2790_ITU TSS_X.25 2790 ITU TSS X.25

User Manual: Pdf 2790_ITU-TSS_X.25

Open the PDF directly: View PDF .
Page Count: 22

Download
Open PDF In Browser	View PDF

DATAPRO

Data Networking

2790

Standards

lTV -TSS Packet Switched
Networking Standards
X Series
In this report:

Datapro Summary

X.21 Interface
Specifications ....................... 5

In 1984 the ITU-TSS (fonnerly CCITT) published standards on wide-ranging topics, including X.25 packet switching. A set of revisions to the X Series, the Blue Books, was published
in 1989. Since that time, standards ratification and publication has been an ongoing, continuous process. Each future addition or revision to the X Series standards will be made
available, as soon as it is finalized, in individual gray booklets. Since the major building
blocks of the X standards were completed by 1984, all post-1984 technical changes are
relatively minor; they are discussed in this report, however. Major developments in packet
switching center around the development of ISDN-related technologies, such as fast packet
switching and frame relay, which provide integration of voice, video, and data, and support
much higher throughput than traditional X.25 networks. ISDN's relationship with traditional
X.25 packet switching is also discussed.

Recommendation
X.25 .................................... 10
Connections Between Packet
Switched Data
Networks ............................ 16
Trends in Packet
Switching ............................ 18

A packet switched network permits a user's data
terminal equipment (PC, host computer, or terminal) to communicate with the equipment of
other geographically dispersed users. A packet
assembler/disassembler (PAD), also referred to
as data circuit-terminating equipment (DTE),
serves as a network entry/exit point, packetizing
and depacketizing data according to the rules
specified by the X Series recommendations of
the International Telecommunications Union's
Telecommunications Standardization Sector
(ITU-TSS, formerly known as CCITT).
In the early days of packet switching, each
Public Data Network (PDN) defined its own network access protocol, which permitted an appropriately equipped computer to communicate
with other devices on the network through a
physical connection to the PDN. Each of these
protocols used a multiplexing technique that enabled a computer to establish and maintain one
or more virtual circuits to other network communicating equipment. No industry standard for
packet switching existed, however, and most

-By Martin Dintzis
Assistant Analyst

C 1994 McGraw-Hili. IllCOf\lOrated. Reproduction Prohibited.
Datapro Information SenIices Group. Delran NJ 08075 USA

computer manufacturers were reluctant to provide the necessary software to handle the variety
of network access protocols.
With the adoption of the X Series Recommendations by the ITU-TSS in 1976, the PDNs
could offer a standard network access protocol.
The ITU-TSS published revisions to these standards in 1984 and 1989. Since that time, the ratification and publication of revisions have become a continuous, ongoing process.
This report focuses on Recommendations
X.3, X.28, and X.29 (informally called the Interactive Terminal Interface [ITI] standards); X.21;
X.2S; and X.7S.

Packet Assemblyl
Disassembly
Recommendations X.3, X.28, and X.29 define
the procedures by which asynchronous terminals, computers, and other devices, often referred to as data terminal equipment (DTE),
communicate with other devices via a packet
switched network. Packet assemblers/disassemblers, also referred to as DTE, commonly serve
as network entry/exit points.
X.3 defines the basic and user-selectable
functions of a PAD. It also lists 22 parameters
necessary to characterize a specific device (e.g.,
bit rate, the escape character, and flow control

FEBRUARY 1994

2790
Standards

technique). The proper setting of these values enables the PAD to
correctly interpret the communicating device and vice versa.
X.28, a related standard, defines the procedures for character
intenPange and service initialization, the exchange of control information, and the exchange of user data between an asynchronous terminal device and a PAD. X.29 dermes the procedures for
the exchange of PAD control information and the manner in
which user data is transferred between a packet mode D1E and a
PAD or between two PADs.

Recommendation X.3
TSS Recommendation X.3, Packet Assembly/Disassembly Facility in a Public Data Network, outlines the procedures for packet
assembly/disassembly in asynchronous transmissions. These
functions can 1?e programmed and built into a microprocessor- .
based "black box" that is placed between the terminal and the
X.25 network at either the customer's premises or the entry point
of the network node.
The PAD performs a number of functions, some of which allow it to be configured, by either an asynchronous terminal device or another (remote) PAD, so that its operation is adapted to
the asynchronous terminal's characteristics. The PAD's basic
functions include the following:
• The assembly of characters into packets;
• The disassembly of the user data field;
• Virtual call setup, clearing, resetting, and interrupt procedures;
• Generation of service signals;
• A mechanism for forwarding packets when the proper conditions exist;
• A mechanism for transmitting data characters, including start,
stop, and parity elements;
• A mechanism for handling a break signal from an asynchronous terminal;
• Editing of PAD command signals;
• A mechanism for setting and reading the current value of PAD
parameters;
• A mechanism for the selection of a standard profile (optional);
• Automatic detection of data rate, code, parity, and operational
characteristics (optional); and
• A mechanism for the remote DTE to request a virtual call between an asynchronous terminal and another DTE (optional).
The PAD's operation depends on the selectable values of internal
variables called PAD parameters. A set of parameters exists independently for each asynchronous terminal. The current value (the
binary representation of the decimal value) of each PAD parameter delimits the operational characteristics of the related function. The initial value of each parameter is set according to a
predetermined set of values, the initial standard profile. Twentytwo PAD parameters have been standardized by the ITU-TSS.
They are as follows:
• PAD recall using a character-allows an asynchronous terminal to initiate an escape from the data transfer state or the
connection-in-progress state in order to send PAD command
signals. This parameter has the following selectable values: not
possible, possible by character 110 (OLE), or possible by a
user-defined graphics character.
• Echo-enables characters received from the asynchronous terminal to be interpreted by the PAD and transmitted back to the
asynchronous terminal. Selectable values are no echo (0) and
echo (1).
FEBRUARY 1994

ITU.TSS Packet Switched
Networking Standards

Data Networking

X Series

• Selection of data fonvardingcharacters-allows the asynchronous terminal to send defined sets of characters, which the
PAD recognizes as an indication to complete the packet assembly and to forward a complete packet sequence as defined in
X.25. The basic functions of this parameter are encoded and
represented by a decimal value. The functions include no data
forwarding character (represented by decimal 0); alphanumeric
characters A-Z, a-z, and 0-9 (decimal I); CR (decimal 2); ESC,
BEL, ENQ, and ACK (decimal 4); DEL, CAN, and DC2 (decimal 8); EfX and BOT (decimal 16); Hr, LF, VT, and FF (decimal 32); and all other characters in columns 0 and I of International Alphabet No.5 (lAS) not included in the above (decimal
64).
• Selection of idle timer delay-permits the selection of the
duration of a time interval between successive characters.
When data received from the asynchronous terminal exceeds
this interval, the PAD terminates the assembly of a packet and
forwards it as defined in the X.25 protocol.
• Ancillary control-defines flow control between the PAD and
the asynchronous terminal. Decimal 0 represents no use of
X-on (DCI) and X-off (DC3); decimal I represents use of XonIX-off (data transfer); and decimal 2 represents the use of
X-onlX-off (data transfer and command).
• Control of PAD service signals-provides the asynchronous
terminal with the capability to decide whether and in what format PAD service signals are transmitted.
• Selection of operation of the PAD on receipt of the break
signal-after receiving a break signal from the asynchronous
terminal, the PAD may do nothing, send an interrupt packet to
a packet mode D1E or another PAD, reset, or send an indication of break PAD message to a packet mode D1E or another
PAD.
• Discard output-permits a PAD to discard the content of user
sequences in packets rather than disassembling and transmitting them to the asynchronous terminal. Selections include normal data delivery or discard output.
• Padding after carriage return-permits the PAD to automatically insert padding characters in the character stream sent to
the asynchronous terminal after the occurrence of a carriage
return character. This enables the asynchronous terminal printing device to perform the carriage return function correctly. A
value between 0 and 255 indicates the number of padding characters the PAD will generate.
• Une folding-permits the PAD to automatically insert appropriate format effectors in the character stream sent to the asynchronous terminal. No line folding or a predetermined maximum number of graphics characters per line may be selected.
• Binary speed-a read-only parameter that neither DTE can
change. It enables the packet mode D1E to access a characteristic (known by the PAD) of the asynchronous terminal device.
Speeds from 50 bps to 64K bps are represented.
• Flow control of the PAD by the startIstop mode DT&-govems flow control between the asynchronous terminal and the
PAD. The asynchronous terminal transmits special characters
to indicate whether it is ready to accept characters from the
PAD. In lAS, these special characters switch an ancillary transmit device on and off. Decimal 0 represents no use of X-on
(DCI) and X-off (DC3); decimal 1 represents use of X-onlXoff.
• Line-feed insertion after carriage return-permits the PAD
to automatically insert a line-feed character in the character
stream sent to or received from the asynchronous terminal or

j 994 McGraw-Hili, Incorporated. Reproduction Prohibited.
Datapro Information Selvices Group. Delran NJ08075 USA

'<......_-

Data Networking

ITU·TSS Packet Swltolled
Networking Standards
X . .rI••

2790

Standards

Table 1. TSS Recommendatlons-XSerie.
TSS
Recommendation

Description

X.1

International user classes of service in public data networks: class 8 (2400 bps); class 9 (4800 bps);
class 10 (9600 bps); or class 11 (48,000 bps)
International user facilities in public data networks
Packet assembly/disassembly (PAD) facility in a public data network; lists options and defaults for
interactive asynchronous terminal connection to X.25 packet networks
General structure of signals of International Alphabet NO.5 (IA5) code for data transmission over public
data networks (IA5 is described in TSS V.3)
Interface between data terminal equipment (DTE) and data circuit-terminating equipment (DCE) for
async transmission services on public data networks
V.21-compatible interface between DTE and DCE for async transmission services on public data
network
General-purpose interface between DTE and DCE for synchronous operation on public data networks
For use on public data networks by DTE that are designed to interface to synchronous V-Series
modems
List of definitions of interchange circuits between DTE and DCE on public data networks
Interface between DTE and DCE for terminals operating in the packet mode on public data networks
Electrical characteristics for unbalanced double-current interchange circuits for data communications
equipment
Electrical characteristics for balanced double-current interchange circuits for data communications
equipment
DTElDCE interface for asynchronous device access to the PAD facility of a public data network in the
same country
Procedure for the exchange of control information and user data between a packet mode DTE and a
PAD facility
Procedure for communications between users and packet networks through the switched telephone
network and through circuit switched public data networks
Expanded X.25 recommendation for internetwork communications between packet switched networks;
interface is defined between two STEs (Signaling Terminal Equipments) that are a part of ISDEs
(International Data Switching Exchanges), with expanded support for wideband links, extended
sequencing, and an expanded network utility field for international call establishment
Hypothetical reference connections for public synchronous data networks
Network parameters in public data networks
Call progress signals in public data networks
International numbering scheme for multinetwork communications containing a 4-digit DNIX (Data
Network Identification Code), 3-cligit area code, 5-digit host identification, a 0- to 2-digit subaddress

X.2
X.3
X.4
X.20
X.20 bis
X.21
X.21 bis
X.24
X.25
X.26
X.27

X.28
X.29
X.32
X.75

X.92
X.95
X.96
X.121

after echo of each carriage return character. This function applies only in the data transfer state.
• Line-feed padding-permits the PAD to automatically insert
padding characters in the character stream transmitted to the
asynchronous terminal after the occurrence of a line-feed character. This enables the asynchronous terminal printing mechanism to perform the line-feed operation correctly. This function
applies only in the data transfer state.
• Editiog-enables character delete, line delete, and line display
editing capabilities. During the PAD command state, the editing function is always available; use or nonuse of the editing
function in the data transfer state is selectable.
• Character delete, Hne delete, and line display-all editing
functions represented by one user-selectable character from
lAS.
• Editing PAD service signals-enable the asynchronous terminal to edit PAD service signals for printing devices and display
terminals; also used for editing via one character from lAS.
Editing is not selectable.
1994 McGraw-Hili. Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

• Echo mask-when echo is enabled, echo mask designates that
selected defined groups of characters sent by the asynchronous
terminal are not transmitted back. The following may be selected: no echo mask; no echo of CR; no echo of LF; no echo of
VT, HT. and FF; no echo of BEL and BS; no echo of ESC and
ENQ; no echo of ACK, NAK, STX. SOH, EOT, ETB, and
ETX; no echo of editing characters; or no echo of all other
characters.
• Parity treatment-permits the PAD to check parity in the
datastreanl from the asynchronous terminal andlor generate
parity in the datastream to the asynchronous terminal. No parity checking or generation, parity checking, or parity generation are selectable.
• Page wait-allows the PAD to suspend transmission of additional characters to the asynchronous terminal after the PAD
has transmitted a specified number of line-feed characters.

Recommendation X.28
TSS Recommendation X.28, titled DTEIDCE Interface for a
Start/Stop Mode Data Terminal Equipment Accessing the Packet
FEBRUARY 1994

2790
Standards

Assembly/Disassembly Facility (PAD) in a Public Data Network
Situated in the Same Country, describes the interfacing procedures that allow the PAD to be connected to an asynchronous
tenninal. X.28 covers four areas:
• Procedures to establish an access infonnation path between an
asynchronous tenninal and a PAD;
• Procedures for character interchange and service initialization
between an asynchronous tenninal and a PAD;
• Procedures for the exchange of control infonnation between an
asynchronous tenninal and a PAD; and
• Procedures for the exchange of user data between an asynchronous tenninal and a PAD.
Modems standardized for use on public switched or leased line
facilities establish the procedures for providing an access path
(DTFJDCE interface). Procedures for both V and X Series interfaces are defined.
Transmission speeds up to 1200 bps are specified for V-Series
interfaces; they are in accordance with either the V.21, V.22, or
V.23 standard, depending on facility type and speed. The V-Series
specifications define the procedures for setting up and disconnecting the access infonnation path by both the DTE and the
PAD.
X-Series interfaces also are used with switched or leased line
facilities. The physical characteristics for the DTFJDCE interface
are specified in X.20 or X.20 bis. Procedures for setting up and
disconnecting the path by both the DTE and the PAD are defined.
X.28 specifies procedures for character interchange and service initialization between an asynchronous tenninal and a PAD.
Characters sent and received must confonn to lAS. The PAD
transmits and expects to receive only eight-bit characters. The
eighth bit, the last bit preceding the stop element, is used for
parity checking.
X.28 describes the action the PAD takes when the value of
parameter 21 (X.3, parity treatment) is set to 0, 1, 2, or 3. If
parameter 21 is set to 0, the PAD inspects only the first seven bits
and ignores the eighth bit. When parameter 21 is set to 1, the PAD
treats the eighth bit of the character as a parity bit and checks this
bit against the type of parity-odd, even, space (0), or mark (1}used between the PAD and the asynchronous tenninal. If it is set
to 2, the PAD replaces the eighth bit of the characters to be sent to
the tenninal with the bit that corresponds to the type of parity
used between the PAD and tenninal. When the value is set to 3,
the PAD checks the parity bit for characters received from the
asynchronous tenninal and generates the parity bit for characters
to be sent to the asynchronous tenninal (as in values 1 and 2).
Once the access infonnationpath is established, the asynchronous tenninal and the PAD exchange binary 1 across the interface. This places the interface in the active link state (state 1).
When the interface is in the active link state, the DTE transmits a
sequence of characters that indicates service request (state 2) and
initializes the PAD. The service request pennits the PAD to detect
the data rate, code, and parity used by the asynchronous tenninal
(DTE) and to select the initial profile of the PAD. The service
request may be bypassed, if the tenninal is connected to the PAD
via a leased line and the PAD knows the speed, code, and initial
profile of the tenninal or if a default value is used. After the
request service signal is transmitted, the DTE transmits binary 1,
which places the interface in the DTE waiting state (state 3A). If
parameter 6 (X.3, control of PAD service signals) is set to 0, the
interface immediately enters the PAD waiting state (state S) after
receipt of service request. If parameter 6 is set to other than 0, the

ITU·TS$ Packet Swltc"eeI
Networking Stanclards
x SerIes

Data Networking

PAD transmits the PAD identification PAD service signal (indicates PAD and port identity; is network dependent), and the interface enters the service ready state (state 4). The DTE then transmits a selection PAD command signal (state 6), and the PAD
transmits an acknowledgment PAD service signal, followed by
binary 1, which places the interface in the connection-in-progress
state (state 7).
If parameter 6 is 0, the PAD will not transmit PAD service
signals. In this case, the interface is placed in the connection-inprogress state after receipt of a valid selection PAD command
signal.
Once the DTE receives the PAD service signal (state 8) or a
sequence of signals in response to a PAD command signal, the
interface is placed in either the PAD waiting state (state S) or the
data transfer state (state 9).
Afault condition exists if a valid service request signal is not
received by the PAD within a selectable number of seconds after
the transmission of binary 1. If a fault condition occurs, the PAD
perfonns clearing by disconnecting the access infonnation path.
The procedures for the exchange of control infonnation between an asynchronous tenninal and a PAD include PAD command signals, PAD service signals, break signals, and prompt
PAD service signals. PAD command signals flow from the DTE
to the PAD; they set up and clear a virtual call, select a standard
profile (pAD parameters) that is either ITU-TSS or network defined, request current values of PAD parameters, send an interrupt
requesting circuit status, and reset a virtual call. PAD service signals flow from the PAD to the DTE; they transmit call progress
signals, acknowledge PAD command signals, and transmit operating infonnation of the PAD to the tenninal. Either the PAD or
the tenninal can transmit the break signal. It provides signaling
without losing character transparency. The prompt PAD service
signal indicates the PAD's readiness to receive a PAD command
signal.
The temporary storage of characters in an editing buffer provides editing functions in the PAD. These functions pennit the
asynchronous tenninal to edit characters input to the PAD before
the PAD processes them. They include character delete, line delete, and line display. Character delete removes the last character
in the editing; line delete removes the contents of the editing
buffer. Line display causes the PAD to send a fonnat effector
followed by the contents of the editing buffer to the tenninal.
Procedures for the exchange of user data between an asynchronous tenninal and a PAD apply during the data transfer state.
The values of the parameters set in X.3 detennine which characters are transmitted during the data transfer state. For example, if
parameters 1 (pAD recall using a character), 12 (flow control of
the PAD by the start/stop mode DTE), IS (editing), and 22 (page
wait) are set to 0, any character sequence may be transmitted by
the asynchronous tenninal for delivery to the remote DTE during
the data transfer state.
User data is sent to the asynchronous tenninal in octets (eightbit characters) at the appropriate transmission rate for the asynchronous tenninal; the start/stop bits are added to the data characters. Octets are assembled into packets (see X.2S) and
forwarded when enough data has been received to fill a packet,
when the maximum assembly timer delay period has elapsed,
when a data forwarding character is transmitted, or when a break
signal is transmitted (parameter 7 is set to other than 0).

Recommendation x.a
TSS Recommendation X.29, titled Procedures for the Exchange
of Control Information and User Data Between a Packet Assembly/Disassembly Facility (PAD) and a Packet Mode DTE or Another PAD, provides the final step. X.29 describes the interfacing
procedures that allow the PAD to communicate with the X.2S

FEBRUARY 1994

1994 McGraw-Hili. Incorporated. Reproduction Prohibited.
Datapro Infonnatlon Services Group. Delran NJ 08075 USA

ITU-TIS PIIcket SWitched
Networkl... Standards
X Series

Data Networking

network. It defines the procedures for the exchange of PAD control infonnation and the manner in which user data is transferred
between a packet mode DTE and a PAD or between two PADs.
Recommendation X.29 specifies that control infonnation and
user data are exchanged between a PAD and a packet mode DTE
or between PADs using the data fields described in X.2S. Interface characteristics-mechanical, electrical, functional, and procedural-are also defined as in X.2S.
X.29 specifies that the call user data field of an incoming call
or call request packet going to/from the PAD or the packet mode
DTE must consist of protocol identifier and call data fields. A call
request packet need not contain a call user data field to be accepted by the PAD. If the call user data field is present, the PAD
transmits it, unchanged, to its destination.
A call data field's octets consist of user characters sent from
the DTE to the PAD during the call establishment phase. This
field is limited to 12 octets. The octet's bits are numbered 8 to I;
bit I, the low order bit, is transmitted first.
Bits 8, 7, 6, and S of octet 1 of a user data field of complete
packet sequences are the control identifier field. This field, which
consists of four octets, identifies the facility to be controlled. The
control identifier field coding for messages to control a PAD for
an asynchronous tenninal is 0000. When the control identifier
field is set to 0000, bits 4, 3, 2, and 1 of octet 1 are defined as the
message code field, which is used to identify specific types of
PAD messages.
User sequences perfonn data exchange. They are transferred
in the user data fields of complete packet sequences with the Q bit
set to O. Only one user sequence exists per complete packet sequence. The PAD transmits all data packets with the D bit set to O.
The DTE sends a data packet to the PAD with the D bit set to 1.
When the PAD receives a data packet with the D bit set to I, it
sends the corresponding acknowledgment. The PAD may reset
the virtual call, if it does not support the D bit procedure.

Table 2. TSS X.21 Interchange Circuits

2790

Standards

Control infonnation is exchanged via PAD messages, which
contain a control identifier field and a message code field that
may be followed by a parameter field. PAD messages are transferred in the user data fields of complete packet sequences with
the Q bit set to 1. Only one PAD message exists per complete
packet sequence. The PAD sends all data packets with the D bit
set to O. The DTE may send data packets to the PAD with both the
D bit and the Q bit set to I. When the PAD receives a data packet
with both the Q and D bits set to 1, the corresponding acknowledgment is transmitted. The PAD may reset the virtual call if it
does not support the D bit procedure. (Figure S shows the bit
positions for the Q and D bits.)
The PAD forwards a data packet when a set, read, or set and
read PAD message is received or when any of the conditions
listed in X.28 exist (e.g., enough data has been received to fill a
packet, the maximum assembly timer delay period has elapsed, or
a data forwarding character is transmitted). The PAD never forwards an empty Data packet.
By sending a set, read, or set and read message to the PAD,
. one can read and change the current values of PAD parameters.
Upon receipt of one of these messages, the PAD delivers to the
DTE any previously received data before it acts on the PAD message. The PAD responds to a read or set and read PAD message
by sending a parameter indication PAD message, which contains
a parameter field listing parameter references and current values.
Set allows the changing of parameters.
X.29 also discusses invitation to clear procedures, which are
used to request that the virtu",1 call be cleared by the PAD; interrupt and discard procedures, which are used to indicate that the
asynchronous tenninal has requested that the PAD discard received user sequences; reset procedures, as defined in X.25; error
handling procedures by the PAD, which define the actions to be
taken by the PAD when errors are detected; and procedures for
inviting the PAD to reselect the called DTE (optional), which are
used by a packet mode DTE to request that the PAD clear the
virtual call.

X.21 Interface Specifications
Interchange Name
Circuit

Signal
ground or
common
return
DTE
common
return
DCE
common
return

Transmit

Receive

Control
Indication

S
B

Signal
element
timing
Byte timing

Direction
to DCE
from DeE

Circuit

Type

Ground

X
X

Data
Transfer

X
X

Control

C 1994 McGraw-Hili. Incorporated. Reproduction Prohibited.
Datapro Infonnatlon Services Group. Delran NJ 08075 USA

Timing

The trends in communications engineering lean toward all-digital
networks and the integration of voice and data. Prospective users
of these digital, integrated networks are concerned about an interface that can provide access to voice, data, video teleconferencing, and related services. Currently, a wide variety of connectors,
electrical standards, and user procedures for various services and
networks exists-leading to almost insunnountable technical and
economical problems. Therefore, it is likely that standards organizations will develop a universal service access interface. Although it would require certain extensions, X.21 is currently the
most likely to become a future standard for a universal interface
in distributed system implementations.
TSS Recommendation X.21, Interface Between Data Terminal Equipment (DTE) and Data Circuit-Terminating Equipment
(DCE) for Synchronous Operation on Public Data Networks, defines the physical characteristics and control procedures for an
interface between DTEs and DeBs.
X.21 is the designated interface for TSS Recommendation
X.2S, a packet-switching protocol. X.21 can also be used in a
non-packet switched environment. At least two X.21-based public data circuit switched networks are currently implemented, one
in Scandinavia and one in Japan.
The X.21 standard has not gained wide acceptance in the
United States. The reluctance in the U.S. to embrace the X.21
standard is due in part to the finn entrenchment of RS-232-C.
Another factor is the cost of implementing X.21. Since X.21

FEBRUARY 1994

2790
Standards

transmits and interprets coded character strings, more intelligence
must be built into the interface, at a higher cost than traditional
pin-per-function. interfaces.
Certain characteristics of X.21 should ensure a more widespread acceptance in the coming years. One immense advantage
X.21 has over traditional interfaces is its capability to assign an
almost unlimited number of functions, because there are no functional boundaries associated with connector size. Also, X.21 offers a much more sophisticated level of control over the communications process. Another important feature of X.21 is its
inherent dialing functions, including the provision for reporting
the reasons why a call was not completed. This eliminates the
need for a separate data call interface, such as RS-232-C's companion RS-366-A, and in switched facilities, results in improved
response times.
Another important aspect is its relationship to X.25. As the
packet-switching technique becomes more widely implemented,
the demand will be greater for equipment to meet the X.2l standard. Internationally, the combination of X.2l, the ISO HDLC
protocol, and X.25 has been used to form an effective communications path. Another boost for X.21 is IBM's recognition.
X.2l has some. shortcomings. It does not permit the transmission of control information during data transfer. Also, it precludes
the insertion of data encryption hardware between the DTE and
the DCE. Another drawback is the need to modify the DTFJDCE
master/slave protocol techniques and to supply special crossover
cables to facilitate DTE-to-DTE or DCE-to-DCE interconnection.
X.21 uses a different interfacing technique than that which is
normally associated with physical-level interfaces. Instead of assigning each function a specific pin on the connector (e.g., Y.24
and EIA RS-232), X.21 assigns coded character strings to each
function.
The following is a summary of the X.21 standard, including
the functional descriptions of the interchange circuits, phases of
operation, electrical characteristics, and mechanical characteristics.

Functional Descriptions of Interchange Circuits
Four types of X.21 interchange circuits are defined: Ground, Data
Transfer, Control, and TIming. These circuits, outlined in Table 2,
are described below.
Ground and Common Return Circuits-include two types
of common return circuits, DTE Common Return and DeE Common Return, and one ground circuit, Signal Ground.
Signal Ground (Circuit G) establishes the common reference
potential for unbalanced double-current interchange circuits. If
required, it reduces environmental signal interference.
Lowering signaling rates may require two common return
conductors. In this case, two circuits, DTE Common Return (Circuit Ga) and DCE Common Return (Circuit Gb), are necessary.
For a further explanation of these circuits, see the Electrical Characteristics section of this report.
Data Transfer Circuits-include two Transmit and Receive
data transfer circuits.
Transmit (Circuit T) transfers signals from the DTE to the
DeE during the data transfer phase. It also transfers call control
signals to the DeE during call establishment and other call control phases.
Receive (Circuit R) receives signals transmitted by the DeE
from a remote DTE during the data transfer phase. This circuit
also transfers call control signals from the DeE during the call
establishment and other call control phases.
Control Circuits-include Control and Indication circuits.
Control (Circuit C) transmits signals that control the DeE for
a particular signaling process. The representation of this signal
requires additional coding of the Transmit circuit, as specified for
FEBRUARY 1994

ITU-TSS Packet Switched
Network.... Standards
X .......

Data Networking

the procedural characteristics of the tnterface. During the data
phase, Circuit C remains in the ON condition.
Indication (Circuit 1) indicates the call control process to the
DTE. The representation of this signal requires additional coding
of the Receive circuit. When Circuit I is on, it signifies that signals on the Receive circuit contain information from the remote
DTE. When Circuit I is off, it signifies a control signaling condition, defined by the Circuit R bit patterns, as specified by the
procedural characteristics of the interface.
Timing Circuits-includes Signal Element TIming and Byte
TIming.
Signal Element TIming (Circuit S) provides the DTE with signal element timing information. For this function, Circuit S turns
on and off for nominally equal periods of time.
X.21 defines different roles for the DTE and DeE in regard to
signal element timing. During the off-ta-on transition, the DTE
presents a binary signal on Circuit T and a condition on Circuit C.
The DCE presents a binary signal on Circuit R and a condition on
Circuit I during the off-to-on transition. The DeE transfers the
signal element timing across the interface as long as the timing
source is capable of generating this information.
Byte TIming (Circuit B) provides the DTE with eight-bit timing information for synchronous transmission. Use of this circuit
is not mandatory. Circuit B turns off whenever Circuit S is in $e
ON condition, indicating the last bit of the eight-bit byte. At all
other times within the period of the eight-bit byte, Circuit B remains on.

Phases of Operation
The X.2l standard defines four phases of operation: the Quiescent Phase, the Call Control Phase, the Data Transfer Phase, and
the Clearing Phase.
Each step of the operational phases places the DTE and DCE
in a certain state. See Table 3 for a listing of these states and their
associated signals on the interchange circuits.
Quiescent Phase-the quiescent phase is the period during
which the DTE and the DCE signal their capability to enter the
call control phase or the data transfer phase. It is characterized by
the appearance of basic quiescent signals from the DTE and DeE.
Various combinations of these quiescent signals result in different
interface states, or quiescent states.
There are three DTE quiescent signals. DTE Ready indicates
the readiness of the DTE to enter the other operational phases.
DTE Uncontrolled Not Ready indicates the DTE is incapable of
entering certain operational phases, usually due to an abnormal
condition. DTE Controlled Not Ready indicates that although the
DTE is operational, it is temporarily incapable of accepting incoming calls for circuit switched service.
There are two DeE quiescent signals: DeE Ready and DCB
Not Ready. DCE Ready indicates the DCE is ready to enter operational phases. DCE Not Ready indicates that no service is
available; it is also signaled whenever possible during network
fault conditions and during the period when test loops are activated.
There are six quiescent states:
• Ready
• DTE Controlled Not Ready, DCE Ready
• DTE Ready, DCE Not Ready
• DTE Uncontrolled Not Ready, DCE Not Ready
• DTE Controlled Not Ready, DCE Not Ready
• DTE Uncontrolled Not Ready, DCE Ready
See Figure 1 for a diagram of the quiescent states and the transitions that are allowed between these states.
1994 McGraw-HUI, Incorporated. Reproduction PrOhibited.
Datapro Information Services Group. Delran NJ 08075 USA

Data Networking

27"

ITU·TSS hcket Switched
NetwOl'klng S..ndllrda

Standards

X Serle.

Table 3. X.21 States: Names, Signalling, and Transitions

"'-

Signals on T,C and ft,' Circuits

Transitions'

Phase

of
Operatlon

State
Number

State Name

1
2
3
4

Ready
call request
Proceed-to-select
Selection signal

5
6A
6B
7
8
9
10

DTEWaiting
DCEWaiting
DCEWaiting
Call progress signal
Incoming call
Call accepted
DCE provided information
DCE provided information
Connection in progress
Ready for data
Data transfer

CC
CC
CC

10 bis
11
12
13

13R
13S
14
15
16

17
18
19

Receive data
Send data
DTE Controlled not ready, DCE
ready
Call collision
DTE Clear request

DCE Clear confirmation
DTE Ready, DCE Not ready
DCE Not ready
DCE Clear indication

DTE

DeE

to state
number

to state

2,138,14,24

8,13R,18
3,15

number

1
0
0

1A5

OFF
ON
ON
ON

1
1
1
1
1
1
1
1
D

.ON
ON
ON
ON
OFF
ON
ON
ON
ON
ON
ON

1
D

OFF
ON

ON
OFF

13
7

1
13

01
0
0
(see
Note)

OFF
ON
OFF

1
BEL

OFF
OFF

1,24

23
3

a
a

0
1
D

OFF
OFF
ON

0
0
0
0

OFF
OFF
OFF
OFF

CC
CC
CC

CC
CC
CC
CC
CC
CC
CC
DT

a
CC
C

+
+

SYN
SYN
IA5
BEL
BEL

1A5
1A5
1
1
D

OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON

4,15
5
6A,11,12
7,10,11,12
10bis, 11,12
6A,10,11,12
15,9

13R

68,11,12
6A,11,12
6B,11,12
12
13
13S,DCE
not ready

21
1
1,13,13S

(see
Note)
20
21
22

23
24

DTE Clear confirmation
DCEReady
DTE Uncontrolled not ready, DCE
not ready

C
C

0
0

OFF
OFF

0
1

OFF
OFF

OFF

DTE Controlled not ready, DCE not
ready
DTE Uncontrolled not ready, DCE
ready

a
a

OFF

18,22

OFF

Any
state
(see
Note)

OFF

"All other transitions are considered Invalid.
Note: DCE Clear indication or DTE Clear request may be entered from any state except Ready.
Key to Table:

Q-Qulescent Phase
CC-CaI1 Control Phase
DT-Data Transfer Phase
T-Transmit interchange circuit
c;.control Interchange circuit
R-Receive Interchange circuit
/-Indication interchange circuit

o 1994 McGraw-HiIl, Incorporated. Reproduction ProhibHed.
Datapro Information SeMces Group. Delran NJ 08075 USA

oand 1-8teady binary conditions
01-Alternate binary 0 and binary 1
X-Any value
OFF-Continuous off (binary 1)
ON-Continuous on (binary 0)
lA5-Characters from CCITT Alphabet #5
+-IA5 character 2111
BEL-IA5 character 017
SYN-IA5 character 1/6

FEBRUARY 1994

.wlto.....

ITU-TSS Pllcket
Network....· _. . . .

Data Networking

x .......

Fi;tue 1.
flrM,eMt

S,.,.,

n
I
c

Stale number
Signel on T circulI
Signal on C circuit
Signal on R clrcuH
Signal on I circuit

OCE

DCE
'

TranSition with Indlcation
of whether DTE or OCE
Is responsible fer lransltion

The above diagram indicates transitions that are allowed in X.21 networks. Other transitions are possible and may be allowed in some networks.
See Table J for a listing o/possible transitions.

Call Control Phase-the call control phase for circuit
switched· service contains many elements and procedures. Characters used. for call control are selected from lAS, a seven-bit plus
parity international code outlined in TSS Recommendation V.3.
Each call control sequence to and from the DeE is preceded by
two or more continuous SYN characters. For error checking of
call control characters, odd parity is specified.
The following elements of the call control procedure are outlined in X.21: events of call control procedures, unsuccessful call,
call collision, direct call, and facility registration/cancellation
procedure. These elements are summarized below.
The events of the call control procedures include the following:
• Call Request, signaled by the DTE to indicate a request for a
calI;
• Proceed to Select, used when the network is prepared to receive
selection information. It is transmitted by the DCE to the DTE
within three seconds of the call request signal.
• Selection Signal Sequence,tranStnitted by the DTE. A selection
sequence consists of a facility request block, an address block,
a facility request block followed by an address block, or a faCility registration/cancellation block. A facility request block
comprises one or morefacility request signals, which consist of

feBRUARY UI94

a facility request code containing one or more facility parameters. An address block contains one or more address signals.
Address signals consist of either a full address signal or an
abbreviated address signal.
• DTE Waiting.
• Incoming Call, indicated by the DCE. In response, the DTE
signals Clear Request, DTE Uncontrolled Not Ready, or DTE
Controlled Not Ready.
• DCE Waiting.
• Call Progress Signal Sequence is transmitted by the DeE to the
calling DTE to indicate that circumstances have arisen to prevent the connection from being established, to report the
progress made toward establishing the call, or to signal that
problems have been detected and that the call needs to be
cleared and reset.
• DCE-Provided Information Sequence, transmitted from the
DeE to the calling DTE. It consists of DeE-provided information blocks, such as line identification and charging information. Line IdentificatiOn is transmitted by the DeE to the calling DTE during the DCE-Provided Information state
immediately after all call progress signals, if any, are transmitted. Both calling and called line identification are optional.
Line identification consists of the international data number, as
assigned in TSS Recommendation X.121, International Numbering Plan for Public Data Networks. The DeE transmits
Charging Information during the DeE-Provided Information
C 1994 McGraw-HIH. Incorporated. Aeproductlon Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

Data Networking

ITU·TSS Packet Switched
Networking Sblndanl.

2790

X..,le.

state. It informs the subscriber of either the monetary charges
for a call, the duration of the call, or the number of units used
during the call.

• Connection-In-Progress, indicated by the DCE.
• Ready for Data, transmitted by the DCE when the connection
is available for data transfer between DTEs.
An unsuccessful call occurs when a required connection cannot
be established. In this case, the DCE indicates the failure and its
reason to the calling DTE through a call progress signal.
A call collision can occur in one of two ways: a DTE detects a
call collision when it receives Incoming Call in response to Call
Request. A DCE detects a call collision when it receives Call
Request in response to Incoming Call. When theDCE detects a
call collision, it will indicate Proceed to select and cancel the
incoming call.
The DTE indicates a request for a direct call by signaling DTE
Waiting after receiving the Proceed to Select signal. If necessary,
the DTE may choose an addressed call by presenting the correct
Selection signal.
The facility registration/cancellation procedure is optional. A
facility registration/cancellation signal consists of up to four elements in order: facility request code, indicator, registration parameter, and address signal. Not all of these elements are required
in the facility registration/cancellation signal. Also, a number of
these signals may be linked to form a block. In response to acceptance or rejection of the facility registration/cancellation action,
the network provides the appropriate Call Progress Signal.
Data Transfer Phase-when the DTE is in the data transfer
phase, any bit sequence may be transmitted. X.21 defines the data
transfer phase for three types of connections: switched; leased,
point to point; and leased, centralized multipoint.

Table 4. X.21 Pin Assignments
Pin
Number

9
2
10
3

11
4

12
5

See Note (3)
Ga
T

See Note (3)
T(B)
T(A)
C(B)
C(A)
R(B)
R(A)
I(B)
I(A)

C
R(B)
R(A)
I(B)
I(A)
S(B)

S(B)

SeA)
B(B)
B(A)

Reserved for
future use
G

7
15

Employing
X.27

Employing
X.26

Notes:
(1) X.21 pin assignments are defined by ISO 4903-1980.
(2) Where balanced ciruits are, the associated pairs are
designated "A" and "B. "
(3) Pin 1 is reserved for connection of the shield or shielded
interconnecting cable.
@

1994 McGraw-Hili, Incorporated. Reproduction Prohibited.

Datapro Information Services Group. Delran NJ 08075 USA

Standards

For operation over switched facilities, the DTE may send bits
to a corresponding DTE after receiving the Ready for Data signal.
During data transfer, control and interchange circuits are in the
ON condition, and data is transmitted over the transmit and receive circuits. Data transfer may be terminated by clearing,
which is defined below.
Two basic signals are used for operation over leased, point-topoint facilities. Send Data transmits data by the DTE on Circuit T;
the remote DTE's Receive Data signal receives data over Circuit
R. To terminate the data transfer, the DTE signal places its transmit circuit in the binary 1 condition. The DCE indicates termination of data transfer by placing its receive circuit in the binary 1
condition, its control circuit in the OFF condition, and its indications circuit in the OFF condition.
Both the central and remote DTEs use the Send Data and Receive Data signals for operation over leased, multipointfacilities.
The central DTE delivers data transmitted to all remote DTEs;
remote DTEs (one at a time) transmit data to the central DTE. A
remote DTE may send data to the central DTE while the central
DTE is sending to all remote DTEs.
Clearing Pbase-either the DTE or the DCE may initiate
clearing. The DTE indicates its desire to enter the clearing phase
by transmitting DTE Clear Request. The DCE responds by signaling DCE Clear Confirmation, followed by DCE Ready.
Clearing by the DCE takes place when it transmits DCE Clear
Indication. The DTE responds with the DTE Clear Confirmation
signal, followed by the DCE signaling DCE Ready.

Electrical Characteristics
X.21 uses two types of electrical characteristics, each for different
system requirements.
For synchronous operation at 9600 bps and below, the interchange circuits at the DCE side of the interface must comply with
TSS Recommendation X.27. The DTE side can comply with either X.27 or another TSS Recommendation, X.26. For synchronous operation at signaling rates above 9600 bps, interchange
circuits at both the DTE and DCE sides of the interchange circuits
must comply with X.27.
X.26 is defined in TSS Standard V.lO. It describes the electrical characteristics for unbalanced interchange circuits. X.26 calls
for both a DTE and DeE common grounding arrangement. The
maximum suggested cable length is 1,000 meters, and the maximum data rate is lOOK bps.
X.27 is defined by TSS Standard V.II, which describes the
electrical characteristics for balanced operation. Maximum suggested cable length is 1,000 meters, and the maximum data rate is
10M bps.

Mechanical Characteristics
The mechanical characteristics for X.21 are outlined in the ISO
Standard 4903, approved by the International Organization for
Standardization (ISO). The standard, entitled Data Communication-15-pin DTEIDCE Interface Connector and Pin Assignments, was published in June 1980.
ISO 4903 assigns connector pin numbers to a 15-pin interface
between DTE and DeE equipment. Table 4 presents a chart of
these X.21 pin assignments as they relate to the X.26 and X.27
standards.

X.21 Bis
Although X.21 is the specified interface for X.25, alternative interfaces also exist. One of these is X.21 bis.
TSS Recommendation X.21 bis, the physical and functional
equivalent to TSS V.24, defines 25 interchange circuits between
DTEs and DCEs. TSS V.24 is compatible with EIA Standard RS232-C. The X.21 bis recommendation, accepted as the interim

FEBRUARY 1994

ITU.TSS Packet Swllc"'ed

27.

Data Networking

Networking Stan""'"

Standards

x SerIes

Figure 2.
A Sampl, User T,rmilUll Conjiguration/or Operating on a Publk DIltll N,twork (PDN) in PtlCket Mode, with X.25 tlS the Int,r/ac, to th,
N,twork

User

Public Packet-Switching
Network

Location

Virtual Circuit
DTE
Computer with
X.25 Level 2
and Level 3
Software

FrontEnd
Processor

Host
Computer

DTE

I
I
I

Intelligent
Terminal with
X.25 Level 2
and Level 3
Software

I
I
I
I

I
Modem

I
I

X.25 Level 2 and 3 Software

DTE
Intelligent
Terminal with
X.25 Level 2
and Level 3
Software

•

Modem

DTE

Modem

I
I
I
I
I
I

tI)

E
G>

::::E

)(

:2
::I

::::E

Modem

I
I
I
I
I
I
I
I
I
I
I

Modem

---

DCE

Node
Processor
with
X.25
Level 2
and
Level 3
Software

I
I
I

r---

~
~

Signifies an interface that must conform to X.25 Level 2, Physical Interface Standards.

interface for X.2S, will be gradually replaced by X.21 as more
equipment is manufactured to meet X.21 specifications.

The development of Recommendation X.2S was stimulated by
the need for a standard interface between the packet-switching
networks already developed or being developed by many industrial nations and by the requirement that no terminal equipment be
denied access to packet switched services.
X.25 is a dynamic standard with many variations in the U.S.
and abroad. Currently, X.2S-based packet switched networks exist in Australia, Austria, Belgium, Canada, France, Ireland, Germany, Hong Kong, Italy, Japan, Mexico, the Netherlands, Portugal, Singapore, the Soviet Union, South Africa, Spain,
Switzerland, the United Kingdom, and the United States. Since
X.2S is a dynamic standard with many extensions and optional
features, these networks are not totally compatible with one another. Those located in Europe have the highest level of mutual
compatibility.

Since the establishment of X.2S, additional user-level protocols have been developed. These protocols provide the interfaces
between different types of terminals and the X.2S interface. X.3,
X.28, and X.29, informally called the Interactive Terminal Interface (lTI), were the first of the protocols to interface to X.2S.
They relate to the support of asynchronous, low-speed terminals
by packet switched networks. These are logical complements to
X.2S because they permit specific sets of terminals to interface to
the packet networks using the X.2S interface.
The X.2S interface standard provides for the connection of
terminals and computers to public packet-switching networks.
X.2S outlines three layers of operation: the Physical Layer, the
Link Layer, and the Packet Layer. These layers parallel the bottom three layers of the ISO Reference Model for Open Systems
Interconnection. The Physical Layer calls for TSS X.21 as the
physical and electrical interface but accepts X.21 bis, a functional
equivalent of RS-232-C, as an interim standard. The Link Layer
uses the procedures of the HDLC protocol standard. The Packet
Layer defines procedures for constructing and controlling a data
packet.

FEBRUARY 1994

Recommendation X.25

1994 McGraw-Hili. Incorporated. Reproduction ProhibHed.
Dalapro Infonnation Services Group. Delran NJ 08075 USA

Data Networking

ITU·TSS Packet Switched
Networking Standards
X Series

The 1984 revision of Recommendation X.25 added specifications for X.21 access and expanded the potential of packet operations, allowing users to actively gain access to the X.2S port,
identify themselves, and validate their connection through passwords. This change reoriented the X.25 standard toward switched
access through both X.21 facilities and the public telephone network. It now supports X.32 with regard to the public switched
telephone network or a circuit switched public data network,
dial-in and dial-out access, backup for leased line connections,
long-distance access to the network, and teletex.
Datagram was deleted in 1984, while the following packetlevel services were made available as options:
• Registered Private Operating Agent (RPOA) Selection permits the use of one or more networks to route a call to its
destination. If the user selects only one network, either the basic or extended format of the RPOA Selection can be used; if
more than one network is chosen, the extended format is used.
• Call Redirection permits the rerouting of calls if the first tried
route fails.
• Call Redirection Notification informs the recipient of the forwarded call that the call has been redirected.
• Called Line Address Modified Notification tells the caller,
within a call confirmation packet, that the call has been redirected.
• Hunt Group distributes incoming calls that have an address
associated with the hunt group.
• Charging Information gives the caller information on time
and charges and requires a new field in the call-clearing packet
format.
• Local Charging Prevention is a security facility that prevents
reverse or third-party call charges.
• Network User Identification accommodates user ID, billing,
and online facilities registration. This permits users to communicate directly with the packet data network to change the parameters of their subscriptions.
The packet level is an octet-oriented (eight bits per octet) structure. Packet sizes can vary from 1,024 to 2,048 octets, but only
within a network. Network-to-network exchange is limited to 128
octets. Closed user group facilities can accommodate very large
private-packet networks, although the number of closed user
groups to which a DlE can belong is network dependent. (See
Figure 3.)
Link-level changes implemented in 1984 created a clear separation between the Link Access Procedure (LAP) and the Link
Access Procedure Balanced (LAPB). Multilink procedures for a
single interface were also implemented. LAPB underwent an
alignment with the single-link procedure of X.7S. An extended
numbering option (modulo 128) was added to LAPB to enable the
sequencing of 127 frames and the use of satellite facilities. In
addition, the 1984 revisions to X.25 refined the procedure for the
implementation of the D-bit, polished technical accuracy, and defined the rules for new fields and formats.

X.25 Communications

(

X.25 is titled Interface Between Data Terminal Equipment (DTE)
and Data Circuit-Terminating Equipment (DCE) for Terminals
Operating in the Packet Mode on Public Data Networks. It provides a precise set of procedures for communications between
DTE and DCE for terminal equipment operating in a packet environment. The DCE in this case is a node processor that serves as
an entry/exit point on the packet network side of the user/network
interface.

1994 McGraw-HiII, Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

2790

Standards

The Data Terminal Equipment is a programmable device on
the user side of the user/network interface. The DlE is located at
the user site when the onsite equipment supports X.25; at such
installations, the DTE can be either a computer, a front-end processor, or an intelligent terminal, as shown in Figure 2, The DTE
can be a group of intelligent terminals (multiplexed to avoid the
use of multiple lines) that transmit data over the packet network
to a remotely located host. It can also be a processor acting on
calls received from multiple locations that communicate over the
packet network.
Regardless of the device or application, all DTEs present standard formatted data and control information to the DCE over
standard communications facilities. Devices operate over the network in the virtual circuit mode. Essentially, the user causes the
network to establish a logical circuit connection with the receiving station for the transmission of mUltiple contiguous packets.
(The actual physical circuit over which individual packets are
transmitted can vary during a session, but the logical circuit ensures presentation of each packet to the receiving station in the
proper order.) Information delivery is so rapid that the user appears to have an end-to-end, dedicated channel.
Users who do not support X.25 can access the public data
network for packet data transmission; however, they cannot transmit directly to a DCE or network node. They must transmit to a
special network-operated PAD, discussed earlier in this report.
Terminal transmissions are stored in a buffer at the PAD. There
they are assembled into packets and sent to the device at the other
end of the virtual circuit. Packets arriving at the PAD from the
network are reassembled into the appropriate format before they
are sent on to the terminal. The PAD is programmed and configured to interface properly with the protocol and physical characteristics of the user's device. Data presented to the PAD is reformatted into X.25 format and forwarded to the DCE.
Recommendation X.2S is divided into those specifications required for a device or network to comply and those that are optional. Approximately one-third are required; the remaining twothirds of the specifications are optional.
The excess throughput capacity inherent in the X.25 standard
allows for future network growth and technological progress. For
example, a single X.25 interface can theoretically handle 4,095
virtual channels, packet sizes up to 2048 bytes each, and packet
sequencing up to 128 packets per logical channel. Most network
suppliers' nodal processors are too small to handle this much
traffic through a single interface. Therefore, in practice, the support offered over each interface is limited to the current capacity
of the network's access node.
Functional Layers
Recommendation X.25 defines three functional layers: the Physical Layer (Level 1), the Link Layer (Level 2), and the Packet
Layer (Level 3). These are consistent with the first three layers of
the ISO Reference Model for Open Systems Interconnection
(OSI). (Although OSI labels its third layer the Network Layer, its
function parallels that of the Packet Layer. Both provide the
means to establish, maintain, and terminate connections.)
Levell: Physical Level-outlines the physical, functional,
and electrical characteristics of the D1EIDCE interface. X.21
uses the transmission of coded character strings across a 15-pin
connector to define standard interface functions, e.g., Transmit
Data. Its specifications are defined in TSS Recommendation
X.21.
Although X.21 is the recommended physical-level interface
for X.25, the availability of data communications equipment with
X.21 capabilities is limited, especially in the United States. As a
result, the lTU-TSS has accepted an interim recommendation,
X.21 bis, as the X.25 physical interface. X.21 bis is functionally

FEBRUARY 1994

ITU·TSS Packet Switched
Networking Standards
X SerIes

2790
Standards

Data Networking

Figure 3.
FrtIIIUI FomuJIll/or Plleut Mode Tmnllmisllion

~I""----------------------------Frame·""'---------------------------"~I
...
.....-------packet-----t.~1
.
Flag

Control
Field

Address

Packet
Control
Information

User Data

Field
Sequence
Check

Flag

Fields Generated by:
Application Program
Level 3 Software
Level 2 Software

Size,

.bl1I

Description

Value is 01111110.

Value for DTE to DCE command frame is "10000000".
Value for DTE to DCE response frame is "11000000".

Flag
Address

Value for DCE to DTE command frame is "11000000".
Value for DCE to DTE response frame is "10000000".

8
Control

I Frame:
Bit 1 is "0".
Bits 2, 3, 4, are N(S) (transmitter send sequence count).
Bit 5 is P/F (PoIVFinal) bit.
Bits 6, 7, 8 are N(R) (transmitter receive sequence count).
S Frame:
Bit 1 is "1".
Bit 2 is "0".
Bits 3, 4 identify Supervisory Command.
Bit 5 is P/F (Poll/Final) bit.
Bits 6, 7, 8 are N(R) (transmitter receiver sequence count).
U Frame:
Bitlisl.
Bit2 is 1.
Bits 3, 4 are first part of Unnumbered Frame Command identifier.
Bit 5 is P/F (Poll/Final) bit.
Bits 6, 7,8 are second part of Unnumbered Frame Command identifier.

Packet
Control
User Data

24
1,024
16

Control information.
1,024 data bits are normal maximum; exceptional maximum is 2,040 data bits.
Check bits for user data.

Field Sequence
Check
8

Value is "0111 H 10".

Flag

equivalent to EIA ~.s-232-C, which assigns a separate function to
each pin across a 25-pin connector.
Level 2: Link Level-describes the link access procedures
used for data interchange between a DCE and a DTE. In the TSS
Recommendation X.21, International User Classes of Service in
Public Data Networks, X.25 specifies that the DTE and DCE
must operate in the following user classes of service: class 8
(2400 bps), class 9 (4800 bps), class to (9600 bps), or class 11
(48K bps). The procedure calls for a full-duplex facility so that

the DTE and DCE can conduct two-way, simultaneous, independent transmissions. The procedures use the principles and terminology of the High-level Data Link Control (HDLC) protocol
specified by the International Organization for Standardization.
Level 2 comprises three procedures: the Link Access Procedure, the Link Access Procedure Balanced, and the Multilink Procedure (MLP). The original X.25 data link control procedure embraced LAP, which is similar to the HDLC Asynchronous
Response Mode (ARM). Due to some incompatibilities with

FEBRUARY 1994

1994 McGraw-HiII, Incorporated. Reproduction Prohibited.
Oatapro Information Services Group. Delran NJ 08075 USA

Data Networking

ITU·TSS Packet Switched
Networking Standards

2790

Standards

X Series

Table 5. X.25 Level 2 Commands and Responses
Commands (1)

Responses

Format
(information)

Information
transfer
Supervisory

Unnumberad

RR (3)
RNR (3)
REJ (3)

(receive ready)
(receive not ready)
(reject)

RNR
RNR
REJ

(receive not ready)
(receive not ready)
(reject)

SARM
(2,3)

(set asynchronous
response mode)

OM
(2)

(disconnected
mode}

SABM
(2)

(set asynchronous
balanced mode)

DISC

(disconnect)

Encoding of Control Field
4
5
8

I-N(S}-I P

0
0
0

0
1

0
0

P/F

N(R}
N(R}
N(R}

CMDR
FRMR

(unnumbered
acknowledgment)
(connand reject)
(frame reject)

I--N(R}-I

P/F
P/F
P/F

0
0

(1) The need for, and use of, additional commands and responses are for further study.
(2) DTEs do not have to implement both SARM and SABM; furthermore, OM and SABM need to be used if SARM only is used.
(3) RR, RNR, and REJ supervisory command frames are not used by the DeE when SARM ;s used (LAP).

some elements of procedures, Level 2 of the X.25 Recommendation was revised to incorporate LAPB. Similar to the Asynchronous Balanced Mode (ABM) of HDLC, it provides for a
"balanced" configuration; that is, each side ofthe link consists of
a combined primary/secondary station. The Multilink Procedure
is used for data transmission on one or more single links. Each
link conforms to the X.25-defined frame structure and to the elements of procedure described in LAPB.
Software in both DTE and DCE performs Level 2 processing.
This software appends control information onto packets that are
ready for transmission, maintains control of the transmissions,

performs transmission error checking, and strips a successfully
received frame down to a packet. The packet consists of packet
control information and (optionally) user data; packets are discussed in detail later in this report.
HDLC specifies certain control fields that must be appended
to both ends of a packet, resulting in a transmission format called
a frame. Appended in front of a packet are a beginning Flag field,
an Address field, and a Control field. Appended behind the packet
are a Frame Check Sequence field and a closing Flag field.
Figure 3 gives the size and description of each field. The receiving device uses the two Flag fields to ascertain the beginning

Table 6. Summary of Packets Exchanged Between Terminals During a Virtual ~all
Events

Activity at Calling DTE

Activity at Called DTE

Call Initiation

Call Request packet is sent

Incoming Call packet is received
Call Accepted packet is sent

Call Connected packet is received

Data Transfer

Disconnect

Data packet sent

Data packet received
Ready Receive packet sent

Ready Receive packet received
Data packet A sent'
Data packet 8 received'

Data packet 8 sent'
Data packet A received'

Clear Request packet sent

Clear Indication packet received
Clear Confirmation packet sent

Clear Confirmation packet received

"Two-way (full-duplex) transmission of data packets between terminal eqUipment.
@ 1994 McGraw-Hili, Incorporated. Reproduction Prohibited.

Datapro Information Services Group. Delran NJ 08075 USA

FEBRUARY 1994

Data Networking

ITU.TSS Packet Switched

2790
Standards

Networking Standard.
X Serl••

Figure 4.
X.25 User and Network Software Relationships
1-------UserLocalion - - - - - - I - - - - - C o m m u n i c a t i o n s - - - - - - t - - - - - - P u b I i C PacketFacility
Switching Network

DCE
DTE

Network Node Procassor

r - ---Communicetions Accass Method

User
Program

r----., r----.,
1 X.25
1 Level 3.

1 1 X.25
1 1 Level 2.

1
1

1_ P~ck~t ~~I...! 1_ F~~ ~v~ ...!

l ,~~"

I - - - - --I

1
1
I

1
1
I X.25 Level 3.
I Packet Level

1
1
1

I
I

1
1

X.25 Level 2.
Frame Level

To
Other
Network
Nodes

L _____ .1 ______ I

Physicallnlerfaca

l
and ending of a frame. A single Flag field can be used as the
closing flag for one frame and the opening flag of the next frame.
The Address field, under HDLC, is used to identify the station(s)
for which the command is intended in command frames or to
identify the station sending the response in response frames.
The Control field identifies the type of frame and supplies
control information pertinent to that type of frame. A frame can be
either an Information Frame (I-Frame), a Supervisory Frame (SFrame), or an Unnumbered Frame (U-Frame). See Table 5 for the
encoding of this field.
An Information Frame contains a user packet. The Control
field of the I-Frame contains the N(S) transmitter Send Sequence
Count, the N(R) transmitter Receive Sequence Count, and the
PolllFinal (P/F) bit. N(S) is the sequence number of this frame
sent from this transmitter to this receiver. N(R) is the sequence
number of the next frame this transmitter expects to get from the
receiver when the receiver becomes a transmitter. The PoIIlFinal
bit indicates whether a transmission acknowledgment is required
or when the final frame in a stream has been transmitted.
The Supervisory Frame transmits one of three supervisory
commands and cannot contain user data. The Control field of the
S-Frame contains the command, the P/F bit, and an N(R). Table 6
summarizes the commands and responses possible in the Control
field.
The Unnumbered Frame extends the number of link control
functions, without incrementing the sequence counts at either the
sending or receiving station.
A U-Frame control field contains the Unnumbered command
and the P/F bit. One of the U-Frarne commands initializes the
DTFJDCE network to the ARM of operation, which permits both
DTE and DCE to initiate transmission to each other.
The Frame Check Sequence (FCS) field contains a 16-bit
check pattern created at framing time by generating check bits
based upon the binary value of the data in the packet. The receiving device performs the same calculation and compares its results
with the value that the sending device had placed in the FCS field.
If these values do not agree, the frame has a transmission error
and is discarded. Discarding causes the receiving device to send
an S-Frame with the Reject Recovery command. This S-Frame
contains the frame numbered N(R), thus acknowledging receipt

FEBRUARY 1994

X.25 Levell.
Physlcat Interface

of all frames numbered N(R)-I and below, and initiates retransmission of frames N(R) and above.
In effect, Level 2 envelops the packet in control information
and prescribes procedures that ensure a high degree of transmission accuracy and detection of lost frames during transmission.
Level 3: Packet Level-describes the packet format and control procedures for the exchange of packets between the DTE and
the DeE. In addition to the user data packet format, there are
several formats for DTFJDCE administrative messages.
The software for formatting packets and controlling packet
exchange is the Level 3 software resident in both the DTE and the
DCE. Figure 4 shows the relationship of Level 3 and Level 2
software operating in the DTE and the DeE. In the DTE, a user
application program normally presents the data to be transmitted
to the operating system's communications access method. The
access method invokes Level 3 software to append the packet
header information, then invokes Level 2 software to create the
frame. Packet header formats are discussed below.
Once the frame is created, it is ready for transmission. Upon
receiving the frame, the receiving DCE invokes Level 2 software
to perform error detection and sequence checking and to strip the
accepted frame down to a packet. The packet is presented to
Level 3 software for packet-level processing and to prepare the
packet for transmission to the destination DCE. The physical address of the destination DTE, derived from the caIl initiation, is
inserted into the packet during Level 3 processing. At the network
destination DeE, Level 3 software reformats the packet control
information and presents the resulting packet to Level 2 software;
Level 2 software includes the packet in a frame for transmission
to the destination DTE (user). At the destination DTE (user),
Level 2 software performs error detection and sequence checking
and strips accepted frames down to the packet. Level 3 software is
then applied to the packet to strip the header information from the
packet and pass the user information to the appropriate application program via the communications access method. Table 7
summarizes the types of packets exchanged between terminals.
The packet header consists of three octets. In Octet I, the first
four bits represent the Logical Channel Group Number, and the
final four bits represent the General Format Identifier. Octet 2
represents the Logical Channel Number, and Octet 3 represents
the Packet Type Identifier. See Figure 5.
@

1994 McGraw-Hili, Incorporated. Reproduction

Oatapro Infonnalion Services Group. Delran

Prohlb~ed.

NJ 08075 USA

Data Networking

ITU·TSS Packet Switched
Net_rldng Standa"'"
xSerie.

2790

Standards

Table 7. Packet Types and Their Use in Various Functions

Function

Can Setup and Cleaning

Data and Interrupt

Flow Control and Reset

Restart

Diagnostics

Packet Type
From DCE to OlE
From OlE to DCE

In Incoming Call
Call Connected
Clear Indication
DCE Clear
Confirmation
DCE Data
DCE Interrupt
DCE Interrupt
Confirmation
DCE RR (Receive
Ready)
DCE RNR (Receive
Not Ready)
Reset Indication
DCE Reset
Confirmation
Restart Indication
DCE Restart
Confirmation
Diagnostics

Call Request
Call Accepted
Clear Request
DTE Clear
Confirmation
DTE Data
DTE Interrupt
DTE Interrupt
Confirmation
DTERR
DTE RNR
DTE REJ*
Reset Request
DTE Reset
Confirmation

X
X
X

Restart Request
DTE Restart
Confirmation

The Logical Channel Group Number and the Logical Channel
Number provide the routing information that directs the packet
over one of the logical channels. The numbering system for the
logical channels is dynamic and refers to a switched data path
within the packet network.
The General Format Identifier indicates the general format of
the rest of the header. Specific bit patterns are established for the
following: Call Setup packets; Clearing, Flow Control, Interrupt,
Reset, Restart, and Diagnostic packets; and Data packets.
The Packet Type Identifier establishes the packet's function.
Examples of functions include call setup, priorities, and data. If
applicable, it may identify the packet's place in sequence. See
Table 7 for the various packet types.

Packet·Level Procedures
X.25 Level 3 defines procedures for call initiation, data transfer,
interrupts, reset, restart, and clearing. Some of these procedures
are summarized below.
Call Initiation: The Level 3 software in a DTE initiates a
transmission by sending a Call Request packet. This packet enables the calling DTE to request the opening of a logical channel.
The calling DTE designates the channel number based upon the
original assignments that were made when the user subscribed to
the network. The Call Request packet also informs the network of
the calling DTE's address and of the called DTE's address. Until
the call is disconnected, the network retains the addresses of both
devices associated with the logical channel number. Therefore,
each Data packet that is transmitted needs to contain only the
logical channel number. The network appends the destination address just before routing the packet. At the time the Call Request
is issued, the logical channel indicated by the calling DTE must
be in a Ready state, that is, not being used to handle another call.
Upon receipt of the Call Request by the called DTE, the specified
logical channel is designated as being in the DTE Waiting state.
The packet is then transmitted to the destination DCE.

1994 McGraw-Hili. Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

Service
Switched
Virtual
Circuit

Perm.
Virtual
Circuit

X
X
X

X
X

X
X
X
X
X

X
X

The destination DCE transmits an Incoming Call packet to the
originating DTE and places the logical channel in the DCE Waiting state.
The called DTE indicates its willingness to accept the call by
transmitting a Call Accepted packet across the DTFlDCE interface. The Call Accepted packet specifies the same logical channel
that is indicated by the Incoming Call packet. This places the
specified logical channel in the Data Transfer state. (The logical
channel at the calling DCE is still in the Wait state.)
A Call Connected packet is then sent by the DCE to the calling
DTE and sets the logical channel status to the Data Transfer state;
the logical channel is then ready for transfer of Data packets. This
applies only for virtual circuit connections; for permanent virtual
circuit connections, the assigned logical channels are always in
the Data Transfer state.
It is possible for a DCE to receive a Call Request (from a
DTE) and an Incoming Call (from the network) simultaneously,
with both messages specifying the same logical channel. This is
called a call collision; when a collision occurs, the DCE cancels
the Incoming Call and proceeds to process the Call Request.
nata Transfer: Once the logical channel is in the Data Transfer state, Data, Flow Control, or Reset Information packets can be
transmitted.
The Data packet format includes the packet send and receive
sequence numbers. The calling DTE can transmit up to a predetermined number of packets before requiring a response. This
number is referred to as the window. All packet networks must
accommodate a lower window edge of at least eight, permitting at
least seven packets to be outstanding before an acknowledgment
is required. The upper window value (the maximum number of
outstanding packets) may not exceed 127 (modulo 128).
Upon receipt of an authorized packet, the receiving device can
either authorize additional transmission by transmitting a Receive
Ready packet to the sending device or deny authorization to transmit by issuing a Receive Not Ready packet.
When transmitting Data packets, the Interrupt packet can bypass flow control (authorization procedures). A DTE can transmit
FEBRUARY 1994

ITUoTSS Packet SWitched
Networking StandIIrcI.
X Serle.

2790
Standards

Data Networking

Fig/ll'e 5.
X.25 PIIe"t He_r Fot7nllI

Octet 1

Octet 2

1 2 3 4: 5 6 7 8 1 2 3 4 5 6 7 8
Logical
Channel
Group
Number

General
Format
Identifier

Logical Channel Number

an Interrupt packet to another DTE. To avoid swamping the receiver with Interrupt. packets, the sender cannot send a second
Interrupt packet until the first Interrupt packet is acknowledged
by an Interrupt Confirmation packet.
While Data packets or Interrupt packets are being transmitted,
issuing a Reset Request packet can reset the call. A DCE initiates
a reset by issuing a Reset Indication packet. In either case, the
logical channel is placed in the Data Transfer state. Any Data,
Interrupt, Receive Ready (RR), or Receive Not Ready (RNR)
packets in the network at the time of reset are ignored. If a DTE
and DCE transmit Reset packets simultaneously, a reset collision
occurs. The net effect is to place the logical channel in the Data
Transfer state, ready for Data and Interrupt packets.
A DTE or a DCE initiates a request for clearing (disconnecting) a logical channel. Upon confirmation by the other device, the
logical channel is placed in the Ready state.
Error Handling (Packet Level): When an error occurs, the
DCE transmits Reset, Clear, and Restart indication packets to the
DTE. Reset reinitializes a virtual call or permanent virtual circuit.
Reset removes all Data and Interrupt packets on the logical channel and resets the lower window edge to zero. It affects only one
logical channel number (one user). Reset procedures, which apply only in the Data Transfer state, handle specific error events,
such as local procedure error, remote procedure error, network
congestion, incompatible destination, network out of order, etc.
Clear affects only one logical channel number (one user). It clears
a session when, for example, the host or host link crashes. In this
case; the network initiates Clear procedures on the terminal link.
It effectively logs off the terminal user. Clear resets the lower
window edge to O. Restart affects all users (all logical channels on
a given link) and clears all calls on that link. Restart is used when
a failed link returns to service; it makes sure that all calls were
cleared when the link went down. Restart also is used when either
side (DTEor DCE) reports error conditions at the packet level;
the good side initiates the restart procedures.
In some networks, a Diagnostic packet indicates unusual error
conditions. A Diagnostic packet from the DCE provides information on events that are considered unrecoverable at the packet
level; the information permits the DTE to analyze and initiate
recovery by higher levels. No conftrmation is required on receipt
of a Diagnostic packet.

FEBRUARY 1994

Octet

3--1

M-Bit

D-Bit Q..Bit

Bits

1 2 3 4 5 6 7 8
Packet Type Identifier

Implementation Considerations for Users
X.2S recognizes that DTEs differ in their degree of sophistication.
A simple DTE may have fixed packet formats and built-in parameter values, while a sophisticated DTE may work with varying
packet formats and provide variable parameters to take advantage
of the many network functions and signaling capabilities offered.
At the physical level, the transmission rate DTE uses to access
an X.2S network should be governed by the throughput and response time requirements of the user. Factors to consider include
the maximum number of virtual circuits operating simultaneously
and traffic characteristics of throughput-critical and responsetime-critical virtual circuits.
At the link level, LAPs and LAPBs are defined. A DTE might
implement only LAPB. Parameters that are part of the LAPB
procedures can be set to constants for all network connections.
For example, a timer (Tl) may be set according to the slowest
required line speed; a constant such as three seconds may be used.
Also, the maximum permitted number of unacknowledged
I-Frames (information frames) may be determined. The network
provider must agree to this. All networks support a window of at
least seven I-Frames. The maximum packet size (number of bits)
of the I-Frame must also be established.
If the DTE handles certain character sizes (octets), the frame
level should be capable of accommodating any number of bits
correctly (generate acknowledgment, calculate frame check sequence). How such a properly received frame is processed depends on the individual system and may include actions such as
discarding the frame, padding it, clearing the call, or resetting.

Connections Between Packet
Switched Data Networks
TSS Recommendation X.7S describes the procedures for the interconnection of packet switched networks. Many public packet
switched data networks support X.7S procedures, including
AT&T Accunet, TransCanada's Datapac, Graphnet Freedom Network II, US Sprint's SprintNet, and WangPac networks.
X.7S is similar to X.25 in that it specifies procedures for the
physical, link, and packet levels. Signal Terminal Equipment
(STE), which acts as a bridge node between networks, implements X.7S procedures. A related standard, TSS X.121, defines
the international numbering plan for public data networks.

Cl 1994 McGraw-HiII. Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

Data Networking

ITU·TSS Pecket SwRebecl
Networking Standards
X Series

Recommendation X.75
Recommendation X.75, Terminal and Transmit Call Control Procedures and Data Transfer System on International Circuits Between Packet-Switched Data Networks, provides the rules for
transmitting data between different data networks. The basic system structure is made up of communicating elements that function independently. These elements include the physical circuits,
which comprise links Al or G I (as defined in X.92), and a set of
mechanical, electrical, functional, and procedural interface characteristics; packet transfer procedures, which operate over the
physical circuits and provide for the transport of packets between
STEs; and packet signaling procedures, which use the packet data
transfer procedures and provide for the exchange of call control
information and user traffic between STEs.
Figure 6 shows the basic system structure of the signaling and
data transfer procedures. The international link is assumed to be
data link Al and/or data link G1 as defined in Recommendation
X.92. The international link should be capable of supporting fullduplex operation.
Link-level packet transfer procedures between STEs consist
of the Single Link Procedure (SLP) and the Multilink Procedure
(MLP). The SLP is used for data interchange over a single physical circuit between two STEs. When multiple physical circuits are
used in parallel, the SLP is used independently on each circuit.
The MLP is used for data interchange over multiple parallel links.
The MLP may be used over a single link when the communicating parties agree to this procedure. Transmissions are full duplex.
SLP is based upon the LAPB as described in X.25 and uses the
principle and terminology of the International Organization for
Standardization's (ISO's) HOLC. For SLP, either modulo 8 (nonextended mode) or the modulo 128 (extended mode) may be
used. The MLP is based on the multilink procedures specified by
the ISO and performs the functions of distributing packets across
available SLPs.
Three channel states are defined: the link channel state, which
provides transmission in one direction; the active channel state,
which means that the incoming or outgoing channel is receiving
or transmitting a frame; and the idle channel state, which means
that the incoming or outgoing channel is receiving or transmitting
at least 15 contiguous I bits.
Data is transmitted in frames. Each frame must contain an
Opening Flag (8 bits), an Address field (8 bits), a Control field (8
bits), a FCS field (16 bits), and a Closing Flag (8 bits). An Information field of an unspecified number of bits, which follows the
Control field, is optional.

2790
Standards

The Control field contains a command or response, and sequence numbers if applicable. Control field formats may be one
of three types: numbered information transfer (I format), numbered supervisory functions (S format), and unnumbered control
functions (U format). The I format performs information transfer
functions. The S format performs link supervisory control functions, such as acknowledging I frames, requesting transmission of
I frames, and requesting a temporary suspension of transmission
of I frames. The U format provides additional link control functions.
Each I frame is numbered sequentially. The send state variable
V(S) represents the sequence number of the next in-sequence I
frame to be transmitted. The value of the V(S) is incremented by
one with each successive I frame transmission but cannot exceed
N(R) of the last received I or S format frame by more than the
maximum number of outstanding frames. The send sequence
number N(S), contained only in I frames, is set equal to the value
of YeS). The receive state variable V(R) represents the sequence
number of the next in-sequence I frame expected to be received.
The value of V(R) is incremented by one when an error-free,
in-sequence I frame whose N(S) equals V(R) is received. Both I
frames and S frames contain the receive sequence number N(R).
N(R) is the expected send sequence number of the next received I
frame. When the STE transmits N(R), it indicates that all I frames
numbered up to and including N(R)-l have been received correctly.
All frames contain the PolVFinal (P/F) bit, which is referred to
as the P bit in command frames and the F bit in response frames.
The STE solicits a response from another STE by setting the P bit
to I; the answering STE responds by setting the F bit to I. The
following commands and responses are supported:
• Information (I) command-transfers sequentially numbered
frames that contain an information field;
• Receive ready (RR) command and response-used by the
STE to indicate that it is ready to receive an I frame or to
acknowledge a previously received I frame;
• Receive not ready (RNR) command and response-used by
the STE to indicate a busy condition;
• Reject (REJ) command and response-used by the STE to
request retransmission of I frames beginning with frame numbered N(R);
• Set asynchronous balanced mode (SABM) command-an
unnumbered command used to set the addressed STE in the

Figure 6.
Basic System Structure of X.75 Signaling and Data Transfer Procedures

r---------------------,
Higher Level Functions
-Call Control
-Network Control
-User Traffic

{

'--

Packet
Signaling
Procedures

f--

Packet
Transfer
Procedures

Signaling Terminal (STE)

I
I
I

Link
A1 or G1

XN
Interface

GatewayITransit Data Switching Exchange X or Y

1994 McGraw-Hill, Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

FEBRUARY 1994

2790
Standards

asynchronous balanced mode infonnation transfer phase; all
command/response control fields are one octet (eight bits);
• Set asynchronous balanced mode extended (SABME) command-an unnumbered command used to set the addressed
STE in the asynchronous balanced mode infonnation transfer
phase; all numbered command/response control fields are two
octets; all unnumbered command/response control fields are
one octet;
• Disconnect (DISC) command-tenninates the previously set
mode; DISC indicates that the STE transmitting DISC is suspending operation;
• Unnumbered acknowledge (UA) response-used by the STE
to acknowledge mode-setting commands;
• Disconnected mode (DM) response-reports STE status
when it is logically disconnected from the link and is in the
disconnected phase; and
• Frame reject (FRMR) response-indicates an error condition
that is not recoverable by retransmission of the frame.
Packet-level signaling procedures relate to the transfer of packets
at the STE-XlSTE-Y (XlY) interface. Recommendation X.7S
specifies signaling procedures for virtual call setup and clearing,
for pennanent virtual circuit service, for data and interrupt transfer, for flow control, and for reset.
Logical channels are used to complete simultaneous virtual
calls and/or pennanent virtual circuits. A logical channel group
number and a logical channel number (in the range 0 to IS inclusive and 0 to 255 inclusive, respectively) are assigned to each
virtual call and pennanent virtual circuit. The logical channel
group number and the logical channel number are contained in
each packet type, except restart packets.
The procedures for virtual call setup and clearing are used
only when a logical channel is in the packet-level ready (RL)
state. If call setup is possible, and no call or call attempt exists, the
logical channel is in the ready (PL) state (within the RL state).
Call setup is initiated when the STE sends a Call Request packet
across the XlY interface. The Call Request packet specifies a
logical channel in the PL state. The logical channel is then placed
in the call request state. The called STE indicates acceptance by
the called DTE by sending a Call Connected packet across the
XlY interface. It specifies the same logical channel as that requested by the Call Request packet. The logical channel is then
placed in the flow control ready (DL) state within the data transfer (P4) state.
A logical channel (in any state) can be cleared when the STE
sends a Clear Request packet, which specifies the logical channel, across the XlY interface. Upon receipt of a Clear Request
packet, STE-X or STE-Y frees the logical channel and transmits a
Clear Confirmation packet that specifies the same channel. This
places the logical channel in the ready state within the RL state.
Permanent virtual circuits require no call setup or clearing.
Data transfer procedures apply independently to each logical
channel at the XlY interface. In the data transfer (P4) state, Data,
Interrupt, Flow Control, and Reset packets may be sent and received by the STE. The data transfer state exists within the
packet-level ready state of a logical channel. Each Data packet
contains a sequence number called the packet send sequence
number peS); only Data packets contain the peS).
The procedures for flow control and reset apply only in the
data transfer state. Flow control applies to Data packets. A window is defined for each direction of transmission at the XlY interface. The lowest number in the window is called the lower
window edge. The maximum window edge does not exceed
modulo 8 or 128, is unique to each logical channel, and is reserved for a period of time. For a particular call, two window
FEBRUARY 1994

ITU·TSS P.cket SwHcheci
Networking SgnUni.

Data Networking

X .......

sizes, one for each direction of transmission, may be selected. The
packet receive sequence number peR) (modulo 8 or 128) conveys
information from the receiver for the transmission of data packets. The PeR) becomes the lower window edge when transmitted
across the XlY interface, thereby authorizing additional data
packets to cross the XlY interface.
Reset procedures are used to reinitialize a single call and apply
only in the data transfer state. The STE sends a Reset Request
packet that specifies the logical channel to indicate a request for
reset. The logical channel is placed in the reset request state. The
requested STE confinns by sending a Reset Confirmation packet,
which places the logical channel in the flow control ready state.
Restart procedures are used to clear all calls simultaneously.
When the STE sends a Restart Request packet, the XlY interface
for eacb logical channel is placed in the restart request state. In
this state all packets, except Restart Request and Restart Confirmation packets, are discarded by the XlY interface. An STE confirms by sending a Restart Confirmation packet, which places all
channels in the ready state.
Packet fonnats are based on the general structure of packets as
defined in X.2S.

Trends in Packet Switching
X.2S packet switching is widely supported in existing data processing and data communications equipment. All major host
computer and communications processor vendors, for example,
have incorporated X.2S interfaces into their products. This is part
of an overall trend in accepting international standards and the
increasing availability of products conforming to these standards.
The ITU-TSS published revisions to the X Series standards in
1984 and in 1989. Since that time, the ratification and publication
of revisions has become a continuous, ongoing process. Since the
major building blocks for X.2S were laid by 1984, all subsequent
changes have been, and will continue to be, relatively minor.
Some recent changes have revolved around efforts to make TSS
X Series standards compatible with those of the International Organization for Standardization. Currently, discussions on how to
provide greater interoperability between various X.2S networks
and between X.2S and frame-relay networks is taking place. Major developments in packet switching today, however, center not
around X.2S, but around the development of new ISDN-related
technologies, such as frame relay and Broadband ISDN, which
provide much higher throughput through simplified packetization
and routing schemes. This section discusses post-1984 changes to
X.2S and its relationship to ISDN.

Major Changes in 1988 Revisions
In the 1988 revised standards, there were no changes at the physical and link levels. At the packet level, however, a new facility for
redirecting calls, Call Deflection, was established. In 1984 the
ITU-TSS had made available a new Call Redirection facility, allowing the network to redirect all calls destined for a given address. This redirection could occur when the destination was out
of order or busy, or it could be based on time of day or other
criteria. The 1988 facility extended this capability, allowing the
destination subscriber to clear incoming calls to another party on
a call-selective basis. The Clear Request packet contains the Call
Deflection infonnation that profiles the desired alternate party.
Relationship Between lTU-TSS and ISO ElTorts
TSS X.2S packet-level protocol specifies a virtual circuit service;
the ISO has issued a compatible version of the packet standard,
ISO 8208. In recent years, ITU-TSS and ISO organizations have
worked on standards to carry longer addresses in the DTE field to
facilitate interworking with ISDN (E.164).

1994 McGraw-Hili. Incorporated. Reproduction Prohib~ed.
Dalapro Information Services Group. Dalran NJ 08075 USA

Data Networking

ITU·TSS P.cket Switched
Networking Stllndllrds
XSe,'e.

2790

Standards

In 1988 the ITU-TSS also modified the Address Extension
facilities to be consistent with ISO address length. Previously, a
provisional 32-decimaIlI6-octet field had been recommended;
this address length was increased to 40 decimalsl20 octets. The
ISO also added these address recommendations to Addendum 2
ofISO 8348 (Connection-Mode Network Service); the ITU-TSS
adoption is X.213.

With the exception of Broadband ISDN, all of the above interfaces can be carried on unloaded copper loops. Using fiber has
also been considered, as it would make the local loop more robust. Out-of-band signaling makes possible a new class of services. In addition, the 16K bps D-channel connects directly to the
BOC's packet switched network, providing the subscriber with
the data multiplexing advantages packet switching offers.

Connectionless Issues: Relationships With ISO EtTorts
At any layer of the OSI Reference Model, two basic forms of
operation are possible: connection oriented and connectionless.
Connection-oriented service involves a connection establishment
phase, a data transfer phase, and a connection termination phase.
A logical connection is set up between end entities prior to exchanging data. In a connectionless service, typical of local area
networks, each packet is independently routed to the destination.
No connection establishment activities are required since each
data unit is independent of the previous or subsequent one.
Each transmission mode has a niche where it represents the
best approach. For example, file transfers may benefit from a
connection-oriented service, while point-of-sale inquiries may be
best served by a connection less service.
Traditionally, the ITU-TSS has pursued a connection-oriented
philosophy, while ISO has addressed both connection and connectionless services. The original OSI standard, ISO 7498, is connection oriented. ISO provided connection less service, however,
by issuing an addendum to that protocol: ISO 7498/DADI. ISO
has issued a standard for connection-mode network service (ISO
8348), while the lTU-TSS has issued an identical service, X.213.
In regard to X.25 itself, however, ISO has decided not to pursue
the connectionless service, formerly known as "datagram" service.

lTU-TSS ISDN Standards
ISDN provides a specific protocol that users can employ to signal
the network. Currently, a three-layer protocol suite is defined. At
the Basic Rate, the Physical Layer manages a 192K bps, fullduplex bitstream using time-compression techniques and timedivision methods to recover the two B-channels and one D-channel. The remaining 48K bps stream is used for Physical Layer
control information. The defined standards are 1.420 (Basic Rate
Interface Definition), 1.430 (Basic Rate Interface Layer 1),1.421
(primary Rate Interface Definition), and 1.431 (Primary Rate Interface Layer 1).
The Data Link Layer is not defined for transparent B-channels
used for circuit switched voice or data, but it is defined for the
D-channel. For Narrowband ISDN, the D-channel employs a
LAP-D Link Layer protocol, which is a subset of the ISO HDLC
Data Link protocol, as specified in TSS Recommendations Q.920
(1.440) and X.921 (1.441). It provides statistical multiplexing for
three channel types: signaling information for managing the Bchannels; packet switched service over the D-channel; and optional channels, used for telemetry of other applications.
The Network Layer protocol for the signaling channel is
specified in TSS Q.930 (1.450) and Q.931 (1.451) specifications.
It provides the mechanism for establishing and terminating connections on the B-channels and other network control functions.
For the packet switched service over the D-channel, the Network
Layer protocol is X.25.
A technique is required for specifying whether user-to-network signaling, user packet data, or user telemetry data is being
sent over the D-channel. This technique involves the use of a
service access point identifier (SAPI).
Each layer in the OSI Reference Model communicates with
the layers above and below it across an interface. The interface is
through one or more service access points (SAPs). SAPs have a
number of uses, including subaddressing for intemetworking
situations, Transport Layer applications, and for user data packet
service over an ISDN D-channel.
Considering the applications to the Transport Layer, one
should note that two general types of addressing in a communications architecture are available. Each host on the network is assigned a network address, allowing the network to deliver data to
the proper computer. Each process within a host, moreover, is
given an address that is unique within the host itself, allowing the
Transport Layer to deliver data to the proper process. These process addresses are identified using SAPs. A similar approach is
followed for ISDN.
LAP-D must deal with two levels of multiplexing. First, at a
subscriber location, multiple-user devices may be sharing the
same physical interface. Second, each user device may support
multiple types of traffic, including packet switched data and signaling. To accomplish this type of mUltiplexing, LAP-D employs
a two-part address consisting of a terminal endpoint identifier
(TEl) and a SAPI. "JYpically, each user terminal is given a distinguishing TEL The SAPI identifies the traffic type and the Data
Link Layer services directed to Layer 3. For example, the SAPI
value of 0 directs the frames to Layer 3 for call-control procedures; a SAPI value of 16 indicates a packet communication procedure. See Figure 7.

Packet Switching in ISDN
The goal of the Integrated Services Digital Network (ISDN) is to
provide an end-to-end digital path over a set of standardized user
interfaces, giving the user the capability to signal the network
through an out-of-band channel. (In contrast, in X.25, the user
signals the network in-band by issuing packets such as Call Request, Call Accepted, etc.)
Currently, different types of interfaces to the telephone network exist for different services. These interfaces include twowire switched, two-wire dedicated, four-wire dedicated, and
DDS. ISDN provides a small set of interfaces that can be used for
multiple applications. The lTU-TSS has defined the following
interfaces for ISDN:
• 2B+D-two 64K bps channels and a 16K bps packet/signaling
channel (also called the Basic Rate Interface).
• 23B+D-twenty-three 64K bps channels and a 64K bps
packet/signaling channel (also called the Primary Rate Interface).
• 3HO+D-three 384K bps channels and one 64K bps packet/
signaling channel.
• HlI-nonchannelized 1.536M bps.
• Hl2-nonchannelized I.920M bps.
• Multislotted-multiples of 64K bps channels (up to I.536M
bps) under the customer's control.
• Broadband-high data rates, based on an approach called synchronous optical network (Sonet), building on multiples of
51.84M bps. Sonet standards negotiations began in 1986. The
lTU-TSS approved phase I of the standards in 1988 and phase
II in 1989. This architecture has been called Broadband ISDN,
in contrast to the other interfaces that have been considered part
of Narrowband ISDN.
Ii:> 1994 McGraw-Hili. Incorporated. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

FEBRUARY 1994

ITU~TSS Packet SwItched
Networking ShindanII.
XSeri••

figure 7." '.

...,

Data Networking

SAP~ Action/or a BtlSk Rat, JU:1um",1

Signaling Packet Telemetry

J•

.J •

J •~

Layer 4 .
________________

..r----- - -;Iir:._, - - - -. - - - - -"
r-------------,~E------------11rr:,.------------,
'
·
r
J.

Network
Layer 3
B1 Entity

Layer 3
B2 Entity

Layer 3
P Entity
ISAP1=16)

Layer 3
s Entity
ISAP1=O)

~"
-----~------------~-----------~~
1~

Layer 2
B1 Entity
64K bos

Layer 2
B2 Entity
64K bos

Layer 3
t Entity

Data Link
Layer 2
LAPD Entity
64K bps

-------------------

~-----~~:~----~
Protocols

I~_______________________________

~~l~J.

Layer 1
ISDN Physical Entity
192K bps

______________________________________________

~a1~~1

Frame Relay
Frame relay is a rapidly emerging, standards-based addressing
technique that has great potential in LANIWAN networking and
other interactive applications requiring high-throughput, low-delay transmission. Frame relay is based on the TSS Layer 2 pro~o
col developed for ISDN, Link Access Protocol D (LAPD). U?hke
conventional X.25 packet switching, frame relay uses vanable
packet lengths and performs error ch'7king only a~ the re~ote
end of transmission. Any errors occumng between mtermedlate
network nodes are assumed caught and corrected by higher-layer
protocols. Thus, intermediate nodes simply forward packets
(called frames) without processing the datastream. In addition,
frames must be received in the order in which they were sent,
unlike some X.25 networks, which involves considerably less
machine processing at the opposite end of transmission. These
efficiencies result in superior performance, with data rates up to
or above TIlEl levels.
Vendors from several different networking disciplines have
established themselves as frame-relay equipment providers.
These include traditional packet switched equipment suppliers
such as Northern Thlecom, Alcatel Data Networks, BBN Systems
and Thchnologies, BT North America, and Hughes Network Systems; T-carrier nodal processor vendors such as Ascom Ti~eplex,
General DataComm, Newbridge Networks, Network Equipment
Technologies, and StrataCom; and LAN internetworking (bridge
and router) vendors.
Most commercial frame-relay products are based on some or
all of the following American National Standards Institute
(ANSI) specifications:

• T1.602: Data Link Layer Signaling Specification for Applications at the User-Network Interface

FEBRUARY 1994

• T1.606: Frame Relaying Bearer Service-Architectural
Framework and Service Description (TIS1I88-225)
• T1 Sl.2: Addendum to TI.606 (Frame Relaying Bearer Service-Architectural Framework and Service Description) ~n
Congestion Management Principles
• T1.617: Signaling Specification for Frame Relay Bearer Service
• T1.618: Core Aspects of Frame Protocol for use with Frame
Relay Bearer Service

The ITU-TSS has released frame-relay specification 1.122, entitled "Framework for Providing Additional Packet Mode Bearer
Services. " TSS 1.462 (X.31), .. Support of Packet Mode Terminal
Equipment by an ISDN," describes a packet mode service.
Broadband ISDN and Cell Relay
Broadband ISDN (BISDN) is the blueprint for public networks in
the mid-1990s (1994 and beyond). It is being developed to support switched (on demand), se~iperman~nt, and permanent
broadband connections for both pOint-to-pomt and pomt-to-multipoint applications. Channels operating at l?5M bps ~d .622M
bps will be available under BISDN, allowmg. transmls~lon of
data, video, and digitized voice. Broadband servtc~s are aimed at
both business applications and residential subscnbers. Connections will support both circuit mode and packet mode services of
a single media, mixed-media, and multimedia.
1994 McGraw-HIII, Incorporated. Reproduction Prohibited.
Datapro InfOrmation Services Group. Detran NJ 08075 USA

\",.

Data Networking

ITU.TSS hoket Switched
NetwoIIclng StancIIIrds
x SerIes

BISON's foundation is Asynchronous Transfer Mode (ATM),
a high-bandwidth, low-delay switching and multiplexing technology in which information is packetized into fixed-size slots called
cells. A cell consists of an information field that is transported
transparently by the network and a header containing routing information. With simplified protocols and cells with a fixed, short
length (53 bytes), ATM makes very high data rates possible.
The ITU-TSS has issued several Broadband ISDN specifications in its I-Series recommendations. 1.361, the BISDN ATM
Layer Specification. defines the ATM cell structure and coding,
including header formats and coding at both the User-Network

C 1994 McGrilw-HHi. Incorponlted. Reproduction Prohibited.
Datapro Information Services Group. Delran NJ 08075 USA

2780

Standards

Interface (UNI) and Network Node Interface (NNI). It also defines ATM protocol procedures. 1.311, entitled BISDN General
Network Aspects, describes networking techniques, signaling
principles, traffic control, and resource management for BISON.
It defines ATM virtual section, virtual path, and virtual channel
concepts.
Several off-the-shelf ATM products were released in late 1993
by vendors of packet and frame-relay switches, T Carrier multiplexers, and LAN routers vendors. Many more end-user ATM
products, along with AN public carrier services, will gradually
become available in 1994 and 1995.•

FEBRUARY 1994

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-c041 52.342996, 2008/05/07-21:37:19
Create Date                     : 2018:10:11 11:40:21-08:00
Modify Date                     : 2018:10:11 12:10:55-07:00
Metadata Date                   : 2018:10:11 12:10:55-07:00
Producer                        : Adobe Acrobat 9.0 Paper Capture Plug-in
Format                          : application/pdf
Document ID                     : uuid:4b500a2a-55ae-344d-8cba-f9234457fed4
Instance ID                     : uuid:3d397653-3618-5a47-8173-5617c6847a3a
Page Layout                     : SinglePage
Page Mode                       : UseNone
Page Count                      : 22

EXIF Metadata provided by EXIF.tools

2790_ITU TSS_X.25 2790 ITU TSS X.25

Navigation menu

Versions of this User Manual:

Views

Navigation