Chapter 15 The Integrated Services Digital Network

The PSTN has taken shape over more than a century of progressive technical advancements. Before the advent of reliable computers, PSTN features were constrained by the limits of relays and wired logic that were complex, inflexible, and expensive to maintain. The first departure from electromechanical switches was Western Electric’s No. 1 ESS, the first of which was installed in 1965. That system had to meet a rigid set of requirements of interoperability with existing systems and the need to preserve investment in outside plant, station equipment, and interoffice trunking. The No. 1 ESS was crude by today’s standards, but it launched a

transition into the intelligent network.

Except for agreements on the use of electromagnetic spectrum, international standards were weak and ineffective in the United States in the mid-20th century. Network transmission and data protocol standards were, to a large degree, dictated by AT&T and IBM, respectively, resulting from their market power. Both companies worked with international bodies to devise mutual standards, but the results were often incompatible. The OSI model is derived from IBM’s SNA, but SNA is a working network architecture, while OSI is not. AT&T, unwilling to wait for international bodies to agree on digital carrier and common channel signaling standards, moved forward with their own standards. SS7 has since replaced AT&T’s common channel interoffice signaling, but the digital hierarchies remained incompatible until SONET/SDH standards brought them together at higher levels.

 In this context, the Consultative Committee on International Telephone and Telegraph (CCITT), which has since evolved into ITU-T, began working on modernization of the PSTN in the 1970s, but the original ISDN guidelines were not produced until 1984 in CCITT recommendation I.120, and this was far from complete. Computers at the time were primitive and costly and 64 Kbps of bandwidth was more than enough for almost any application that was then foreseen. One possible exception was large file transfers, but at that time a 100 MB disk was considered huge. By the time the recommendations matured, data requirements had changed so dramatically that ISDN was largely irrelevant for data.

ISDN was developed to resolve what was then perceived to be deficiencies in the PSTN that hampered both voice and data transmission. One was the need for multiple interfaces depending on the type of service. Service delivery to customer premises was on copper cable pairs and different interfaces were needed for various types of service. ISDN was the only technology at the time that could offer end-to-end digital connectivity, which data needed.

 In the vision of the time, ISDN would breach the transmission speed limits

of the PSTN for connectivity to value-added networks, which were then accessed over low-speed modems. It would provide flexibility for developers to bring new products to market and it promised interoperability between products of different vendors so that any ISDN telephone could operate on any vendor’s switch. For voice users, a separate signaling channel would give the users insights into the nature of incoming calls, even when the line was active. It would enable distinctive treatment of the call, depending on who was calling.

ISDN might have fulfilled its potential if it had reached the market earlier and had been priced attractively and promoted actively, but none of these happened in North America. The standards were slow to develop and when they did, products were not interoperable. Some IXCs used proprietary ISDN protocols to deliver services their customers were requesting. For example, calling number delivery was a desirable service to many call centers. Some IXCs delivered caller ID in-band and others used a proprietary ISDN. Wags joked that the acronym stood for “innovations subscribers don’t need,” or “I still don’t know.”

ISDN requires digital switching. Many of the LECs’ No. 1 ESS switches had to be replaced to offer the service and the LECs did not perceive that the demand existed. For existing digital switches, some LECs were slow to invest in the upgrade to ISDN until they were certain that the demand existed. Many LECs charged such a premium for the service that customers had no incentive to subscribe except for dial-up video conferencing. Perhaps the crowning blow to basicrate ISDN, which is aimed at residences and small businesses, is the availability of  DSL and cable Internet access. In comparison to the speeds available with these services, ISDN is too slow and too costly.

 Primary-rate ISDN is a viable service offering. Where it is priced competitively with analog or non-ISDN digital PBX trunks, PRI offers many advantages and is usually chosen over other alternatives. Trunks can be used interchangeably between incoming, outgoing, and DID, which increases trunk efficiency. The ability to place end-to-end digital calls anywhere in the world is far from assured, however, because of analog local loops.

Several key applications generate demand for ISDN. For business subscribers, ISDN is an effective medium for videoconferencing. ISDN lines are effective for backing up private networks such as frame relay. Where the LEC tariffs are reasonable, ISDN trunking offers key advantages for PBX owners. Another major trend is telecommuting. As more incentives develop for workers to spend at least part of the week working from home, ISDN can be an economical way of retaining a link to the office at reasonable speed, although here also it is losing the battle to IP.

ISDN TECHNOLOGY

ISDN lines come in two varieties, BRI and PRI. BRI consists of two 64 Kbps B (bearer) channels plus one 16 Kbps D (data) channel. In ISDN shorthand the interface is called 2B+D. The D channel carries signaling between the central office switch and terminating equipment, which could be a telephone set, personal computer, video conferencing system, router, or similar device. The B channels are used for so-called bearer services such as voice, data, or video. Through a process known as bonding, the two channels of a BRI can be tied together to provide 128 Kbps of bandwidth. Most LECs offer BRI as a line-side option from local switches, and most major PBXs support BRI line cards. A special station set is required for ISDN, or as shown in Figure 15-1, ordinary station equipment can be installed behind a terminal adapter (TA). The station equipment can be a telephone set, a card in a key telephone system, a card in a PC, or a direct interface into a Group 4 fax machine or a video codec. Unlike a POTS telephone, an ISDN instrument has intelligence that enables it to perform selected functions in concert with the central office.

A PRI interface is a 24/32 channel T1/E1 circuit with one or two channels reserved for signaling. A PRI can connect to any device designed to be PRI compatible, including PBXs, routers, T1/E1 multiplexers, central office switches, tandem switches, or computers. The equipment can select 64 Kbps channels singly, in pairs, or with some systems, in an N X 64 configuration. N X 64 means contiguous bandwidth equal to N quantity of 64 Kbps channels. For example, 384 Kbps is a popular bandwidth for conference-quality video, and is defined as an HO channel. The LECs provide dial tone over PRIs, and the major IXCs provide direct customer access over PRIs. The separate data channel allows communication between the IXC’s switch and the PBX so channels can be switched at will to outgoing, 800, 900, video, data, or any other such service that can be delivered over T1/E1.

The objectives of ISDN are

_ To provide end-to-end digital connectivity

_ To gain the economies of digital transmission, switching, and signaling

_ To provide users with direct control over their telecommunications services

_ To provide a universal network interface for voice and data

ISDN ARCHITECTURE

ISDN network architecture is based on ITU-T standards, with standards in the United States largely driven by Telcordia. ISDN standards are formulated on the first three layers of the OSI Model. The standards specify how information is encoded and how supplementary services such as calling features are provided.

In the United States, the two major switch manufacturers, Nortel and Lucent Technologies, developed their own ISDN protocols initially. This first edition was known as NI-1, which defines the most commonly needed set of features for regular telephone service plus several Centrex features. The main objectives of NI-1 are terminal portability and switch interoperability. Terminal portability enables users to move to a new location and have their telephone work without concern about the type of central office system. Switch interoperability enables switches of different manufacture to communicate over SS-7. Unfortunately, these objectives were not all realized because the switches were not completely compatible.

 A subsequent version, NI-2, expands on NI-1 by defining the PRI standard and expanding BRI capabilities. In NI-1, some BRI features were proprietary. In NI-2, they are standardized. Features such as D-channel backup and nonfacility associated signaling are introduced. NFAS permits multiple PRIs to share D channels. The current version of ISDN is ISDN 2000.

BRI Network Terminations

BRI intends to simplify the POTS network interfaces, but it introduces a language and interfaces that can be somewhat bewildering. Figure 15-2 illustrates the interfaces and the equipment designations. In most of the world the LEC provides the interface termination from the local central office, but in the United States the customer must furnish a device known as the network termination 1 (NT1). The NT1 communicates with the central office over a two-wire U interface and provides a four-wire S/T interface toward the customer. The S/T interface provides separate transmit and receive paths for attached devices.

Many devices have the NT1 built in. The NT1 handles physical layer functions,

which include these:

_ Separation of the 144 Kbps line signal into its constituent B and D channels

_ Termination and conversion between two-wire and four-wire interfaces

_ Monitoring of performance and maintenance functions

_ Provision of loop timing

An NT-2 is an optional terminator connecting between the NT1 and ISDN compatible devices. A terminator that includes both NT1 and NT-2 functions is known as an NT12. The NT-2 allows up to eight devices to share a passive bus. Only two of the devices can be active at one time, but any of them can be addressed through the NT-2. The D channel can be used for data communications from other devices on the bus. The protocol provides a method for allowing multiple devices to contend for the signaling channel. The NT-2, which may be built into a PBX, multiplexer, local area network, or terminal controller, performs the data link and network layer functions.

Terminal equipment that is BRI compatible can plug directly into an NT1 interface or across a shared bus. In ISDN terminology compatible equipment is known as TE-1. Equipment that is not BRI compatible is known as TE-2 equipment, and must plug into a TA. TAs are required for non-ISDN voice terminals and equipment with non-ISDN interfaces.

ISDN standards define five points of demarcation:

_ R interface is a link between non-ISDN equipment and an ISDN TA.

_ S interface connects ISDN terminals to NT-2 and NT12 devices.

_ T interface connects NT-2 and NT1 devices.

_ U interface connects NT1 and NT12 devices to the public network.

_ V interface, located in the ISDN node, separates the line termination

equipment from the exchange termination equipment.

 The ISDN access line replaces the local loop of pre-ISDN services. It is a single twisted-pair metallic line with a maximum length of 6500 m. An NT1 and the ISDN switch communicate over a full-duplex connection by using 2B1Q (2 binary, 1 quaternary) signal with echo cancellation. The transmitting power of the fourwire input to the NT1 splits between the line and an equalizing network in the NT1. An electronic filter determines whether a line signal is original data or an echo caused by a mismatch between the line and the network. Echoes are canceled out so only the original signal remains.

Service Profile Identifiers (SPIDs)

SPIDs are an optional feature that identifies the services and features the central office switch provides. SPIDs are configured at device initialization. The SPID format is usually the 10-digit phone number of the ISDN line, plus a four-digit number that identifies features on the line. The SPID varies with LEC, central office equipment type, and generic features the switch supports. The SPID can be the most confusing aspect of ordering a BRI line although some devices are self-configuring.

PRI Network Terminations

ISDN PRI is a standardized architecture for the interface between CPE and the

PSTN. The PRI protocol is in three layers:

_ Layer 1 (DS-1): Electrical, synchronization, and framing

_ Layer 2 (LAPD): Packetization, error detection, and flow control

_ Layer 3 (Q.931): Call control messages

In North America a PRI is 23B+D, while in the European system it is 30B+2D. B channels can be bonded to provide H channels, which provide higher bandwidth. The H channel standards are

_ H0 = 384 Kbps (6 B channels)

_ H10 = 1472 Kbps (23 B channels)

_ H11 = 1536 Kbps (24 B channels)

_ H12 = 1920 Kbps (30 B channels)—International (E1) only

PRI offers features that are valuable to many customers. Depending on the carrier’s service offering, any of the CLASS features discussed in Chapter 12 can be provided. PRI service offers the following advantages compared to using analog trunks:

_ Hardware costs are reduced in the PBX. In most PBXs, a PRI card costs

about the same as an eight-port analog trunk card, and uses one-third

the number of slots.

_ ISDN trunks with their call-by-call service selection provide listed directory number, outgoing, DID, and other services without the need for special trunk groups. Depending on the size of the system, trunk requirements can be reduced substantially.

_ Transmission performance is enhanced. Digital trunks can be operated with no loss and imperceptible noise. By contrast, the loss and noise of analog trunks increase with the distance from the central office.

_ ISDN trunks can be provided over self-healing networks, reducing vulnerability to outage.

_ Trunks can be added up to the capacity of a PRI with no wiring work or hardware additions on the customer’s premises.

 The following is a brief description of some of the more popular features:

_ Call-by-call service selection: Without call by call, a CPE switch must be equipped with multiple specialized trunk groups. For example, separate groups might be required for DID, two-way CO trunks, outgoing local trunks, foreign exchange trunks, toll-free lines, and tie trunks to a distant PBX. With call by call, any B channel can be designated for any purpose. The CPE switch and the central office negotiate the call and select the channel or channels to assign it to.

_ D-channel backup: A B channel can be selected as backup to the D channel to prevent service failure in case the signaling channel fails.

_ Nonfacility associated signaling: NFAS enables multiple PRIs to share one or more D channels.

_ Release link trunk: This feature is similar to antitromboning. If a user on a CPE switch receives a call from the central office and forwards or transfers it to another node on the PSTN, the two switches recognize that connection to the CPE switch is no longer required and drops the trunks.

_ Trunk antitromboning: Tromboning refers to a situation in which two B channels associated with the same D channel are being used in parallel. For example, if a user in a CPE switch receives an incoming call, and conferences on a third party on the PSTN, two B channels are tied up. If the two switches are equipped for antitromboning, they release one of the B channels.

_ Two B-channel transfer: This feature enables a user to transfer two B channels to another termination.

_ Wideband dial-up: This feature permits dialing a connection using multiple B channels.

ISDN APPLICATION ISSUES

The primary issues surrounding ISDN are cost and availability. Where the service is both available and cost effective, managers will find little reason not to use it. Cost must be examined closely before reaching a conclusion. Even if the cost is greater than an equivalent number of analog lines, the two-way nature of ISDN

means that fewer trunks will be needed for the same grade of service.

Anyone who is selecting a new PBX, or even a key system, should consider ISDN compatibility. Applications are emerging that require the digital connectivity of ISDN, and as the service becomes common, applications such as these will develop:

_ Video conferencing, primarily desktop devices

_ High-speed access to remote databases such as the various on-line services and Internet

_ High-speed remote access to LANs

_ High-speed image applications such as Group 4 facsimile

_ Second line in the home applications driven by the development of telecommuting

_ Dial backup for routers to restore leased line failures

The same list of applications can also apply to Internet, so to some degree IP and ISDN compete to provide the same services.

Many LECs offer attractive ISDN tariff rates where facilities permit. Note, however, that in many cases only one switching system in a wire center is equipped for ISDN, which means that a line cannot be moved between switches without a number change. Some LECs permit customers to avoid the number change by using number portability. The major IXCs provide PRI as an alternative to the T1/E1 connections over which they offer bulk outgoing and incoming services. With call-by-call service selection the channels can be allotted to any service, which increases utilization and reduces costs.