It is strongly
felt that it would be necessary to determine how efficiently, accurately,
and rapidly a human operator can make decisions concerning conflict
elimination when all the pertinent data is presente.d.to him in a suitable
formo The degree of efficiency would have to be gauged in respect to
the handling of all types of separations as well as accelerated traffic
conditions 0 From a cursory examination, it appears that a scheme in
which the human made the basic decision as to how the conflict. should be
resolved~ coupled with the use of the computer for supplying details and
for checking~ would hold a good deal of promisee
For the reasons described above~ in succeeding sections wherein
flow diagrams are presented g blocks labelled Conflict Elimination will be
used in lieu of specification of the exact means by which this action
shall be carried out&
Co

Altitude Assignments

A problem closely related to that of conflict elimination is the
assignment of cruising altitudes for aircrafte Although pilots and operation offices are permitted to initially request a cruising altitude, in
many cases it is not desirable to permit the flights to proceed at the
requested altitudea The problem is basically that of assigning aircraft
to altitudes so as to permit the optimum handling when it later' becomes
necessary to l"nd thema Officially stated~
Insofar aspracticable~ cruising altitudes of aircraft
flying to the same. destination shall be as.signed in a
manner that will be correct for an approach sequence
at destinationo

Report R=203

....129The first aircraft estimated to arrive over the point
from which approaches are commenced will normally be
the first aircraft to approacho Other aircraft will
normallY have priority in the order of their estimated
arrivals over such pointo
Altitudes at holding points shall be assigned in a
manner that will facilitate clearing each aircraft
to approach in its proper priorityo Normally the
first aircraft to arrive over a holding point should
be at the lowest altitude" with following aircraft
at successively higher altitudeso l

The situation which these regulations are intenBdto prevent are
those iri which the first aircraft to land arrives at a higher altitude thana
following ai.rcraft~ and must be brought down through the altitude of the second.
aircraft~
If the two aircraft will arrive over this point at about the same
time~ then it is seen that this changing of the altitude becomes a rather
difficult task& It is not hard to envision situations more complex than that,.;- ,
above, situations in which it is imperative that proper assignment of altit~8'
be made beforehande
The most desirable state of affairs would be that in which an
optimum assignment of altitude had been made when each aircraft entered the
control sy-sterno For the reasons noted in the preceding sections of the cha.pter..,
such preassignment is rather difficult to achieve
The only convenil3nt situ.,a.tion
in which it can. be success.fullya.ccompli~~hecl.is when the aircraft in questiom
are both dispatched frQDlthe~ame airport, in which case the action could ~e
llndertakell by the Approya.J,. Request Programc In other cases a controller handling
the sector including a particular airport will eith~:r make an att$mpt to
prQperly sequence aircraft when they enter that sector or will request the con~.
trollers '. Qf adj acent sectors to change the a1 ti tudes of aircraft in their sectors
so tha.t the proper assignment will ha.Ve.been achie:ved o . The method bywh.icb
clearanc.es i.Qr arrivals are handled will perm.it the computer to makesueh a check
on the a.ltitude assignments of all aircraft at least three -reporting points
before they begin their descento
CI

Basica.lIY3> the altitude assignment has all the aspect~Qfthe
conflict.eliminations discussed in the previous sectiQn,and the disc.us$ion
there applies to a great extent hereo As for the mechanization, it would appear
tl1at it might be carried out in pa.rt by the Approval Request Program, ,and in
part by the section of the Clearance Program dealing with arriva,lso HQwever,
as in the case of conflict elimina,tion~ a number of questions arise as to how'
the rea~signrnentsQf altitudes should be madeo It is felt thai;, this again is
an appropriate spot for human intervention~ and a.lthough the author has drawn
up several flow diagrams for the optimization of altitudes assignment by-the
Gomputerj) AereJ) again, the action will be represented in the flow diagrams by
only a. single blocko
.

Reference 16 J) page 10

Report R-203

-1;0Dc

Subdivision of

Relationship of Subprograms

Progr~

The Clearance Program logically divides itself into four main
Progress Report, Ascent Clearance, Cruise Clearance, and
Descent Clearancec The subprogrmms are schematically represented in
Figure 31 in which the g~nera1 plan of action of the Clearance Progra.
is illustratedo
subprograms~

As the names imply, the Ascent, Cruise, and Descent Clearance
Subprograms handle the clearance for those three major sections of a
flight
Initial clearances for flights are obtained via the Approval
Bequest Program or upon receipt of a Clearance Bequest message. Depending
upon the condition of the aircraft -- on the ground or in the air -- a
three-point clearance is then processed and issued" These clearances m~
be extended or revised in accordance with the results of the progress
reports received from the aircraftc The three subprograms mqy also be
entered as a result of a request for coordination on a clearance by
another areao
0

Progress Report Subprogram

The outline of action for the Progress Report Subprogram is
shown in Figure 32" Action of the subprogram will commence upon the
receipt of progress reports when aircraft pass en-route reporting
points, departure notices, holding passages, or messages from aircraft
"stepping" up through altitudes will not be handled in this program.
As previously outlined, the main function of the program
is to determine whether the aircraft is within an allowable deviation
from the time it was estimated to pass the reporting point •. If so,
action is taken to extend the clearance, otherwise the program determines
what section of the flight is being handled and a return is made to
check for separation and possibly revise the clearaneeo
30

Cruise. Clearance Subprogram

The Cruise Clearance Subprogram will be used for the following
purposes:
a)

processing a clearance following an ascent to
cruising altitudeo

processing a clearance for an aircraft entering area
at crUising altitude.

processing a clearance for an aircraft entering IFR
control while cruising
0

extending a clearance following a progress report.

checking and revising a clearance following a progress
report
0

Report, ,R-203
-131
Clearance
1
-

Request

Approval Request

Ascent Cl.a~anc.
Subprogram
Coordination
'Request

Cruise Clearanc.e
Subprogram

Progress Report
1

I
I
t
Progress ',eport
Subprogram,

Descent plearance
Subprogram

Plan of Action for Clearance Program
Figure 31.

-132Input~
progress report
Find appropri~te flight
plan in storage.

Report R-203

Is progress report correct as
regards reporting point and altitude?
yes

~gency action

Does aircraft leave control
of computer at this point?
he

yes
eaving area?,

t .

no
Is aircraftwlthin allowable
devia i01:6 from time estimate?

Coor~nation

001
Is aircraft 500-top?

Cancellati n of
stored flight plan.

~s
no
Revise time-estimates at
remaining reporting points.

leaVi~

Is aircraft
airways before next point?

/yes

Set up timedelayed action

Approach control
coordination if
necessary.

y~s

ascent?
0

Prepare to c heck and
reVise~scent clearanc

Is aircraft presentlY
descending?
yes /

I s aircraft
~in

To Mscent
Clearance

Does aircraft have
full clearance for
rest of flight?

SUbprO~ yes

Preparekto
check and revise
descent clearance

prep~:ofor
further
one-point clearance
~

To Cruise Clearance Subprogram
To Descent
Clearance
Subprogram
Progress Report Subprogram
Figure 32.

Is aircraft
in descent?

Prepare to check
and revise threepoint cruise
clearance!
To Cruise
Clearance
SUbprogram

Report R-203

-133-

The action of the subprogram is outlined in general form in Figure 33. Cases
b) and c) are first distingu!shed in the subprogram and are handled separately;
the other cases enter directly into the cyclic part· of the action which will
handle a clearance of one, two, or three reporting points depending upon the
purpose for which it is used. After all of the special conditions are distinguished and isolated, a check for separations is made with the Separation
-Program. When a conflict is qiscovered it is assumed that proper elimination
'of conflicts and assignment of altitude will be made. Following the processing
:of ·a complete clearance, the Qomputer makes appropriate ahanges in the time
estimates over the uncleared' points, sets up the condition registers, and
prepares to dispatch the cle,rance.
In this subprQgram, as well as those for the ascent and descent
clearance, when'preparing for the use of the Separation Program the times at
which the two reporting point, will be passed will be estimated. This estimate
will use the current wind capditions as well as the airspeed of the aircraft.
In accordance with current practice, the observed deviation of the aircraft as
determined in a'previous progress report, if any, will be used to obtain a
better estim.ate; such a proc,ss is equivalent to having calculated the actual
ground speed of the aircraft ~d used it for the new estimate. Little or no
use is made at the present time by the controllers of knowledge of the actual
winds which can be calculated by determi.ningthe true .ground speed and s)lbtracting fromi t. t.be .reported air speed of the aircraft; the author. has at
times.been .. amazed.at"the.accnracy with which controllers estimate times using
only the cruder methods.de.scribed abov.e,.and there appears to be little to be
gained by using a more sophi~ticated method.
Ascent and.Descent Clegrance Subprograms
1i'lowdiagrams for the Ascent and Descent Clearance Subprog:r.u
are... presentedin F'igures34 and 35. The general action. of the Ascent Clearance
Subprogram will he to process the. . clearance up to . the point of the flight at
whieh.thecruising.a1 ti.tude.i~.reac.bed; at this point the clearance "is tal<:en
.lIp. and handled by the Cruise Subprogram. The Descent Clearance Subprogram
similarly handles only the descent po.rtionof the clearance wi th suitable
coordination being made with the Cruise Clearance Subprogram.
As already noted, it will be the duty of the computer to decide
upon the manner in which aircraft will be dispatched to or from the 'Cruising
altitudes when on ascent or descent. Themost . desirable paths, in either case,
are straight-line ascents' or descents at maximum rate of change of altitude.
This line of maximum altitude change represents one boundary on the ascent or
descent, the other being the terrain considerations. If a conflict is discovered on the path of maximum altitude change, it will be the function of the
computer' :to find some other acceptable path between the two boundaries; a path
~th a slow~ rate of change of altitude or one which follows the maximum path
for only a certain distance can be tried. As a means of implementing the
conflict elimination and altitude assignment functions, in processing an ascent
clearance the computer will proceed in a forward direction along the flight
path, starting at the point at which the aircraft joins the airw~s; and in
processing a descent cl~arance the computer will consider points proceeding
back up the descent path frqm the point at which the aircraft will join the

'Reportll-203

. . .133Dl~ut

Is aircraft ~tering area
or be innin~ IFR control?

Set up to han e clearance
for required number of po;J.nt:;J

c~nsider

Prepare to
fi
point for clearance

Can fli ht be carried out
in entirety at SOD-tops?
yes

Ina

Do~ ll.ircrar'l- leave area
at -::is.1POint? Iyes

Set up for threepoint' clearanceo

Coordination

Does descent start here

or :fo1re ne,xt

r~:?

.) To Descent Clearance
S~bp:rogram.

noes aircraft leave
airway at this point?

no
Is

l~_y_e_s_·__~)IClearance completed

flig~t Sao-tops?

yes

Check for separations

conflic~1

Jseparation Program

,~_____y_e~s~________________________~____~

Conrlict·Elimination~.A.ltitude

Assignment

Have sufficient points
. been considered?

noJ,
Prepare .to
with next

yes.

~evise estimates.
Set up ~ondition
over uneleared~registers.

po in ts ..

CrUise Clearance Supprogram
Figure 33.

Prepare to
)dispateh
. lhes sage.

Report R-203

-13;,.
rn~ut

Prepare to ttart at point
where aircraft joins airways

Can ascent be mad VF.Bt

Con,ider first

ascent
Tor separation

'------------T-----C-h~~k

,,~

- (

Separation

PrOgTaDI

Conflict?
yes

------4

Conflict
Eliminatio.

Is aircraft still on
ascent at this point?

Prepare
to
yes
'V
consider next~<------'
.
succeeding point

Prepare to return to Cruise
Clearance Subprogram for
rest of clearance

\."E-k-----~>cruise

Clearance
.~r~~<~-------Program
Set up condition registers
Prepara to

bvious - that
'tbistype' 'of-1iiork~ouldbe--carried--out-rather-ea.-silyby a 'computer ; essentially.9 however 9i this flow control would be but an inherent part of

the optimumsyst-enl there are

a completely

RererenG~

ac~eduled

system o

Report R=203
The problem of :m.aking'"otttaeompleteandad:eqt2at~ ,ehedule for
other, than but"a small ,,'number ,of 'aircraft is of suffieientd!fficulty that
a high"-'speed -computer would"be r-eqtrlredo The problem- isnqt merely that
of setting up a single master:Sclledul--e, theebaracterist~es qf commercial
air pas-senger and tre1ght,travel$l-prlvate fly1ngg -and"'military demands
ah such that one c-ouldnat eXpeet-that ttwould bepossfbl-e ~ create a
strl~tsehedul. "and -foll~ito --Rather·the sehedul-e-11Bking-·ttutt.es will be
of s-ooh'a natur~'thatthey'''-mnst be oarried-otl't"lllmost contin:q:ol1s1y, as UD8~heduled 'aircrS:ft-i"fle 'fli"ght-phnsq-or--ai!" -sohe:dui-ed aircl"a;e't devu.t. sut£icientlyfromthe 'scliedtden 8.-sto'"reqaire" futher "a-ctt~o Because of
the fa~tthat a -sizeable aJlratm.to'f traffic-does follow at'a-tr1,. regular
daily and w..k1y'pattern 9 it'would'proba'blyqbepo5sibl.e toba". a' stamd.8.rd
oore tothes~hedtdesa"l"ot1l'ld lrili-chotherl!clledttlee- would b,,-t'itted o As
noted 9 all sqhedule-s,9 if they are- ·to-beeffectlve-am--Dotre~:t11re a frequent amount of change !Jshould "'be'made withsuff'i:eient lee..,. so as to
be able to take care of sudden ~hanges in wea.ther or small delays or advanee'S -whtch'ai::.rcraftmght -eneounter en""'rout.- G

The initial '~chedttlmg'-or'Tesch8dulmgq-ofrlt: clatt, is quit.
similar in na-tun-'-w-nathematical problems of lIJarlmization:ar min1m1zatioooThere anc-ertain-boundaryc-ooditions -"",desired tfme'of" depa~t-'
ur~ S> route~8peed.ge-tcO ~ "",,,,,,,,a-nd-t-here'--are 'eerta1n:restrfctlon. enforced,
~hierly 'thos-e' of' saf-e"'s-eparation;'the-" dln,ired 'resul tis- a-sehe

.AR is pr:1ma.ri1y a buffer register for pass irlg words

into AC After orders oa x, os x." ad~x, eu x, sa x, and ao x i1:;
contains the number originally. contained inregist,r x. After
orders em x, mr x, mh x, and dv x it contains the JD.8.gnitude of
the eontents of x" The effect of sp x and op x is stated below.
0

other order change, the contents of AR.

BBo- A number stored in. BR always ~ppea.rs as a positive
magnituCle, the Sign ot the ':number being assumed to be that
indicated by the sign digit in AC. This conventien has no
effect on the l.ogieal resul.t of the e>peratlms involving BR except
that when BR eamtains a number that will be used later it is
necessary to retain th~ appropriate sign digit ..
Ala.rms.

If the resul.t of an arithmetic operation exceeds

the register capacity (i "e .. , if overflow occurs), a suitable

alarm is given except as mentioned in conneeticm. with orders
sa x and 81 n ..

Shift orders
A m:W.tipl.ication overflow in 81 is lost
without giv1ng an alarm, but an overflow fl"CIIl rOlmd-off gives an'
alarm.
Orders 81" 0 and 81 0 cmly cause round-off, an ala:r.m
MUg given if an overflow occurs.. The integer n is treated
module 32, ioeo, 81 32 = 81 0, sl 33 al, 1, etc ..
<>

Scale factors
If all the digits in BR are zero and AC
eonta.:ins .... 0, the order sf x leaves AC and BR undisturbed and
stores the number 33 in the last II digit positions of
register x ..
It

Report R-203

-145Division. Let uand v be the numbers !nAC and register x
when the order dv x 1s used. If' lui < Jv I the -correct quotient is
obtained and no overflow can. arise. If' lui> Ivl overflow occurs
and gives aQ,uarm., If' u :: V + 0 the dv order leaves 16 ones
'in BR and roUnd-off ill a. subsequent sl'15 would cause overflow
and give an alarmo If u • v :: 0 a zero quotient 1s obtained.
Cheek order 'fhis order cazmot be used for anything but an
iden.tity cheeko It is intende~ pr1m8.r!lY for use in special.
test problems and in spot~eheeking to assure reliability,
especiall.y in handling film units
0

Display orders., The qk operation sets horizontal. deflection
of a.ll scopes and lea.ves it' setlUltilthe nextq1t.. A new
verticaJ.. a.ef'leet1on must," provided each time a spot is displayed
by qd or sf
The qd or qf e>pera.t1ons can be used without a qh
operatioD. ~y allowing the horizon.tal deflection to be provided
0

by a linear time-base sweep gellerator in. oae of the display
scopes!! in which' ease ecaputed values can be syn.chronized with
the sweep by allowing each aew sweep to cause the cjDPuter to

start over at the beginning

!he temporary display orders

will be repla.ced later by uiD.g an operation like rf to select

the device and operation

to put the number into it.

Orders~

Operation

Order

Name

Fll1lction.

Binary
Cede

road initially

remote unit stop

00001

Stop enemal unit

run forward

00010

Prepare to 'USe exteraaJ. uuit
k in forward. directioa.

run backward

00011

read

00100

Pre].)are to USe extenal _it
k in ba.cltwa:rd direction.
.,
..
.

rf k

rb k

'~.

oodOO

fran external: unit'internal storage is tulle

Take words

until

~..

rd x

!r8.1'U1fer to::'x.,eg1ster x a \rord
"'eraal. unit •.
supplied by''.~t~
'.~'~f ~

re x

record

00101

.Arruse for trausfer of
contents of register x to
external. unit
0

Report R-203
upera';l.on·
Order

F\1llction

BiJuJ.ry

Code
qh..x

h~is

set

display

00110

Transfer contents of AC to
r~gister x; set the horizontal
position of all display scope
beams to correspond to the·
n'UDlerieal val.ue of the
contents of AC.

OOlll

!ransfer contents of AC to
register X; set the vertical

(1l-seope

positIon of the 'beau of the
. ~isplay scopes to correspond
to the nualerical v~'" of the
contents' of AC; diSPlBl' (by
tntensit,ying) a spot Q.a the
face 01 tce D-display scopes

tax

't;a:asf'er to
storage

oiooo

Transfer cont~nts of AC
register x.

td x

tr.ansfer digits

01001

Transfer last 11 digits tram
AC to last 11 digit PQ81tiOD8
of register x. .

t~x

tnamsfer address

01010

Transfer last 11 digits f~am
AR to last 11 digit position
of register 'x.

ck x

check

01011

Stop the computer and ring en
8J.arm if the contents of register
.x is mot identical with. the
conteD.ts of' AC; otherwise proceed to next order.

. . _.

_._a.

01100

Unassigned •

01101

Exchange

'\;0

~e:conten~s of"ACt.,: .•

with the cQnwa:t;a ojt register

(original content. ot AC to
register x, . origiD.aJ. eOl1tents
of register x to AC).

11:

cp x

condition program

OlllO

If n'tllJ.1Jer in AC is nega:ti"9'e,
proceed' as in sp; if number is
positive disregard the cp order,
but clear the AR
0

Repo:rrt

R~203

-141Operation
Order

I------~-___,r___-_t_

llame

Function

Binary

Code
spx

subprogram

Ollll

Take next order trom register
xo . If tlle sp ordelf was at
address y, store y+l in
last 11 digit positions of AR.

ca x

clear and add

10000

Clear AC and BR,theD. put
conteata of register 'x ·into AC.:. "
If neceuary, add in carry from ..--

previous sa addition.
os c

clear & subtract

10001

ele~r

AC' a :m4BR, then put
complement of contents of
register.x ato AC. I f necessary, add in carry from previous
sa additiollo

ad x

su. x

em x

add

..subtra ct

clear and add

10010

Add. contents of register x to
cantents .of AC, storiBg result
in AC o

10011

SUb~raet cent••ts of registe~ x
from coatents of AC, storiJlg
result inAC o

10100

mapitude

Clear ACand JaR, the. put
'positive ma~itude of contents

of register x into AC. If neeesadd in carry from previous

sa~'y
88

sa x

special add

10101

addit1oDo

conteats of register x to
cOIltents of, AC, storing result

Add

in .AC and, Zleta iniDg any overflow for next '--eta, as, or em
order Only orders 1 through
15 .may be used. between the sa
order slld C8, es ,or em. orders
for :which the sa is a prepa
0

ration

add one

10110

Add the number ~ x 2-1 5to the
conten.ts of register x. Store
result in AC and. in register x.

qt x

F-scope

lOlll

Same as operatiOB Cld, except
display 8 spot on. the fa ce of
the F-display seopefi..

-l.48Operat1_
Faction

Order
111" X

Diul.tiply sad
rou;nd eff

11000

Multiply C011~J1ts of register
x by cOllteats of AC; nnmd off
result to 15 numerical dig1ts
ad store in AC
Clear DR.
0

mul:t1plyad hQ1d 1100ll Kult1PlJ contents of register
x by cOl1tents Qf AC aDd reta1n
the tull product ill AC aud the

first l.S

d~glt::p.s1ti0D.8of:sK,

the last digit !pu1t18 of, BR
beiDg .cleared ...

dv x

divide

11010

:Divide ecmtents of AC by cQDteais
of rep,ster x, leaving 16~-.er1ea1
dipt8 of the quotient 111 ~R sad
-4- .G ':in AC aeeordiDg to 8iga Gf

tiae qut1ento

('!.he order s1 15

f'ollow1D.g the dT erder will. rcnmd
off t~ qaotient to 15 Dwmerieal
digits aud store it·1n AC.)

shift .left

11011

MultipJ..,y the aUlliber represeated·
by the conteats of .AC and BR by r!'.
Rcnmd Qff ··the result to 1.5 DtlIIJ.er1eal
digits ad store it in AC. Dlsregardoverf'l.. ca.used by the
~ult1plicat1_,

shift right

'by rotmil-off..

'but not that eau.sed
Clear BIt ..

lllOO

the nlSberrepresuted by
the eoatsnts of AC. &Del Bll by 2-D ..
Round off the result t. 15 Jallll1erieal

Multiply

dig1ts and store it 1a AC.

sf x

seale factor

lllOl

Clear Bll.

Xult1:ply the Illlliber :represented by
the con.tents ofAC ud .13K by 2 sutfieiently often to JDake the pos1t1ve
magn1t'Wie of the product eq,1l8l. to

ar greater than 1/20 Leave the
fiaaJ. product 1m AC ad:BIto Store
the Dumber of .'t1ltlp11eat10llS 8.8
last II digits of register x, tbe
first 5 digits being undisturbed.

Report B-203

-149~rati_

Order
qr

read/

-llllO

Function

Perfora two log1eallf distinct
fttaetiQll8 :

shift right

1) Cause the interim tape
input reader to read one
eh.a.rac.ter from tape into
di&ita Q through b of FF

Sh:1.f't the ·cOIltents of AC

8dU to the right n tiaes.
The sip. digit 1s shifted

like any other digit aai
zeros·_· are introa.eter1~ties of the Magnetic Dna Binaq C-.,uter,
Eng1lJ.eer1ng Research Associates, St Paul: 1948.
0

320

rata Procedures tor Control 'rower Sur.ee Radar fltlld Ground Controlled .t..pproaeh; Civil Aero. Ad.,

Radar Procedures ... _-

vei

Washington, Do Co;

1948

330 ''What We Leaned fran the Berlin Airlift; ft bY' Ho Ao Chaffee aad
R B eorbY:I El.eetronies, Vol. 22 .No o· 9, August, 1949.
I)

340

Air Traffic Control and the National Security; .A.ir e,ordinatiag
tCamt\ittee, U S GOYel;'11llent Printing Of'lice , Washington, D. C:;:
I)

It,-,O .

Navigation and NatioD&l Security; prepared by Ad Hoc CCIIIIlittee on

lavigation,. !'he Nat1ooal.·li11:1.taryistablishm.eD.t, Resee.rek·ud Development Boat"d, WUh1ngtOll, D.. Co: 1948"
Aids to Navigat1QJ1; by Jo So Hall." et al .. , Masso Inat. 'rech.,
Radia:t1on Laboratory Series, McGraw Hill Book Coo, Inco, Vol. 2, .
19470

~a.dar

Radar Bea.ecm.s; by A Roberts, at al
M.&ss Inat of Tech
Radiation.
Laboratory Series, McGray...Hill Book Coo, Inc 0, Volo 3, 1947.
0

380 tlSuneillance Radar Deficiencies and How They Can Be OVercome; n
J Wesley Leas, rroeeedl!g of the m, Vo10 36, No.8, p. 1015,
August 1948
0

390

Distance Measuring Equipmet).t; Civil Aeroo Ad., WashirlgtOll, D. C.:
19490
.

400 tiThe Operation of VHF Omni-Range in the Transltion System;" F. L.
Mosely, Inst
410 I'Devel.opment

of Aero

f;j·t a

Sei

Reprint 240

CO'Qrse Line Can.puter tor the Air Navlgatiol1

System; nCo Wc> Watts, Sr." and F" J

Gross, !neto of Aero. Sci.

Preprint 232, 19490
420

Initial Flight iJ.'ests ad !rheorl of' an Exper:imeut~ P8.rallel
Computer,; FoJo Gross u.d Ho A.. Kay, elvU Aero Ado, Wash.. iD.gton, Do Co, 'fD Report No .. 83, September, 19480
Cc)U'rs~

430

Navigatiena.l Functions of Air 8fra.tfie Control; .General Bailway S1g11&1 Co 0 , . _chester, N. y", Pamphlet

659, 1948

440 Evolu.tion of the Integrated System €bf Air

~f'ie Coatrol; General.

Rall:way Signal Co 0, Rochester, N" Y", Pamphlet

450

652, 1,48.

'!he .Aul1eatioB of Existing Techniques to the Develo}!!e!t of. an
lJg>roved System of Air ~f'ic Control; Wo Do White, Airborne
Iutrum.ents LaboratCllry, .d~ola, Ii Yo, Report 524-1, 1947.
0

Report R-203
-15.3-

460 The Application ora. Moviy Block

S~tem to RCA Tel.erenj

Felton, Frank11n Institute, Report 3::;14, 1947"
470

Interlock Features Qt a. Moving Block System. for Selective Ass1@meat and Control gf Air ~f:1.e; Wo W. Felton ad R. S. Grubmeyer,
Franklin 1n8101tute, 1948"

490

W. w.

An EvaJ.:wa;t1e of the Traffic Capacity of a Fixed-Block ~tie
Control System in the Approaeh Zone; Walt" Rambo and R.C co WhM1er
Jr", Airoeme Instruments Laboratory, Mineola, l'i Y", Report 511t.J,
0

19410

500

A Prel:hn:1naJ::y: Aru.tlysis of ~e~hnica1.Problems Involved in the De"leloJi!l.!nt of Aitport· 'lime utilization Equipent, ANDB Technical 11_0radum. NOol; L .. Ro Philpott, 19490

Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c043 52.372728, 2009/01/18-15:56:37 Create Date : 2010:06:28 18:05:53-08:00 Modify Date : 2010:06:28 18:48-07:00 Metadata Date : 2010:06:28 18:48-07:00 Producer : Adobe Acrobat 9.32 Paper Capture Plug-in Format : application/pdf Document ID : uuid:2c931d4a-6ca6-4fef-b5b2-6988b5d9542d Instance ID : uuid:c4612856-0fc3-47fb-b0e8-cf37040b461b Page Layout : SinglePage Page Mode : UseNone Page Count : 170
EXIF Metadata provided by EXIF.tools

R 203_The_Application_of_a_High Speed_Digital_Computer_to_the_Present Day_Air_Traffic_Control_System_Jan52 203 The Application Of A High Speed Digital Computer To Present Day Air Traffic Control System Jan52

R-203_The_Application_of_a_High-Speed_Digital_Computer_to_the_Present-Day_Air_Traffic_Control_System_Jan52 R-203_The_Application_of_a_High-Speed_Digital_Computer_to_the_Present-Day_Air_Traffic_Control_System_Jan52

Navigation menu

Versions of this User Manual:

Views

Navigation