ISDN AND BROADBAND ISDN

ISDN AND BROADBAND ISDN

Rapid advances in computer and communication technologies have resulted

in the increasing merging of these two fields. The lines have blurred among

computing, switching, and digital transmission equipment, and the same

digital techniques are being used for data, voice, and image transmission. Merging

and evolving technologies, coupled with increasing demands for efficient and timely

collection, processing, and dissemination of information, are leading to the development

of integrated systems that transmit and process all types of data. The ultimate

goal of this evolution is the integrated services digital network (ISDN).

The ISDN is intended to be a worldwide public telecommunications network

to replace existing public telecommunications networks and deliver a wide variety

of services. The ISDN is defined by the standardization of user interfaces and is

implemented as a set of digital switches and paths supporting a broad range of traffic

types and providing value-added processing services. In practice, there are multiple

networks, implemented within national boundaries, but from the user's point

of view, there will be a single, uniformly accessible, worldwide network.

The impact of ISDN on both users and vendors will be profound. To control

ISDN evolution and impact, a massive effort at standardization is underway. Although

ISDN standards are still evolving, both the technology and the emerging

implementation strategy are well understood.

Despite the fact that ISDN has yet to achieve the hoped for universal deployment,

it is already in its second generation. The first generation, sometimes referred

to as narrowband ISDN, is based on the use of a 64-kbps channel as the basic unit

of switching and has a circuit-switching orientation. The major technical contribution

of the narrowband ISDN effort has been frame relay. The second generation,

referred to as broadband ISDN (B-ISDN), supports very high data rates (100s of

Mbps) and has a packet-switching orientation. The major technical contribution of

the broadband ISDN effort has been asynchronous transfer mode (ATM), also

known as cell relay.

This appendix provides an overview of narrowband ISDN and broadband

ISDN.

OVERVIEW OF ISDN

ISDN Concept

The concept of ISDN is best introduced by considering it from several different

viewpoints:

* Principles of ISDN

* The user interface

* Objectives

* Services

Principles of ISDN

Standards for ISDN have been defined by ITU-T (formerly CCITT), a topic that we

explore later in this section. Table A.l, which is the complete text of one of the

ISDN-related standards, states the principles of ISDN from the point of view of

CCITT. Let us look at each of these points in turn:

1. Support of voice and nonvoice applications using a limited set of standardized

facilities. This principle defines both the purpose of ISDN and the means

of achieving it. The ISDN supports a variety of services related to voice communications

(telephone calls) and nonvoice communications (digital data

exchange). These services are to be provided in conformance with standards

(ITU-T recommendations) that specify a small number of interfaces and data

transmission facilities.

2. Support ,for switched and nonswitched applications. ISDN supports both circuit

switching and packet switching. In addition, ISDN supports nonswitched

services in the form of dedicated lines.

3. Reliance on 64-kbps connections. ISDN provides circuit-switched and packetswitched

connections at 64 kbps; this is the fundamental building block of

ISDN. This rate was chosen because, at the time, it was the standard rate for

digitized voice, and, hence, was being introduced into the evolving integrated

digital networks (IDNs). Although this data rate is useful, it is unfortunately

restrictive to rely solely on it. Future developments in ISDN will permit

greater flexibility.

4. Intelligence in the network. An ISDN is expected to be able to provide sophisticated

services beyond the simple setup of a circuit-switched call.

5. Layered protocol architecture. The protocols for user access to ISDN exhibit a

layered architecture and can be mapped into the OSI model. This procedure

has a number of advantages:

Standards already developed for OSI-related applications may be used

on ISDN. An example is X.25 level 3 for access to packet-switching services

in ISDN.

New ISDN-related standards can be based on existing standards, reducing

the cost of new implementations. An example is LAPD, which is

based on LAPB.

Standards can be developed and implemented independently for various

layers and functions within a layer; this allows for the gradual implementation

of ISDN services at a pace appropriate for a given provider or a

given customer base.

6. Variety of configurations. More than one physical configuration is possible for

implementing ISDN; this allows for differences in national policy (singlesource

versus competition), in the states of technology, and in the needs and

existing equipment of the customer base.

The User Interface

Figure A.l is a conceptual view of the lSDN from a user, or customer, point of view.

The user has access to the ISDN by means of a local interface to a "digital pipe" of

a certain bit rate. Pipes of various sizes are available to satisfy differing needs. For

example, a residential customer may require only sufficient capacity to handle a

telephone and a videotex terminal. An office will undoubtedly wish to connect to

the ISDN via an on-premise digital PBX, and will require a much higher capacity

pipe.

At any given point in time, the pipe to the user's premises has a fixed capacity,

but the traffic on the pipe may be a variable mix up to the capacity limit. Thus,

a user may access circuit-switched and packet-switched services, as well as other services,

in a dynamic mix of signal types and bit rates. To provide these services, the

ISDN requires rather complex control signals to instruct it how to sort out the time

multiplexed data and provide the required services. These control signals are also

multiplexed onto the same digital pipe.

An important aspect of the interface is that the user may, at any time, employ

less than the maximum capacity of the pipe, and will be charged according to the

capacity used rather than "connect time." This characteristic significantly diminishes

the value of current user design efforts that are geared to optimize circuit

utilization by use of concentrators, multiplexers, packet switches, and other linesharing

arrangements.

Objectives

Activities currently under way are leading to the development of a worldwide

ISDN. This effort involves national governments, data processing and communications

companies, standards organizations, and other agencies. Certain common

objectives are, by and large, shared by this disparate group. We list here the key

objectives:

Standardization. It is essential that a single set of ISDN standards be provided

to permit universal access and to permit the development of cost-effective

equipment.

Transparency. The most important service to be provided is a transparent

transmission service, thereby permitting users to develop applications and

protocols with the confidence that they will not be affected by the underlying

ISDN.

Separation of competitive functions. It must be possible to separate out functions

that could be provided competitively as opposed to those that are fun

damentally part of the ISDN. In many countries, a single, government-owned

entity provides all services. Some countries desire (in the case of the United

States, require) that certain enhanced services be offered competitively (e.g.,

videotex, electronic mail).

@ Leased and switched services. The ISDN should provide dedicated point-topoint

services as well as switched services, thereby allowing the user to optimize

implementation of switching and routing techniques.

@ Cost-related tariffs. The price for ISDN service should be related to cost, and

should be independent of the type of data being carried. One type of service

should not be in the position of subsidizing others.

Smooth migration. The conversion to ISDN will be gradual, and the evolving

network must coexist with existing equipment and services. Thus, ISDN interfaces

should evolve from current interfaces, and provide a migration path

for users.

@ Multiplexed support. In addition to providing low-capacity support to individual

users, multiplexed support must be provided to accommodate userowned

PBX and local network equipment.

There are, of course, other objectives that could be named. Those listed above

are certainly among the most important and widely accepted, and each helps to

define the character of the ISDN.

Architecture

Figure A.2 is a block diagram of ISDN. ISDN supports a new physical connecter for

users, a digital subscriber loop (link from end user to central or end office), and

modifications to all central office equipment.

The area to which most attention has been paid by standards organizations is

that of user access. A common physical interface has been defined to provide, in

essence, a DTE-DCE connection. The same interface should be usable for telephone,

computer terminal, and videotex terminal. Protocols are needed for the

exchange of control information between user device and the network. Provision

must be made for high-speed interfaces to, for example, a digital PBX or a LAN.

The subscriber loop portion of today's telephone network consists of twisted

pair links between the subscriber and the central office, carrying 4-kHz analog signals.

Under the ISDN, one or two twisted pairs are used to provide a basic fullduplex

digital communications link.

The digital central office connects the numerous ISDN subscriber loop signals

to the IDN. In addition to providing access to the circuit-switched network, the central

office provides subscriber access to dedicated lines, packet-switched networks,

and time-shared, transaction-oriented computer services. Multiplexed access via

digital PBX and LAN must also be accommodated.

Standards

The development of ISDN is governed by a set of recommendations issued by

ISDN, called the I-series Recommendations. These Recommendations, or stan

dards, were first issued in 1984. A more complete set was issued in 1988. Most of the

Recommendations have been updated, at irregular intervals, since that time. The

bulk of the description of ISDN is contained in the I-series Recommendations, with

some related topics covered in other Recommendations. The characterization of

ISDN contained in these Recommendations is centered on three main areas:

1. The standardization of services offered to users, so as to enable services to be

internationally compatible.

2. The standardization of user-network interfaces, so as to enable terminal

equipment to be portable, and to assist in (1).

3. The standardization of ISDN capabilities to the degree necessary to allow

user-network and network-network interworking, and thus achieve (1)

and (2).

The I-series Recommendations are broken up into six main groupings, labeled

1.100 through 1.600.

1.100 Series-General Concepts

The 1.100 series serves as a general introduction to ISDN. The general structure of

the ISDN recommendations is presented as well as a glossary of terms. 1.120 provides

an overall description of ISDN and the expected evolution of ISDNs. 1.130

introduces terminology and concepts that are used in the 1.200 series to specify

services.

1.200 Series-Service Capabilities

The 1.200 series is in a sense the most important part of the ITU-T ISDN recommendations.

Here, the services to be provided to users are specified. We may look

on this as a set of requirements that the ISDN must satisfy. In the ISDN glossary ,

(1.112), the term service is defined as

That which is offered by an Administration or recognized private operating

agency (RPOA) to its customers in order to satisfy a specific telecommunication

requirement.

Although this is a very general definition, the term "service" has come to have a very

specific meaning in ITU-T, a meaning that is somewhat different from the use of that

term in an OSI context. For ITU-T, a standardized service is characterized by

Complete, guaranteed end-to-end compatibility

ITU-T-standardized terminals, including procedures

Listing of the service subscribers in an international directory

ITU-T-standardized testing and maintenance procedures

Charging and accounting rules

There are three fully standardized ITU-T services: telegraphy, telephony, and

data. There are four additional telematic services in the process of being standardized:

teletex, facsimile, videotex, and message handling. The goal with all of these

services is to ensure high-quality international telecommunications for the end user,

regardless of the make of the terminal equipment and the type of network used

nationally to support the service.

1.300 Series-Network Aspects

Whereas the 1.200 series focuses on the user, in terms of the services provided, the

1.300 series focuses on the network, in terms of how the network goes about providing

those services. A protocol reference model is presented that, while based on

the 7-layer OSI model, attempts to account for the complexity of a connection that

may involve two or more users (e.g., a conference call) plus a related commonchannel

signaling dialogue. Issues such as numbering and addressing are covered.

There is also a discussion of ISDN connection types.

1.400 Series-User-Network Interfaces

The 1.400 series deals with the interface between the user and the network. Three

major topics are addressed:

Physical configurations. The issue of how ISDN functions are configured into

equipment. The standards specify functional groupings and define reference

points between those groupings.

Transmission rates. The data rates and combinations of data rates to be

offered to the user.

Protocol specifications. The protocols at OSI layers 1 through 3 that specify

the user-network interaction.

1.500 Series-Internetwork Interfaces

ISDN supports services that are also provided on older circuit-switched and packetswitched

networks. Thus, it is necessary to provide interworking between an ISDN

and other types of networks to allow communications between terminals belonging

to equivalent services offered through different networks. The 1.500 series deals

with the various network issues that arise in attempting to define interfaces between

ISDN and other types of networks.

1.600 Series-Maintenance Principles

This series provides guidance for maintenance of the ISDN subscriber installation,

the network portion of the ISDN basic access, primary access, and higher data-rate

services. Maintenance principles and functions are related to the reference configuration

and general architecture of ISDN. A key function that is identified in the

series is loopback. In general, loopback testing is used for failure localization and

verification.

A.2 ISDN CHANNELS

The digital pipe between the central office and the ISDN user is used to carry a

number of communication channels. The capacity of the pipe, and therefore the

number of channels carried, may vary from user to user. The transmission structure

of any access link is constructed from the following types of channels:

B channel: 64 kbps

D channel: 16 or 64 kbps

H channel: 384(H0), 1536(H11), and 1920 (H12) kbps

The B channel is the basic user channel. It can be used to carry digital data,

PCM-encoded digital voice, or a mixture of lower-rate traffic, including digital data

and digitized voice encoded at a fraction of 64 kbps. In the case of mixed traffic, all

traffic must be destined for the same endpoint. Four kinds of connections can be set

up over a B channel:

Circuit-switched. This is equivalent to switched digital service available today.

The user places a call, and a circuit-switched connection is established with

another network user. An interesting feature is that call-establishment dialogue

does not take place over the B channel, but is done over the D, as

explained below.

Packet-switched. The user is connected to a packet-switching node, and data

are exchanged with other users via X.25.

Frame mode. The user is connected to a frame relay node, and data are

exchanged with other users via LAPF.

Semipermanent. This is a connection to another user set up by prior arrangement,

and not requiring a call-establishment protocol; this is equivalent to a

leased line.

The designation of 64 kbps as the standard user channel rate highlights the

fundamental contradiction in standards activities. This rate was chosen as the most

effective for digitized voice, yet the technology has progressed to the point at which

32 kbps, or even less, produces equally satisfactory voice reproduction. To be effective,

a standard must freeze the technology at some defined point. Yet by the time

the standard is approved, it may already be obsolete.

The D channel serves two purposes. First, it carries signaling information to

control circuit-switched calls on associated B channels at the user interface. In addition,

the D channel may be used for packet-switching or low-speed (e.g., 100 bps)

telemetry at times when no signaling information is waiting. Table A.2 summarizes

the types of data traffic to be supported on B and D channels.

H channels are provided for user information at higher bit rates. The user may

employ such a channel as a high-speed trunk, or the channel may be subdivided

according to the user's own TDM scheme. Examples of applications include fast

facsimile, video, high-speed data, high-quality audio, and multiple information

streams at lower data rates.

These channel types are grouped into transmission structures that are offered

as a package to the user. The best-defined structures at this time are the basic channel

structure (basic access) and the primary channel structure (primary access),

which are depicted in Figure A.3.

Basic access consists of two full-duplex 64-kbps B channels and a full-duplex

16-kbps D channel. The total bit rate, by simple arithmetic, is 144 kbps. However,

framing, synchronization, and other overhead bits bring the total bit rate on a basic

access link to 192 kbps. The frame structure for basic access was shown in Figure

7.10. Each frame of 48 bits includes 16 bits from each of the B channels and 4 bits

from the D channel.

The basic service is intended to meet the needs of most individual users,

including residential and very small offices. It allows the simultaneous use of voice

and several data applications, such as packet-switched access, a link to a central

alarm service, facsimile, videotex, and so on. These services could be accessed

through a single multifunction terminal or several separate terminals. In either case,

a single physical interface is provided. Most existing two-wire local loops can support

this interface.

In some cases, one or both of the B channels remain unused; this results in a

B+D or D interface, rather than the 2B+D interface. However, to simplify the network

implementation, the data rate at the interface remains at 192 kbps. Nevertheless,

for those subscribers with more modest transmission requirements, there may

be a cost savings in using a reduced basic interface.

Primary access is intended for users with greater capacity requirements, such

as offices with a digital PBX or a local network. Because of differences in the digital

transmission hierarchies used in different countries, it was not possible to get

agreement on a single data rate. The United States, Canada, and Japan make use of

a transmission structure based on 1.544 Mbps; this corresponds to the T1 transmission

facility using the DS-1 transmission format. In Europe, 2.048 Mbps is the standard

rate. Both of these data rates are provided as a primary interface service. Typically,

the channel structure for the 1.544-Mbps rate is 23 B channels plus one

64-kbps D channel and, for the 2.048-Mbps rate, 30 B channels plus one 64-kbps D

channel. Again, it is possible for a customer with lower requirements to employ

fewer B channels, in which case the channel structure is nB+D, where n ranges

from 1 to 23, or 1 to 30 for the two primary services. Also, a customer with high

data-rate demands may be provided with more than one primary physical interface.

In this case, a single D channel on one of the interfaces may suffice for all signaling

needs, and the other interfaces may consist solely of B channels (24B or 31B). The

frame structure for primary access was shown in Figure 7.11.

The primary interface may also be used to support H channels. Some of these

structures include a 64-kbps D channel for control signaling. When no D channel is

present, it is assumed that a D channel on another primary interface at the same

subscriber location will provide any required signaling. The following structures are

recognized:

Primary rate interface HO channel structures. This interface supports multiple

384-kbps HO channels. The structures are 3H0 + D and 4H0 for the

1.544-Mbps interface, and 5H0 + D for the 2.048-Mbps interface.

Primary rate interface H1 channel structures. The HI1 channel structure consists

of one 1536-kbps HI1 channel. The H12 channel structure consists of one

1920-kbps H12 channel and one D channel.

Primary rate interface structures for mixtures of B and HO channels. This

interface consists of 0 or 1 D channels plus any possible combination of B and

HO channels, up to the capacity of the physical interface (e.g., 3H0 + 5B + D

and 3H0 + 6B).

A.3 USER ACCESS

To define the requirements for ISDN user access, an understanding of the anticipated

configuration of user premises equipment and of the necessary standard

interfaces is critical. The first step is to group functions that may exist on the user's

premises. Figure A.4 shows the CCITT approach to this task, using

Functional groupings. Certain finite arrangements of physical equipment or

combinations of equipment.

Reference points. Conceptual points used to separate groups of functions.

The architecture on the subscriber's premises is broken up functionally into

groupings separated by reference points. This separation permits interface standards

to be developed at each reference point; this effectively organizes the standards

work and provides guidance to the equipment providers. Once stable interface

standards exist, technical improvements on either side of an interface can be

made without impacting adjacent functional groupings. Finally, with stable interfaces,

the subscriber is free to procure equipment from different suppliers for the

various functional groupings, so long as the equipment conforms to the relevant

interface standards.

Network termination I (NT1) includes functions associated with the physical

and electrical termination of the ISDN on the user's premises; these correspond to

OSI layer 1. The NT1 may be controlled by the ISDN provider and forms a boundary

to the network. This boundary isolates the user from the transmission technology

of the subscriber loop and presents a physical connector interface for user

device attachment. In addition, the NT1 performs line maintenance functions such

as loopback testing and performance monitoring. The NT1 supports multiple channels

(e.g., 2B + D); at the physical level, the bit streams of these channels are multiplexed

together, using synchronous time-division multiplexing. Finally, the NT1

interface might support multiple devices in a multidrop arrangement. For example,

a residential interface might include a telephone, personal computer, and alarm system,

all attached to a single NT1 interface via a multidrop line.

Network termination 2 (NT2) is an intelligent device that can perform switching

and concentration functions; it may include functionality up through layer 3 of

the OSI model. Examples of NT2 are a digital PBX, a terminal controller, and a

LAN. An example of a switching function is the construction of a private network

using semipermanent circuits among a number of sites, each of which could include

a PBX that acts as a circuit switch, or a host computer that acts as a packet switch.

The concentration function simply means that multiple devices, attached to a digital

PBX, LAN, or terminal controller, may transmit data across an ISDN.

Network termination 1, 2 (NT12) is a single piece of equipment that contains

the combined functions of NT1 and NT2; this points out one of the regulatory issues

associated with ISDN interface development. In many countries, the ISDN

provider will own the NT12 and provide full service to the user. In the United

States, there is a need for a network termination with a limited number of functions

to permit competitive provision of user premises equipment. Hence, the user

premises network functions are split into NT1 and NT2.

Terminal equipment refers to subscriber equipment that makes use of ISDN;

two types are defined. Terminal equipment type I (TE1) refers to devices that support

the standard ISDN interface. Examples are digital telephones, integrated

voiceldata terminals, and digital facsimile equipment. Terminal equipment type 2

(TE2) encompasses existing non-ISDN equipment. Examples are terminals with a

physical interface, such as EIA-232-E, and host computers with an X.25 interface.

Such equipment requires a terminal adapter (TA) to plug into an ISDN interface.

The definitions of the functional groupings also define, by implication, the reference

points. Reference point T (terminal) corresponds to a minimal ISDN network

termination at the customer's premises; it separates the network provider's

equipment from the user's equipment. Reference point S (system) corresponds to

the interface of individual ISDN terminals and separates user terminal equipment

from network-related communications functions. Reference point R (rate) provides

a non-ISDN interface between user equipment that is not ISDN-compatible and

adapter equipment. Typically, this interface will comply with an older interface

standard, such as EIA-232-E.

ISDN PROTOCOL

ISDN ARCHITECTURE

Figure A.5 illustrates, in the context of the OSI model, the protocols defined or referenced

in the ISDN documents. As a network, ISDN is essentially unconcerned

with user layers 4-7. These are end-to-end layers employed by the user for the

exchange of information. Network access is concerned only with layers 1-3. Layer

1, defined in 1.430 and 1.431, specifies the physical interface for both basic and primary

access. Because B and D channels are multiplexed over the same physical

interface, these standards apply to both types of channels. Above this layer, the protocol

structure differs for the two channels.

For the D channel, a new data link layer standard, LAPD (Link Access Protocol,

D channel) has been defined. This standard is based on HDLC, modified to

meet ISDN requirements. All transmission on the D channel is in the form of

LAPD frames that are exchanged between the subscriber equipment and an ISDN

switching element. Three applications are supported: control signaling, packetswitching,

and telemetry. For control signaling, a call control protocol has been

defined (1.451lQ.931). This protocol is used to establish, maintain, and terminate

connections on B channels; thus, it is a protocol between the user and the network.

Above layer 3, there is the possibility for higher-layer functions associated with

user-to-user control signaling. These functions are a subject for further study. The

D channel can also be used to provide packet-switching services to the subscriber.

In this case, the X.25 level-3 protocol is used, and X.25 packets are transmitted in

LAPD frames. The X.25 level-3 protocol is used to establish virtual circuits on the

D channel to other users and to exchange packetized data. The final application

area, telemetry, is a subject for further study.

The B channel can be used for circuit switching, semipermanent circuits, and

packet-switching. For circuit switching, a circuit is set up on a B channel, on demand.

The D-channel call control protocol is used for this purpose. Once the circuit is set

up, it may be used for data transfer between the users. A semipermanent circuit is a

B-channel circuit that is set up by prior agreement between the connected users and

the network. As with a circuit-switched connection, it provides a transparent data

path between end systems.

With either a circuit-switched connection or a semipermanent circuit, it

appears to the connected stations that they have a direct, full-duplex link with each

other. They are free to use their own formats, protocols, and frame synchronization.

Hence, from the point of view of ISDN, layers 2 through 7 are not visible or

specified.

In the case of packet-switching, a circuit-switched connection is set up on a B

channel between the user and a packet-switched node using the D-channel control

protocol. Once the circuit is set up on the B channel, the user may employ X.25 layers

2 and 3 to establish a virtual circuit to another user over that channel and to

exchange packetized data. As an alternative, the frame relay service may be used.

Frame relay can also be used over H channels and over the D channel.

Some of the protocols shown in Figure AS are summarized in the remainder

of this section. First, we look at the way in which packet-switched and circuitswitched

connections are set up. Next, we examine the control-signaling protocol

and then LAPD. Finally, the physical-layer specifications are reviewed.

ISDN Connections

ISDN provides four types of service for end-to-end communication:

9 Circuit-switched calls over a B channel.

Semipermanent connections over a B channel.

Packet-switched calls over a B channel.

0 Packet-switched calls over the D channel.

Circuit-Switched Calls

The network configuration and protocols for circuit switching involve both the B

and D channels. The B channel is used for the transparent exchange of user data.

The communicating users may employ any protocols they wish for end-to-end communication.

The D channel is used to exchange control information between the

user and the network for call establishment and termination, as well as to gain

access to network facilities.

The B channel is serviced by an NT1 or NT2 using only layer-1 functions. On

the D channel, a three-layer network access protocol is used and is explained below.

Finally, the process of establishing a circuit through ISDN involves the cooperation

of switches internal to ISDN to set up the connection. These switches interact by

using an internal protocol: Signaling-System Number 7.

Semipermanent Connections

A semipermanent connection between agreed points may be provided for an indefinite

period of time after subscription, for a fixed period, or for agreed-upon periods

during a day, a week, or some other interval. As with circuit-switched connections,

only Layer-1 functionality is provided by the network interface. The

call-control protocol is not needed because the connection already exists.

Packet-Switched Calls over a B Channel

The ISDN must also permit user access to packet-switched services for data traffic

(e.g., interactive) that is best serviced by packet switching. There are two possibilities

for implementing this service: Either the packet-switching capability is furnished

by a separate network, referred to as a packet-switched public data network

(PSPDN), or the packet-switching capability is integrated into ISDN. In the former

case, the service is provided over a B channel. In the latter case, the service may be

provided over a B or D channel. We first examine the use of a B channel for packetswitching.

When the packet-switching service is provided by a separate PSPDN, the

access to that service is via a B channel. Both the user and the PSPDN must therefore

be connected as subscribers to the ISDN. In the case of the PSPDN, one or

more of the packet-switching network nodes, referred to as packet handlers, are

connected to ISDN. We can think of each such node as a traditional X.25 DCE supplemented

by the logic needed to access ISDN. That is, the ISDN subscriber

assumes the role of an X.25 DTE, the node in the PSPDN to which it is connected

functions as an X.25 DCE, and the ISDN simply provides the connection from DTE

to DCE. Any ISDN subscriber can then communicate, via X.25, with any user connected

to the PSPDN, including

* Users with a direct, permanent connection to the PSPDN.

* Users of the ISDN that currently enjoy a connection, through the ISDN, to

the PSPDN.

The connection between the user (via a B channel) and the packet handler

with which it communicates may be either semipermanent or circuit-switched. In

the former case, the connection is always there and the user may freely invoke X.25

to set up a virtual circuit to another user. In the latter case, the D channel is

involved, and the following sequence of steps occurs (Figure A.6):

1. The user requests, via the D-channel call-control protocol (1.451/Q.931), a circuit-

switched connection on a B channel to a packet handler.

2. The connection is set up by ISDN, and the user is notified via the D channel

call-control protocol.

3. The user sets up a virtual circuit to another user via the X.25 call establishment

procedure on the B channel (described in Section 3.2). This step requires

first that a data link connection, using LAPB, must be set up between the user

and the packet handler.

4. The user terminates the virtual circuit, using X.25 on the B channel.

5. After one or more virtual calls on the B channel, the user is done and signals

via the D channel to terminate the circuit-switched connection to the packetswitching

node.

6. The connection is terminated by ISDN.

Figure A.7 shows the configuration involved in providing this service. In the

figure, the user is shown to employ a DTE device that expects an interface to an

X.25 DCE. Hence, a terminal adapter is required. Alternatively, the X.25 capability

can be an integrated function of an ISDN TE1 device, dispensing with the need

for a separate TA.

When the packet-switching service is provided by ISDN, the packet-handling

function is provided within the ISDN, either by separate equipment or as part of the

exchange equipment. The user may connect to a packet handler either by a B channel

or the D channel. On a B channel, the connection to the packet handler may be

either switched or semipermanent, and the same procedures described above apply

for switched connections. In this case, rather than establish a B-channel connection

to another ISDN subscriber that is a PSPDN packet handler, the connection is to an

internal element of ISDN that is a packet handler.

Packet-Switched Calls over a D Channel

When the packet-switching service is provided internal to the ISDN, it can also be

accessed on the D channel. For this access, ISDN provides a semipermanent connection

to a packet-switching node within the ISDN. The user employs the X.25

level-3 protocol, as is done in the case of a B-channel virtual call. Here, the level-3

protocol is carried by LAPD frames. Because the D channel is also used for control

signaling, some means is needed to distinguish between X.25 packet traffic and

ISDN control traffic; this is accomplished by means of the link-layer addressing

scheme, as explained below.

Figure A.8 shows the configuration for providing packet-switching within

ISDN. The packet-switching service provided internal to the ISDN over the B and

D channels is logically provided by a single packet-switching network. Thus, virtual

calls can be set up between two D-channel users, two B-channel users, and between

a B and D channel user. In addition, it will be typical to also provide access to X.25

users on other ISDNs and PSPDNs by appropriate interworking procedures.

Common Channel Signaling at the ISDN User Network Interface

ITU-T has developed a standard, 1.451, for common channel signaling. The primary

application of this standard is for the Integrated Services Digital Network. In OSI

terms, 1.451 is a layer-3, or network-layer, protocol. As Figure A.9 indicates, I.451

relies on a link layer protocol to transmit messages over the D channel. 1.451 specifies

procedures for establishing connections on the B channels that share the same

physical interface to ISDN as the D channel. It also provides user-to-user control

signaling over the D channel.

The process of establishing, controlling, and terminating a call occurs as a

result of control-signaling messages exchanged between the user and the network

over a D channel. A common format is used for all messages defined in 1.451, illustrated

in Figure A.lOa. Three fields are common to all messages:

Protocol discriminator. Used to distinguish messages for user-network call

control from other message types. Other sorts of protocols may share the

common signaling channel.

* Call reference. Identifies the user-channel call to which this message refers.

As with X.25 virtual circuit numbers, it has only local significance. The call

reference field comprises three subfields. The length subfield specifies the

length of the remainder of the field in octets. This length is one octet for a

basic-rate interface, and two octets for a primary-rate interface. The flag indicates

which end of the LAPD logical connection initiated the call.

Message type. Identifies which 1.451 message is being sent. The contents of

the remainder of the message depend on the message type.

Following these three common fields, the remainder of the message consists

of a sequence of zero or more information elements, or parameters. These contain

additional information to be conveyed with the message. Thus, the message type

specifies a command or response, and the details are provided by the information

elements. Some information elements must always be included with a given message

(mandatory), and others are optional (additional). Three formats for information

elements are used, as indicated in Figure A.lOb through d.

Table A.3 lists the 1.451 messages. The messages can be grouped along two

dimensions. Messages apply to one of four applications: circuit-mode control,

packet-mode access connection control, user-to-user signaling not associated with

circuit-switched calls, and messages used with the global call reference. In addition,

messages perform functions in one of four categories: call establishment, call information,

call clearing, and miscellaneous.

Circnit-mode control refers to the functions needed to set up, maintain, and

clear a circuit-switched connection on a user channel. This function corresponds

to call control in existing circuit-switching telecommunications networks. Packetmode

access conrzection control refers to the functions needed to set up a circuitswitched

connection (called an access connection in this context) to an ISDN

packet-switching node; this connects the user to the packet-switching network provided

by the ISDN provider. User-to-user signaling messages allow two users to

communicate without setting up a circuit-switched connection. A temporary signaling

connection is established and cleared in a manner similar to the control of a

circuit-switched connection. Signaling takes place over the signaling channel and

thus does not consume user-channel resources. Finally, global call reference refers

to the functions that enable either user or network to return one or more channels

to an idle condition.

Call establishment messages are used to initially set up a call. This group

includes messages between the calling terminal and the network, and between the

network and the called terminal. These messages support the following services:

* Set up a user-channel call in response to user request.

Provide particular network facilities for this call.

Inform calling user of the progress of the call establishment process.

Once a call has been set up, but prior to the disestablishment (termination)

phase, call-information phase messages are sent between user and network. One of

the messages in this group allows the network to relay, without modification, information

between the two users of the call. The nature of this information is beyond

the scope of the standard, but it is assumed that it is control signaling information

that can't or should not be sent directly over the user-channel circuit. The remainder

of the messages allow users to request both the suspension and later resumption

of a call. When a call is suspended, the network remembers the identity of the called

parties and the network facilities supporting the call, but it deactivates the call so

that no additional charges are incurred and so that the corresponding user channel

is freed up. Presumably, the resumption of a call is quicker and cheaper than the

origination of a new call.

Call-clearing messages are sent between user and network in order to terminate

a call. Finally, there are some miscellaneozu messages that may be sent between

user and network at various stages of the call. Some may be sent during call setup;

others may be sent even though no calls exist. The primary function of these messages

is to negotiate network features (supplementary services).

LAPD

All traffic over the D channel employs a link-layer protocol known as LAPD (Link

Access Protocol-D Channel).

LAPD Services

The LAPD standard provides two forms of service to LAPD users: the unacknowledged

information-transfer service and the acknowledged information-transfer service.

The unacknowledged in,formation-transfer service simply provides for the

transfer of frames containing user data with no acknowledgment. The service does

not guarantee that data presented by one user will be delivered to another user, nor

does it inform the sender if the delivery attempt fails. The service does not provide

any flow control or error-control mechanism. This service supports both point-topoint

(deliver to one user) or broadcast (deliver to a number of users); it allows for

fast data transfer and is useful for management procedures such as alarm messages

and messages that need to be broadcast to multiple users.

The acknowledged in,formation-trans,fer is the more common service, and is

similar to that offered by LAP-B and HDLC. With this service, a logical connection

is established between two LAPD users prior to the exchange of data.

LAPD Protocol

The LAPD protocol is based on HDLC. Both user information and protocolcontrol

information and parameters are transmitted in frames. Corresponding to

the two types of service offered by LAPD, there are two types of operation:

Unacknowledged operation. Layer-3 information is transferred in unnumbered

frames. Error detection is used to discard damaged frames, but there is

no error control or flow control.

Acknowledged operation. Layer-3 information is transferred in frames that

include acknowledged sequence numbers. Error control and flow control procedures

are included in the protocol. This type is also referred to in the standard

as multiple-frame operation.

These two types of operation may coexist on a single D channel, and both make use

of the frame format illustrated in Figure A.ll. This format is identical to that of

HDLC (Figure 6.10), with the exception of the address field.

To explain the address field, we need to consider that LAPD has to deal with

two levels of multiplexing. First, at the subscriber site, there may be multiple user

devices sharing the same physical interface. Second, within each user device, there

may be multiple types of traffic: specifically. packet-switched data and control signaling.

To accommodate these levels of multiplexing, LAPD employs a two-part

 

address, consisting of a terminal endpoint identifier (TEI) and a service access point

identifier (SAPI).

Typically, each user device is given a unique terminal endpoint identifier

(TEI). It is also possible for a single device to be assigned more than one TEI; this

might be the case for a terminal concentrator. TEI assignment occurs either automatically

when the equipment first connects to the interface, or manually by the

user. In the latter case, care must be taken that multiple equipment attached to the

same interface do not'have the same TEI. The advantage of the automatic procedure

is that it allows the user to change, add, or delete equipment at will without

prior notification to the network administration. Without this feature, the network

would be obliged to manage a data base for each subscriber that would need to be

updated manually. Table A.4a shows the assignment of TEI numbers.

The service access point identifier (SAPI) identifies a layer-3 user of LAPD,

and thus corresponds to a layer3 protocol entity within a user device. Four specific

values have been assigned, as shown in Table A.4b. A SAPI of 0 is used for callcontrol

procedures for managing B-channel circuits; the value 16 is reserved for

packet-mode communication on the D channel using X.25 level 3; and a value of

63 is used for the exchange of layer-2 management information. Finally, values in

the range 32 to 62 are reserved to support frame-relay connections.

For acknowledged operation, LAPD follows essentially the same procedures

described for HDLC in Lesson 6. For unacknowledged operation, the user information

(UI) frame is used to transmit user data. When a LAPD user wishes to send

data, it passes the data to its LAPD entity, which passes the data in the information

field of a UI frame. When this frame is received, the information field is passed up

to the destination user. There is no acknowledgment returned to the other side.

However, error detection is performed, and frames in error are discarded.

Physical Layer

The ISDN physical layer is presented to the user at either reference point S or T

(Figure A.4). The mechanical interface was described in Lesson 5.

The electrical specification depends on the specific interface. For the basicaccess

interface, pseudoternary coding is used (Figure 4.2). Recall that with

pseudoternary, the line signal may take one of three levels; this is not as efficient as

a two-level code, but it is reasonably simple and inexpensive. At the relatively modest

data rate of the basic-access interface, this is a suitable code.

For the higher-speed, primary-access interface, a more efficient coding

scheme is needed. For the 1.544-Mbps data rate, the B8ZS code is used, while for

the 2.048-Mbps data rate, the HDB3 code is used (Figure 4.6). There is no particular

advantage of one over the other; the specification reflects historical usage.

The functional specification for the physical layer includes the following

functions:

Full-duplex transmission of B-channel data

Full-duplex transmission of D-channel data

Full-duplex transmission of timing signals

Activation and deactivation of physical circuit

Power feeding from the network termination to the terminal

Terminal identification

Faulty-terminal isolation

D-channel-contention access

The final function is required when multiple TE1 terminals share a single

physical interface (i.e., a multipoint line). In that case, no additional functionality is

needed to control access in the B channels, as each channel is dedicated to a particular

circuit at any given time. However, the D channel is available for use by all the

devices for both control signaling and for packet transmission. For incoming data,

the LAPD addressing scheme is sufficient to sort out the proper destination for

each data unit. For outgoing data, some sort of contention resolution protocol

is needed to assure that only one device at a time attempts to transmit.