ISDN - Digital Telephony From End-To-End
Why Integrated Services?
As in the case of the GSM, the ISDN protocols have been developed to take advantage of the fact that most telephony occurs in a digital form once signals have left the local loop. By
providing a framework for customer access to digital services, the coding and decoding of signals can be limited, providing high-quality voice and data services over shared media. As
ISDN user services become more widely available, connection costs will continue to decline, with the possibility of dual 64-kbps connections to the home or office becoming a reality.
ISDN user services include voice, data, fax, and in the higher-rate interconnects, video.
ISDN includes a wide range of protocols. These protocols handle several interfaces including connections to the Customer Premises Equipment (CPE), and a common channel signaling
network that provides a mechanism to control the calls. One of the more challenging aspects of ISDN is getting comfortable with the broad set of jargon that is applied to the types of
network connections and interfaces. For example, the local loop includes an S interface between the T and NT interfaces, different types of U interfaces between the NT and carrier
network. Further complicating matters, many of these interfaces can differ depending on whether you are in the US or Europe, and in the case of the network, whose equipment you
happen to be using.
Given the complexity of the complete telecommunications network, the exchange network
architectures and related signaling protocols comprise a substantial amount of information requiring separate attention.
As in the case of many networks, we'll take it on faith that the
exchange network can be treated as a virtual "cloud" providing users with the necessary services.
The Reference Points
Interfaces and elements within the ISDN Subscriber Access Network (SAN) have been identified by the CCITT
(now ITU) with a range of rather cryptic identifiers. For the SAN, these
interfaces are shown in Figure 1, which is reprinted from the ITU-T blue book.
Figure 1 - Subscriber Access Network Reference Configuration
In this figure, the items in the boxes represent components within the subscriber access network, and represent specific names and functions, that are:
This collection of components can be grouped in a number of configurations, depending on the specific application.
- TE - Terminal equipment, includes any user terminals, phones, computers, etc. Two classes of terminal equipment are included in the reference model, those that directly interface using
the ISDN standards (TE1), and those that conform to other standards, such as X.25, RS-232, X.21 and others; these are referred to as falling into the class of device TE2.
- TA - Terminal Adapter, provides protocol and signaling conversion services to allow TE2 equipment connected over the TA R interface to communicate over the ISDN via the S
- NT - Network Termination. NT1 and NT2 components cooperate to complete the connections between the Customer Premises Equipment (CPE) and the Network. NT2 is most typically
found in the customer premises and can be included in a PABX or other switching device. NT1 equipment is placed between the network equipment and the customer equipment to
provide signal, power, and logical conversions.
- LT - Line Termination. Terminates the digital transmission system on network facilities. Provides fault isolation, signal regeneration and code conversion
- ET - Exchange Termination. This is the point where the SAN is connected to the public Information Exchange Network (IEN)
The Signaling Channels - Core Components For Information Transfer
Interactions between components (and subsets of components) occurs over the signaling channels.
Typically, this information is either data being transferred between end-users and
other end-users, or control information between the end-users and the network.
These signaling channels include the following:
In addition to these channel configuration, a service called "Bonding" provides users with the ability to combine a group of channels into a newer configuration. Several products, both
in Basic Rate and Primary Rate interfaces provide this capability. For example, where the typical Basic Rate interface consists of 2 B channels and a D channels, some products are
available in which the 2 B channels are combined to provide users with a single 128 kbps circuit.
- B Channel - running at 64 kbps, the B channel is most often used for the transfer of digitized voice information. Data information is also transferred over the B channel in either circuit or
packet formats. Voice information is most typically encoded as a 8 kHz stream of 8 bit Pulse Code Modulation (PCM) voice samples. Data information can be transferred either as raw
synchronous, or through one of several rate or format conversion protocols including V.120, and X.21 Terminal adaptation protocols.
- D Channel - operating at either 16 kbps or 64 kbps, depending on the physical interface, the D channel is primarily used for the exchange of signaling information that controls the
connections over the interfaces user signaling channels. In some cases, the D channel can be used to transfer user information, most typically, X.25 packet connections are sent over
this channel. Call control information transferred over the D-Channel generally uses the HDLC variant Link Access Protocol-D (LAPD) as a link protocol for the exchange of layer 3
(Q.931) signaling information.
- H Channel - actually a series of channel types, the H channels are intended to carry user information well in excess of 64 kbps. Channel rates are:
- - H0 - 384 kbps
- - H11 - 1.536 Mbps
- - H12 - 1.920 Mbps
- - H21 - 32.768 Mbps
- - H22 - 44.160 Mbps
- - H4 - 135.168 Mbps
For asynchronous information, the Asynchronous Transfer Mode (ATM) cell-based architecture is frequently applied within the constraints of ISDN. In ATM, several "virtual"
channels are transferred over the network, with the network arbitrating the amount of channel capacity allocated to each virtual circuit based on the channel characteristics negotiated at
connection setup, and the amount of capacity available at any particular instant. The fundamental aspect of this interface is the concept of a "cell" that is 53 octets in length, with 5
octets of control information, and 48 octets of data.
Interfaces - The ISDN Physical (and part of the Link) Layers
The ISDN physical interfaces provide the physical exchange of data between users and the network and synchronize the transfer of information over the various signaling channels
carried by the interface.
The ISDN Basic Rate Interface (BRI) is ISDN's closest match to the single phone interface. With physical interfaces that can operate over either single or dual-pair interfaces, the BRI is
most typically configured to support two B channels for user signaling, and a single 16 kbps D channel for call control and user signaling. The primary types of interfaces usually seen
in the BRI are the S interface, a six-line interface that carries both signals and power and can support several terminals on a line, and the U interface, a two wire interface that multiplexes
signals in each direction over the differential pair. There are at least three standards for the U interface, depending largely upon the locality of the implementation, and the manufacturer
of the switching gear. Each of these interfaces provides services to activate, deactivate and synchronize the channels, providing the BRI layer two protocols with a discrete access to
the signaling channels (2B+D).
There are two general types of ISDN Primary Rate Interface (PRI) standards, the 1544-kbps standard based on T-1 and used in North America and Japan, and the 2048-kbps used in most
of the remainder of the world. The T-1 based PRI uses the DS1 channel assignment as the basis for synchronization of the channels, with the 1544 kbps signal divided into 24 64-kbps
channels. In the 1544 interface, the last channel is used as a D channel, controlling the use of the remaining 23 channels, which can be mixed into D, B, or H0 channels.
Broadband services differ from the primary and basic rate services in that they must accommodate both fixed as well as variable demand circuits. High-rate physical interfaces, such as
SONET are used as a physical layer to carry ATM cells. Within ATM, information streams are broken into a series of packets called "cells" that can each carry 48 octets. Each ATM cell
carries 5 octets of signaling information that is used by the switching equipment to route and manage the flow of information. Features are included within the ATM standards to serve
connections in which the flow of information is not constant. Depending on the negotiated service type, bursts of information are either forwarded or, in overload conditions dropped to
permit higher priority traffic to traverse the network.
Service Protocols - Control and Transfer of User Information
Several types of protocols are used to control and forward information through the ISDN. Control protocols, such as the ISDN call control protocols and the network Signaling System 7
(SS7) are used to manage and direct the connection of users within the network.
At the user interface, circuit based services include synchronous voice and circuit switched data. Synchronous voice signals are typically transmitted in PCM format at 64 kbps, with a
sample rate of 8 kHz, providing an approximate analog channel bandwidth of from 150 Hz to 3200 Hz. Data services are generally provided through use of rate adaption protocols for
digital interfaces including X.21 and other data services ranging from 600 bps to 56 kbps.
There are two ways in which the ISDN can be used to support packet transfer services. In one case, the ISDN can be used to provide a link interface to a packet switch, providing a
service essentially the same as connecting the packet terminal equipment to a switch over a phone line. X.25 packet services can also be obtained over the D channel, with the ISDN
network providing the packet routing services.
Control of the connections is handled through a series of protocols that operate over the D-channel. The link layer protocol used to control the calls is an HDLC variant protocol
referred to as Link Access Protocol - D (LAPD), that provides terminal equipment with a connection oriented link service to the switching equipment. Layer 3 protocols provide call
control services to establish calls, report the status of the calls, and to terminate the calls.
Once in the network, a different set of protocols, based on out-of-band, common channel signaling, take control. The SS7 protocols support a second network used to transfer control
and management information between the switching elements of the network. Signaling points (SP), are the end users of the switching information, and can be switching exchanges,
such as the Nokia DX 220, the AT&T 5ESS and others. Signaling transfer points (STP) are used to forward the information through the control network to permit the SP to coordinate
switching and control activities. A completely separate set of control protocols are managed in this network, with four layers of protocols used to connect elements of the control
Helgert, Hermann J., Integrated Services Digital Networks -- Architectures Protocols Standards, Addison Wesley 1991
Stallings, W. ISDN: An Introduction, Macmillan, New York, 1985