by Michael Starkenburg
Thanks to recent massive advertising by the Regional Bell Operating Companies, Integrated Services Digital Network (ISDN) is quickly gaining acceptance among consumers as the up-and-coming digital communication system. Most people, however, aren't aware that ISDN has been around for over ten years.
The vision of ISDN is to provide digital clarity and reliability in every home and office. ISDN data connections are many times faster than today's analog modems. ISDN can provide both voice and data service simultaneously over existing wiring.
Besides the speed benefits of the data ISDN service, ISDN provides many voice features to the residential and small office user that were previously only found on large scale PBX systems. Call forwarding, call hold, and caller ID are just a few of the features commonly found in ISDN service.
Finally, ISDN is expandable. The service is designed to support applications like a simple home office, and can also replace T-1 connectivity for large installations. As ISDN becomes widely available, it may make video conferencing and other high-bandwidth applications a reality.
The multitude of ISDN acronyms can be daunting to the end user. In this chapter, we will explain some of the basics behind ISDN, and help you understand the technology behind the alphabet soup.
TIP: This chapter provides a good overview, but if you want to drill down on any specific area, there is a wealth of information available on the Internet. In particular, an excellent place to start is Dan Kegel's ISDN page: http://www.alumni.caltech.edu/~dank/isdn/.
The earliest phone systems relied on human operators to make connections between customers as each call was made. This system was quickly replaced by automated switches, which allowed the end user to "dial" their desired party. Besides being less labor-intensive, these switches allowed the phone companies to more quickly scale the phone network to meet the surging demand.
In the 1950s, telephone companies began looking for ways to make the phone network more efficient. How could they put more calls on a switch or a wire? They began to explore how digital technology in the central offices could increase the flexibility and expandability of the existing analog phone network.
Through the 1960s, telephone companies began to deploy digital switches into their central offices. Human voices traveled from offices and residences to these central office switches as an analog signal. At the switch, they were digitized, and sent digitally on to another central office, where they were converted back to an analog signal and sent to their final destination.
To carry a voice signal on a digital carrier, the telephone companies had to develop equipment that could convert the voice signal. They did this by sampling the signal, or taking a measurement many times per second and converting that measurement into a number. That numerical data was then transmitted over the digital link, and converted back into analog at the receiving switch.
While the human voice can produce sounds in the range between 50 and 1500Hz, engineers found that most of what we hear could travel in the range between 300 and 3400Hz. In effect, they cut off the very highest sounds, and the very lowest sounds but still had an acceptable quality signal.
So, the phone companies needed to provide 3Khz of bandwidth for each voice signal, plus 1Khz for separation between calls, for a total of 4Khz. They determined that for an accurate digital signal, they needed to sample at twice that rate, or 8,000 times per second.
Each sample resulted in a number that was represented in an 8-bit digital "word". So, to transmit the human voice digitally, the company needed to provide enough bandwidth to send 8 bits of data, 8 thousand times per second, or 64Kbps. As you'll see later in this chapter, this became the foundation for the architecture of ISDN.
TIP: Standard analog modems work by converting the digital signal into an audible tone. This audible tone travels from your computer to the phone company's central office, where it is converted back to digital as described here. Because of the limitations of all that conversion, analog modems can't effectively transfer any faster than 28.8Kbps. Modem manufacturers use compression techniques to effectively get higher rates (hence 33.6Kbps and 56Kbps modems). ISDN never has to convert the signal from digital to analog and back again, which is why you can transmit so much more data across the same line.
These central office digital switches were connected together by a new type of digital communications link: the T-1 carrier. The T-1 carrier could carry twenty-four of these 64Kbps voice channels, and it used the same amount of copper wire as only two analog voice calls. The T-1 circuit is explained in greater detail in Chapter 16, "T-1 and Fractional T-1."
In an effort to find the most cost-effective solution, engineers developed several different types of digital conversion and switching hardware. Not all of these types of hardware could talk to each other. The phone companies recognized that they needed a digital standard.
In the late 1960s, the Consultative Committee for International Telephony and Telegraph (CCITT) began working on the Integrated Digital Network (IDN). Besides trying to standardize digital hardware, the IDN would combine the functions of switching and transmission into one piece of hardware, further raising the efficiency of the network.
In the IDN, however, the last segment between the central office and the customer premises remained an analog connection. The phone companies would still have to perform expensive analog-digital conversion within their central switch.
The concept of ISDN was introduced in 1972 as a possible solution. By moving the analog-digital conversion equipment into the customer premises, the phone company could provide both data and voice services over a single line. The voice service would be digitized at the customer premises, combined with any data services, and then these Integrated Services would be transmitted to the phone companies central office.
Throughout the 1970s, the telephone network was slowly upgraded to better switching hardware and higher volume connectivity. The phone companies explored digital technology further, and began providing T-1 service directly to customer premises.
The majority of this research was funded and executed by AT&T Bell Labs. In 1984, the Department of Justice broke AT&T up into our existing phone companies, known collectively as the Regional Bell Operating Companies (RBOCs). Each of the seven companies became responsible for their own product development, with help from a new research consortium called Bellcore. Before long, there were seven widely divergent research efforts into ISDN, each with its own requirements. This is a primary reason why, even today, ISDN offerings and prices are different from region to region.
To further complicate this, the Department of Justice ruling forbid the RBOCs from manufacturing customer equipment. Hardware developers were frustrated by the major differences between the RBOCs ISDN implementations, and progress was slow.
In the early 1990s, ISDN began to roll out in major metropolitan markets. Driven by demand from computer and network technology users, the RBOCs began to offer businesses and finally residential customers the opportunity to try the new technology.
Today, ISDN hardware and service is widely available to both businesses and residences. Prices are dropping, and service is reaching an acceptable level. A large number of Internet providers allow ISDN access to their services. It is clear that ISDN is quickly becoming a standard telecommunication offering for the US market.
The existing digital network is incredibly complex. It contains literally thousands of switches and trunk connections, each capable of handling hundreds of calls simultaneously. Describing the network in detail would require its own book, but we can get the general idea by looking at what components make up a connection.
In Figure 13.1, you can see the components of a local call. In this example, the call goes from Subscriber A's local loop to his central office. From there, the call is routed to a Tandem office, and on to a second central office. The second central office routes the call to Subscriber B's local loop. This process causes the clicks and delays you hear in an analog system, but is noiseless and instantaneous in a digital system.
FIGURE 13.1. Local call components.
In Figure 13.2, you see an example of a long distance call. A long distance call travels the local loop-central office-tandem office path like a local call, but is then routed to one of hundreds of long-distance tandem switches.
The local loop, or as it is often called, the last hundred feet, is the piece of the phone system that you are most familiar with. In a traditional phone system, this is the only piece of the network that remains analog. In an ISDN network, the local loop itself is digital.
The first piece of the local loop is commonly referred to as CPE, or customer premises equipment. The telephone company provides service up to a junction box somewhere on your property, called a demarcation point, or demarc. From the demarc, all the hardware and cabling is considered CPE. In an analog system, this would include your telephone equipment and any cabling installed in your residence or office.
FIGURE 13.2. Long distance call components.
In an ISDN system, the local loop is significantly more complex. The cabling is identical in most cases, but the CPE is far different. Since the phone company is sending the full digital signal to the demarc, you need to provide the equipment to handle the physical and electrical termination for this signal, and the equipment to interface it to your telephone or data equipment. These devices, called Network Terminators and Terminal Adapters, are explained later in this chapter.
From the demarc on your property, the phone company runs wire to its local Central Office. Because of the limitations of the copper wire, your Central Office will usually be within 18,000 feet of your premises.
The central office is the core of all the local loops in the area. The copper wire from each residence and business terminates in the central office switch.
A phone company switch is actually a complex computer with hundreds or thousands of input/output ports. Central office switches, also called Class Five switches, handle call-setup, teardown, and monitoring. They manage features like call waiting and call forwarding, and communicate with other switches in other central offices.
NOTE: There are a number of different models of switches in service today. Because each switch has different features and capabilities, you may need to know which switch your local phone company office uses to properly install your ISDN equipment. Common switch models in the United States include the AT&T 5ESS, the Northern Telecom DMS100, and the Siemens EWSD.
The phone companies have been upgrading these switches to be ISDN compatible, and between 75 and 90 percent of North America is now able to get ISDN connectivity. If your local switch is not capable of servicing ISDN subscribers, you may still be able to get ISDN by using a BRITE (basic rate ISDN terminal extension) device. Your local phone company will be able to give you more details about ISDN availability.
Beyond the local central office, the complexity of the network grows exponentially. To help engineers understand this complexity, the network was originally designed with five tiers. These tiers are described in the following sections.
The fifth tier was called the end office, and is what we now call the central office. The central switch is known as a Class Five switch to fit in this hierarchy.
The fourth tier provides interoffice communication, through what is known as a tandem (or Class Four) switch. The tandem switch routes traffic between Class Five switches so that all switches do not need direct trunk connections. Figure 13.3 shows how the network looks with and without tandem switches.
FIGURE 13.3. Networks with and without a Tandem Switch.
The remaining tiers, which handled long distance routing between tandem (Class Four) switches, lost their identity with the breakup of AT&T. Now, all the Class Four and Five switches are grouped into over a hundred geographic areas called Local Access and Transport Areas (LATA). When a call goes from one Class Five switch to another in the same LATA, it is generally handled by the local RBOC. When a call originates in a different LATA from its destination, it is then handled by an Inter Exchange Carrier.
The phone company switches communicate using a language called Signaling System Number 7 (SS7). SS7 is an out of band signaling protocol, which means it does not share bandwidth with the actual signal. Analog systems commonly used in-band systems for call setup and tear-down. This resulted in user perceivable delays while the voice channel was used for setting up the call. By using SS7, the ISDN network can set up calls almost instantly.
Besides providing basic call setup communication between switches, SS7 also enables services like caller ID (formally known as Automatic Number Identification, or ANI). Other supported services, such as automatic callback and call forwarding, are starting to be offered in almost all residential markets, even to analog customers.
ISDN was designed by engineers in large, bureaucratic phone companies. Accordingly, every element of the ISDN standards has an acronym. It's hard to hold a discussion about ISDN that doesn't sound like secret code. Hopefully, this section will help you decipher this code. Most of the information in this section comes directly from the ISDN standards developed by the International Telecommunications Union. For more information on any specific term, device, or protocol, you may want to refer to their website at http://www.itu.ch.
An ISDN transmission circuit is merely a logical grouping of data channels. Each of these channels have a specific purpose and handle a specific amount of bandwidth. By grouping these channels in different configurations, ISDN can be used for almost any digital transmission application, no matter how much bandwidth is required.
The core of any ISDN channel is the bearer channel, or B-channel. A single B-channel carries 64Kbps of digital traffic. This traffic can be a digitized voice signal, digitized video, or raw data. The 64Kbps throughput, as we discussed earlier, is the perfect amount of bandwidth to sample a voice signal.
NOTE: In some areas, the phone companies only provide 56Kbps B-channels. This is for backward compatibility with older phone systems that use that 8K for in-band signaling.
B-channels are usually used in groups of 2, 23, or even more, to provide additional bandwidth or voice lines. To control the transmission of this data, they are always combined with a D-channel.
The D-channel is primarily used for out-of-band messaging. This allows the call setup and tear-down signals to have their own dedicated channel, while the entire bandwidth of the B-channel is left for the actual data.
D-channels can be different sizes, depending on how many B-channels they are controlling. Naturally, as more B-channels are in use, there will be more signaling bandwidth needed. For a Basic Rate Interface (with two B-channels), the D-channel is a 16Kbps pipe, which a Primary Rate Interface (with 23 B-channels) requires a 64Kbps Pipe.
Earlier in the chapter we mentioned SS7, the language used to communicate between phone company switches. The language that travels on a D-channel is the little brother to SS7, and is known as DSS1. DSS1 communicates between your customer premises equipment and the telephone companies switch, and handles the setup and tear-down of calls.
NOTE: For applications that required large amounts of bandwidth, standard configurations of many B-channels were set up. The H-channels, which are described in greater detail later in this chapter, provide bandwidth between 384Kbps and 135Mbps.
Although B-channels and D-channels can be combined in any number of ways, the phone companies wanted to support a manageable number of standard configurations. Two standard access interfaces were developed, the Basic Rate Interface and the Primary Rate Interface.
The Basic Rate Interface (BRI) is a logical grouping of two 64Kbps B-channels and one 16Kbps D-channel.
NOTE: While some people will promote the BRI as being a 144Kbps data channel (64 + 64 + 16 = 144), remember that only the 128Kbps B-channel bandwidth (64 +64 = 128) is commonly available to the user. The 16Kbps data channel is reserved for signaling in most circumstances.
The BRI was intended for residential or home-office use. It was designed as the maximum amount of data that could use existing wiring. A BRI will allow users to access both voice and data services simultaneously.
Depending on your hardware, you can connect up to eight distinct devices to your BRI. This allows you to build a network of devices, both phone, data, and video, and use any three of them at the same time.
While the BRI was designed as the maximum amount of data that could flow over normal wiring, the Primary Rate Interface (PRI) was designed as the maximum amount of data that could flow over a T-1 carrier. A PRI consists of 23 64Kbps B-channels, and a single 64Kbps D-channel, for a total of 1.472Mbps available to the user. If you add the D-channel, you'll find that the total bandwidth of a PRI is 1.536Mbps, almost the same as a T-1 Carrier. An additional 8K of bandwidth is used by the T-1 Carrier for framing.
NOTE: The 1.544Mbps T-1 is available in North America only. In Europe, the corresponding circuit is called an E1, and carries 2.044Mbps. Euro-ISDN will makes use of this additional bandwidth by assigning thirty B-channels to each PRI instead of the North American standard of 23.
The primary use of the PRI today is for large scale voice services. Many Private Branch Exchange (PBX) units include their own ISDN hardware, and will accept a PRI directly from the telephone company. The ability to easily reallocate trunk lines in the PRI makes it a natural for this function.
ISDN standards clearly identify the various devices that connect the end user to the telephone network. Each specific function was defined in the standards as a separate device, although today many ISDN vendors will combine several functions into a single piece of hardware. We'll explain this in greater detail later in the chapter.
After defining the functional devices, the standards went on to define the protocol that each device used to speak with the devices on either side of it. These protocols were assigned the letters R, S, T, and U as identifiers, and were termed reference points.
Figure 13.4 shows the order of functional devices, and the locations of the different reference points. The alphabet soup of the user-network standards gets confusing, so refer back to this figure often.
FIGURE 13.4. Functional devices and reference points.
NOTE: Because vendors can combine several functional devices in a single piece of hardware, you may not see all reference points in your installation. For example, a device that is called an NT1 by the vendor will often provide the functionality of an NT2 device as well. In this case, the vendor will usually label the customer-side interface and "S/T" interface. This interface will then connect to the terminal equipment or terminal adapter. You never see a pure "T" reference point in this configuration.
The ISDN standards define five groups of functional devices: The NT1, NT2, TE1, TE2, and the TA.
The Network Terminator 1 (NT1) is the device that communicates directly with the Central Office switch. The NT1 receives a "U" interface connection from the phone company, and puts out a "T" interface connection for the NT2, which is often in the same piece of physical hardware.
The NT1 handles the physical layer responsibilities of the connection, including physical and electrical termination, line monitoring and diagnostics, and multiplexing of D- and B-channels.
CAUTION: In an analog system, the phone company provides the power to the phone line. This explains why, in a blackout, you can still call the power company to complain. The ISDN system, however, relies on power at your site, generally provided by the NT1 device. In the event of a power failure at the NT1, you will lose the ability to make any ISDN calls, either voice or data. For this reason, it may be wise to keep an analog backup line.
The Network Terminator 2 (NT2) sits between an NT1 device and any terminal equipment or adapters. An NT2 accepts a "T" interface from the NT1, and provides an "S" interface In most small installations, the NT1 and NT2 functions reside in the same piece of hardware. In larger installations, including all PRI installations, a separate NT2 may be used. ISDN Network routers and digital PBX's are examples of common NT2 devices.
The NT2 handles data-link and network layer responsibilities in ISDN installations with many devices, including routing and contention monitoring.
A Terminal Equipment 1 (TE1) device is a piece of user equipment that speaks the "S" interface language natively, and can connect directly to the NT devices. Examples of a TE1 device would be an ISDN workstation (such as the SGI Indy), an ISDN fax, or an ISDN-ready telephone.
Terminal Equipment 2 (TE2) devices are far more common--in fact, every telecommunications device that isn't in the TE1 category is a TE2 device. An analog phone, a PC, and a FAX are all examples of TE2 devices. To attach a TE2 device to the ISDN network, you need the appropriate Terminal Adapter. A TE2 device attaches to the Terminal Adapter through the "R" interface.
These devices connect a TE2 device to the ISDN network. The Terminal Adapter (TA) connects to the NT device using the "S" interface and connects to a TE2 device using the "R" interface.
Terminal adapters are often combined with an NT1 for use with personal computers. Because of this, they are often referred to as ISDN modems. This is actually not accurate, because TA's do not perform analog-digital conversion like modems.
When you order an analog telephone line from the company, you are assigned a simple ten digit number that permanently identifies your connection. Unfortunately, ISDN service isn't that simple. There are five separate identifiers involved in each connection. Two of these, the Directory number and the Service Profile Identifier, are assigned by the phone company when service is ordered. These are probably the only two identifiers you will ever have contact with.
The other three identifiers are dynamically assigned, and are normally transparent to the end user. The important difference between the identifiers is when they are assigned: the SPID is assigned once, when service is set up. The TEI, SAPI, and BC are assigned every time a device is connected to the system.
The directory number (DN) is simply the normal ten-digit phone number the phone company would assign any analog service. In an ISDN service, however, there is much more flexibility in the use of the directory number. The directory number is only a logical mapping, and doesn't have the one-to-one relationship with your line that analog service would have.
A single device can have up to eight directory numbers. Since an ISDN BRI can support up to eight devices, this means that a single BRI can support up to 64 directory numbers. This might allow you to have eight extensions in an office, each with its own directory number. In most BRI installations, you are assigned two directory numbers.
The opposite is also true. You can have several ISDN channels or devices share a single directory number. This is helpful when you want the ability to receive more than one call at a time, without using call-waiting or having the caller hear a busy signal.
The Service Profile Identifier (SPID) is the most important identifier to the ISDN system. Like the directory number, it is assigned by the phone company when service is set up. In fact, most SPIDs consist of the Directory number for that device, plus some extra digits.
The format of the SPID varies widely from region to region. The SPID is usually between 10 and 14 digits long, depending on the policy of your phone company. You may receive one SPID per B-channel, one per ISDN device, or one for your entire line.
Generally, the phone company will issue one SPID for each application. So, for example, you may receive two SPIDs for your BRI: aaa-ppp-nnnn-01, and
aaa-ppp-nnnn-11, where aaa is the area code, ppp-nnnn is the directory number, and the last two digits identify each B-channel. Your SPID is unique across the switch, and must be properly entered into your terminal equipment before your ISDN devices can be recognized by the switch.
When an ISDN device or Terminal Adapter is connected to the line, it follows a hard coded procedure to set up the physical communication with the phone company switch. During the procedure, it sends the SPID to the switch. If the SPID is correct, the switch uses the information it has stored about your service profile to set up the data link. The SPID is not sent again unless the device is disconnected from the network.
TIP: A SPID never seems to be around when you need it. Some phone companies do not include it on the bill, and SPIDs are not published like Directory Numbers. When your service is installed, it would be wise to label your ISDN devices with the SPIDs they will be using. If you have an internal device, you may want to label the wall jack.
A Terminal Endpoint Identifier (TEI) is allocated dynamically for each SPID as the switch recognizes it. The TEI changes every time a device is connected to the network, while the SPID remains the same. The TEI identifies the particular ISDN device to the switch. Because it doesn't have to be manually entered like the SPID, the TEI is usually transparent to the end user.
The Service Address Point Identifier (SAPI) identifies the particular interface on the switch that your devices are connected to. The SAPI is switch-specific, and is never seen by the user.
The TEI and SAPI, when taken together, uniquely identify the link between your equipment and the phone company switch. This identifier, called the bearer code (BC) or call reference, changes every time a new connection is set up.
As ISDN evolved, several different versions of ISDN were developed and tested. Each version had its own protocols, features, and identifiers.
Some feature sets were standardized early on by market forces: everyone wanted to have call forwarding and caller ID. Other features were standardized through technical necessity--the BRI became a standard partially because it was the most bandwidth that could reliably travel over existing wiring.
But a large number of issues remained--services and tariffs varied widely from region to region. The phone companies have been working diligently to produce a set of standards that will allow all ISDN implementations to interact.
A network of governmental and non-profit organizations has been built to create telecommunications standards, including those used in the ISDN. These organizations generally represent a geographic region or group of companies.
The International Telecommunications Union (formerly known as the Consultative Committee for International Telephone and Telegraph or CCITT) has been the primary global force in telecommunications standards since 1865. As a part of the United Nations, the ITU is comprised of governments from around the world, who represent their constituent companies and organizations.
Two industry associations provide the US government with the majority of standards recommendations. The Telecommunications Industry Association (TIA) handles the terminal equipment and the Alliance for Telecommunications Industry Solutions (ATIS) handles the network standards.
In Europe, the European Telecommunications Standards Institute (ETSI) provides recommendations for both equipment and network technology. ISDN is much more common in Europe than in the United States, in part because of the strength of the ETSI. In 1989, 26 European network providers agreed on a standard for Euro-ISDN and implemented it by 1993.
The Open Systems Interconnect model is a description of how telecommunications devices communicate at different levels. The OSI model consists of seven layers, with each subsequent layer being one step closer to the users.
An ISDN addresses the first three layers (physical, data-link, and network) of the OSI model. The higher layers are addressed by the hardware and software in the terminal equipment (such as a PC).
The physical layer of the OSI model defines the hardware, wiring, connectors, and other electrical and physical characteristics of a network connection. In the ISDN standard, the physical layer is represented by the user-network interface discussed earlier in this chapter, including the functional device and reference point specifications. The ISDN standards also address cabling and hardware issues we will discuss later in the chapter.
The OSI Data-link layer handles connection maintenance and error correction. In the ISDN standards, a protocol called Link Access Procedures-D Channel (LAP-D) handles flow control, error retransmission, and packet framing.
In the OSI model, the Network layer defines signaling procedures between the network and users. The ISDN standards for this layer describe the signaling protocols DSS1 and SS7, which we discussed earlier in this chapter.
In the United States, the most widely used and important ISDN standard is called National ISDN. Created by Bellcore in 1991, National ISDN provided standardization on features, protocols, and hardware specifications for the BRI. Since the original release of the National ISDN standard, a group at Bellcore called the National ISDN Council has been proposing improvements to the standard. Today, there are several levels of the National ISDN standard.
The first set of National ISDN (NI-1) proposals called for a set of core services to be offered with all Basic Rate interfaces, including
Today, the NI-1 standard is almost fully deployed. Most terminal hardware and all major switches conform to NI-1.
Released in 1992, the first major goal of National ISDN-2 (NIS) was to standardize the service features for the Primary Rate Interface, as NI-1 did for the Basic Rate Interface. That effort was only partially successful. Many companies felt that the standard didn't offer enough features, so there are still several proprietary PRI products on the market.
A secondary goal of NI-2 was to standardize and simplify the identification process. Bellcore proposed that ISDN vendors implement a series of standards that would allow automatic setup of hardware and easy connection to the ISDN network. The Automatic Terminal Setup system, along with efforts to standardize the format and use of the SPID, should reduce the technical difficulty that has caused ISDN to be such a mystery to end users.
Finally, NI-2 calls for several enhanced features to the existing services, including better billing support and more consistency of features between switch types.
NI-2 support is available in many brands of terminal hardware today, but has not been extensively rolled out in the central office switches. If your terminal equipment supports NI-2, you will be ready to use these features when the local switch in your area is upgraded. Current expectations are that NI-2 will fully roll out in late 1997.
National ISDN-3 (NI-3) is a huge leap forward in terms of ease of use and features, but won't widely be available in the near future.
Some of the features in the NI-3 standard include
By the time the National ISDN-1 specification had been released, several phone companies had already rolled out ISDN services. Most of these services had features which the subscribers (primarily businesses) found useful, but that weren't included in the NI-1 spec.
So, the as these phone companies started to adopt the NI-1 spec, they kept some of the additional features in their services. Services with these additional features are called Custom ISDN services. Services that conform strictly to the NI-1 guidelines are called Standard ISDN.
Generally, if you have NI-1 compliant hardware and software, you can still connect to a custom service, but you cannot make use of the extended features. If you want to use custom features, ask your terminal hardware vendor and phone company for information about their Custom ISDN support.
One of the fundamental differences between the POTS analog system and the ISDN system is the complexity of the hardware installed in customer premises. Luckily, vendors are aware of this barrier to consumer acceptance of their products, and there are a number of resources available to assist you in choosing and installing the right hardware for the job. Some ISDN user groups are listed below: Bellcore http://www.bellcore.com
California ISDN Users Group http://www.ciug.org
Florida ISDN Users Group http://www.ccg4isdn.com/isdn/fiug.html
New York ISDN Users Group http://www.interport.net/~digital/index.html North American ISDN User's Forum http://www.niuf.nist.gov/misc/niuf.html In a typical ISDN installation, you are responsible for three major categories of hardware: cabling within your premises, the network terminator, and the terminal equipment or adapter.
NOTE: The information in this section is heavily focused towards the US ISDN subscriber. European ISDN standards are significantly different: for example, in Europe, your ISDN service provider will be responsible for network termination equipment. Also, it is unlikely that any ISDN equipment purchased in Europe will work in the US without a hassle. If you are planning an ISDN network that includes a European installation, you may want to consult with an ISDN professional in that country.
The first piece of hardware you will be responsible for is the cabling at your premises. The phone company will provide and maintain the cable for your service up to the demarcation point. From that point, you need to ensure that you are using the right type of wire and connectors, and that you have enough to support your application.
People generally take wiring for granted in office and residential installations, but there are a few things you should know about your phone wiring.
Throughout the phone system, copper wiring is run in twisted pairs. The twists in the wire assist in keeping noise levels down, which increases the effectiveness of the wire as a signal carrier. The twisting has to remain constant throughout all the wire, connectors, and jacks in the cabling system.
The Telecommunications Industry Association released standards for wiring that specified three common categories of wire:
Generally speaking, if you need to install new wiring to support your ISDN installation, you should use Category 5 cable. While Category 3 cable is acceptable for ISDN use (and is the minimum installed in most residences), Category 5 cable is capable of handling future enhancements to your data network.
For a residential installation, you should check to ensure that you have at least one pair for each BRI you plan to install, plus one pair for each analog line you plan to keep.
For an office installation, your wiring will depend on the specific terminal equipment you plan on using. An ISDN line will enter your facility on one pair (for a BRI) or two pair (for a PRI carried on a T-1). At the network terminator, which may be your PBX or ISDN router, the number of pairs required may change. In general, you will want at least two pair available to each ISDN device in an office setting.
The Network Terminator, as we described earlier in the chapter, is the electrical and physical termination of the connection between your premises and the telephone company. The NT1 is a relatively simple device--so simple, in fact, that is in often integrated into a piece of terminal equipment.
If you are using only a single ISDN device on a BRI, you might want to consider getting a terminal adapter with an integrated NT1. An integrated device is easier to install and is guaranteed to be compatible with the attached terminal equipment. There are a few reasons why you may still want to consider a standalone NT1 with a single device:
If you are using more than one ISDN device on a BRI, you will probably want to use a standalone device. While you can find integrated NT1/TA's with more than one ISDN device connection, they are generally not available in many configurations. In addition, power and connectivity to all devices on the entire ISDN circuit will be provided by the integrated unit, which may not be desirable. If you need to disconnect or turn off the integrated unit, your other devices will not function.
If you are installing a PRI circuit, the Network Terminator device will probably be integrated into your PBX or ISDN router. If you have an existing PBX that doesn't support ISDN, you can get an intermediate device commonly called a drop and insert unit. The drop and insert unit will manage the ISDN power, call handling and signaling functions, and will deliver the PRI channels into the PBX as standard analog trunk lines. If you are considering replacing existing trunk lines with an ISDN PRI, you should retain an ISDN consultant to help you look at all the possibilities.
The most important function of a Network Terminating device, be it standalone or integrated, is to take a long-run ISDN connection from the phone company and provide a short-run connection to the terminal equipment. An NT device should include one or more of each of the following connections:
TIP: If your NT1 provides an analog POTS jack, that doesn't necessarily mean you can get rid of your POTS service entirely. First, most NT devices do not provide the additional voltage to the POTS jack that tells the analog phone to ring. Second, in the event of a power loss to the NT device, the analog POTS jack will be unusable, along with the ISDN devices. If you are set on getting rid of all your analog lines, you'll want to be sure to get an NT1 that provides a ring generator and a backup power supply--but I don't recommend it.
The second important function of the NT device is to provide power to the ISDN circuit. Depending on the criticality of your ISDN application, you may want some sort of battery backup on your NT. There are three different ways to get this kind of battery backup:
NOTE: If you get an integrated NT/TA, especially one that is designed to live inside a PC, that piece of hardware may need to remain on all the time to provide service to your ISDN devices and analog devices attaches to the NT/TA.
The last, and most important, piece of equipment you will need to choose is the terminal equipment or terminal adapter. The Terminal adapter provides the connectivity between the NT device and your end-user hardware, be it a PC, a LAN, or an analog device.
If you want to connect more than one PC to your ISDN BRI, or if you have an existing LAN, you may want to consider an ISDN router. Most home or small office installations require connectivity for only one PC and possibly an analog phone or fax and will use an ISDN terminal adapter.
The most important decision you will make when choosing your Terminal Adapter is whether to use an external or internal TA.
The external terminal adapter is often referred to as an "ISDN modem" because it looks like a standard modem, and connects to a serial port like a modem. Of course, the word "modem" is a misnomer: ISDN TAs do not MOdulate or DEModulate the data at all. But because they function just like a modem, the external TA is often the simplest choice. You merely connect the TA to the ISDN line and to a serial port on the computer. You can access the ISDN line through your serial port like a modem.
The drawback of an external TA is speed--or lack thereof. With an external TA, you can rarely take advantage of the full 128Kbps of an ISDN BRI. External TAs are generally slower than their internal counterparts for two reasons.
CAUTION: Before committing to an external ISDN TA, be sure you have an available serial port. On a PC, there are only four serial ports available, and two of them are not addressable by all software programs. On most Macs, there are only two serial ports. If you have maxed out your serial ports with other devices, you may want to consider an internal adapter--although it may still be tricky to configure around all those other peripherals.
Enternal TAs need to be configured with your SPID before they can initialize. There are three ways this can be done, depending on your specific TA:
Internal Terminal Adapters fit inside your PC, and connect directly to the PC bus. Because they don't need to use a serial port, they avoid the speed drawbacks of the external Terminal Adapter. You can expect to get the full 128Kbps performance out of your ISDN line using an internal adapter.
Internal cards generally offer POTS jacks and integrated NT devices, and cost less than a comparable external device. They usually support BONDING and MLPPP, and don't have several different pieces to clutter up a work area.
In many ways, an Internal TA is like a network card--it is configured in a similar way, and performs similar functions. Of course, this means that your internal TA may be as difficult to configure as a network card. If you are using an operating system that doesn't support Plug and Play devices, you should prepare yourself for a struggle configuring the device.
TIP: With an internal adapter, if your PC locks up or fails, an attached analog fax or phone may stop working. If you plan on using the analog port for mission critical applications, you may be happier with an external NT device running your analog service, and an internal TA for your data applications.
You should only consider an Internal TA if you have no plans to expand your ISDN network to include another PC or ISDN device. Internal TAs rarely offer the option to connect other ISDN units, although it is common for them to include one or more POTS jacks.
The next category of terminal equipment is more complex. Instead of just interfacing a single device to the ISDN network, ISDN routers perform a "bridging" function. A network bridge connects two dissimilar networks, providing the necessary protocol translation between them. The most common ISDN bridge connects an ISDN BRI to an Ethernet Local Area Network.
Most ISDN-to-Ethernet bridges provide some sort of routing ability. Routing is a more complex version of bridging, where each packet is analyzed to determine if its destination is on the LAN or across the ISDN line. Depending on your specific hardware choice, you can find ISDN-to-Ethernet bridges that route several different protocols.
Which protocol your hardware supports depends on your installation. If you are linking two appletalk LANs in different locations, you'll need an ISDN router which supports this protocol. Similarly, if you want to connect two Novell networks, you'll need a router that supports the IPX protocol. To connect your LAN to the internet, you need to route TCP/IP.
ISDN-to-Ethernet routers generally have an integrated NT device. They provide one RJ11 connection for a BRI "U"-interface from the phone company, and an RJ45 connection for your Ethernet LAN. If your application needs more than 128Kbps bandwidth, you'll want to look for an ISDN router that can accept a PRI interface.
Other important features to look for in your ISDN router include
One of the most useful features of the ISDN network design is the ability to combine 64Kbps B-channels for applications that require larger amounts of bandwidth. While many of today's data applications will be well serviced by a BRI or PRI, you can imagine that future applications will need different solutions.
As you add more B-channels to an ISDN installation, you need to choose a method of controlling and combining the channels into one connection. Each B-channel may travel a different path through the phone network to reach its destination. These different paths may introduce timing differences between data passed over different B-channels. ISDN aggregation techniques can both control multiple B-channel connections, and compensate for timing delays between channels.
When the ITU originally developed the ISDN standards, they included several common B-channel aggregations, called H-channels. H-channels can be set up by sending a special signal to the switch, which will handle opening the lines and controlling the connection. Table 13.1 shows the common H-channel connection types.
Because the local switch recognizes the H-channel standard, it is the easiest method of aggregating bandwidth. No additional hardware is needed, and the call is handled just like any other ISDN call.
| Name | Bandwidth | B-Ch. | Applications |
| H0 | 384Kbps | 6 | Video conferencing, broadcast audio. |
| H11 | 1.563Mbps | 23 | Replaces the T-1. |
| H12 | 1.920Mbps | 30 | Replaces the E-1. |
| H21 | 32.8Mbps | 512 | High speed data. |
| H22 | 44.2Mbps | 690 | High speed data. |
| H3 | 60/70Mbps | 1050 | Broadcast video. |
| H4 | 135Mbps | 2112 | Broadcast video, data trunks. |
Another option for applications that need a specific bandwidth allocation that isn't addressed by an H-channel is to use special ISDN router hardware that dials a groups of B-channels individually.
The drawback of using an H-channel is that you might waste bandwidth if your application has dynamic needs. The H-channel connection keeps all B-channels open for the entire length of the call, regardless of the demand for bandwidth. To solve this, the industry has developed several methods of dynamically allocating bandwidth.
Inverse multiplexing is the process of taking one signal and distributing it over several B-channels. The inverse multiplexers are hardware devices that sit on either side of the ISDN connection. These devices are capable of picking up and dropping additional B-channels as bandwidth needs change during the call.
Because ISDN is priced per B-channel per second, you can save a significant amount by using an inverse multiplexer. On the other hand, the expense of the device, and the additional complexity it causes your installation, may not be worth the line charge savings.
Each inverse multiplexer has its own method of determining when to pick up or drop a line. Because there is little standardization, it is unlikely that two different inverse multiplexers will be able to talk to each other reliably.
Many manufacturers have started adding support for the BONDING protocol to their hardware devices. The BONDING protocol, short for Bandwidth ON Demand Interoperability Group, was developed in an effort to standardize on a single inverse multiplexing design. Many ISDN BRI terminal adapters include BONDING support to allow the end-user to access the full 128Kbps. Because BONDING is a hardware feature, it is fast and efficient, but you need to ensure that both ISDN devices on the connection support it.
The most common method of B-channel aggregation is MultiLink PPP. MLPPP is a software solution that functions very similarly to BONDING. The only major difference is that standard MLPPP connections cannot dynamically allocate additional B-channels during the call. With standard MLPPP, you choose one or two B-channels at the beginning of the call, and you keep that number of B channels open for the entire duration of the call.
A newer standard, developed by ASCEND, fixes this problem. Calls using MLPPP+ can begin with only one B-channel. When a certain percentage of that B-channel is used, the second B-channel is opened up. When the second B-channel is not needed, it is dropped. MLPPP+ is being widely implemented, and is supported now by several manufacturers.
Once you go beyond the 2B+D BRI, ISDN starts to compete with a T-1. Because both services are digital, and they can both carry the same amount of data, which should you use?
The answer is a financial one: which will cost less in your application? T-1 lines are priced on a flat-rate monthly scale. If you use a fairly constant amount of bandwidth, you will probably find that a T-1 or fractional T-1 is more cost effective.
If your application has widely variable bandwidth needs, you may want to use ISDN aggregation. With ISDN, you only pay for the B-channels you have open at a given time.
Once you've decided that ISDN is the correct choice for your data networking application, and chosen the appropriate hardware, you need to get ISDN service from your local carrier.
Until very recently, ordering ISDN was a lengthy and painful process. RBOC's had little standardization, and quite frankly, seemed uninterested in providing acceptable ISDN customer service. Today, however, ordering your ISDN line is fairly simple, depending in large part on which RBOC services your locality.
Each RBOC handles the ISDN service in its local area. The following list of ISDN carriers provides you with a list of contact information for these companies. Ameritech ISDN
National ISDN Hotline: 1-800-832-6328 http://www.ameritech.com/ Bell Atlantic ISDN
ISDN Sales & Tech Center: 1-800-570-4736
Residential ISDN Sales & Service Center: 1-800-204-7332 http://www.bell-atl.com/ BellSouth ISDN
ISDN HotLine: 1-800-428-4736 http://www.bsonline.bellsouth.net/cgi-bin/WebObjects/ISDN Cincinnati Bell ISDN
ISDN Service Center: 513-566-3282 http://www.cinbelltel.com/business/isdninfo.html NYNEX ISDN
ISDN Hotline: 1-800-438-4736 http://www.nynex.com/isdn/isdn.html Pacific Bell ISDN
ISDN Service Center: 1-800-472-4736
24-Hr. ISDN/Availability Hotline: 1-800-995-0346 http://www.pacbell.com/products/business/fastrak/networking/isdn/info/index.html SNET ISDN
ISDN Sales & Technical Support Center: 1-800-430-4736 http://www.snet.com SBC ISDN
ISDN Information: 1-800-792-4736 http://www.sbc.com/swbell/shortsub/isdn_services.html Stentor ISDN
ISDN "Facts by Fax": 1-800-578-4736 http://www.stentor.ca U S WEST
ISDN Fax File Server: 1-800-728-4929 http://www.uswest.com/isdn/index.html
TIP: If you don't know which RBOC you are serviced by, or have questions about the level of service in your area, you can refer to the National ISDN Hotline, provided by Bellcore. They can be reached at 1-800-992-ISDN, or by sending e-mail to isdn@cc.bellcore.com.
ISDN charges, like all telecommunications fees, are heavily regulated. While these tariffs are different for each RBOC, all RBOCS offer a measured rate. In a measured system, you pay a flat monthly fee, plus a second fee per second. Generally, the fee per second is charged for each B-channel, so if you have established a 128Kbps connection, you will pay twice the tariff of a single B-channel 64Kbps connection. In some areas, there is an option for a monthly flat rate instead of measured. If you are a heavy user, this option may appeal to you.
Tariffs are generally set per capability package (which is a set of features, described later in this section). If you don't need all the features of an advanced capability package, you may be able to save money by choosing a simpler one.
ISDN voice features rival the most complicated of PBX systems, but it's unlikely that you will need all of those capabilities. The exact features you can get in your area will vary. Table 13.2 details some of the most common features.
| Category | Features |
| Voice | Caller ID |
| Call Hold | |
| Call Conferencing | |
| Call Forwarding | |
| Call Transfer | |
| Message Waiting Indicator | |
| Data | Caller ID |
| Multi-Line Hunt Group | |
| Basic Business Group |
As you can imagine, there are thousands of combinations of the ISDN voice and data features. In an effort to make the ordering process a little less cumbersome, the North American ISDN User's Forum created a standard set of functionality called "capability packages." Now, instead of having to specify every last detail of your connection, you can simply provide the carrier with the correct order code for the feature set you need. A list of the order codes is provided in Table 13.3.
| Package Title | Included Services |
| Capability Package A | Basic D-channel Packet services. No voice capabilities are provided. |
| Capability Package B | Circuit switched Data on one B-channel. No voice capabilities are provided, basic voice capabilities (no features) are supported. |
| Capability Package C | Alternate Voice/Circuit Mode Data on one B-channel. Only basic voice capabilities (no features) are supported. |
| Capability Package D | Voice on one B-channel and basic D-channel Packet services. Only basic voice capabilities (no features) are supported. |
| Capability Package E | Voice on one B-channel and basic D-channel Packet services. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer. |
| Capability Package F | Voice on one B-channel and basic D-channel Packet services. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer, and uses CACH EKTS. |
| Capability Package G | Voice on one B-channel and Circuit Mode Data on the other B-channel. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer. |
| Capability Package H | Voice on one B-channel and Circuit Mode Data on the other B-channel. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer, and uses CACH EKTS. |
| Capability Package I | Circuit Mode Data on both B-channels. No Voice or Packet capabilities are provided. |
| Capability Package J | Alternate Voice/Circuit Mode Data on one B-channel, and Circuit Mode Data on the other B-channel. Only basic voice capabilities (no features) are provided. |
| Capability Package K | Alternate Voice/Circuit Mode Data on one B-channel, and Circuit Mode Data on the other B-channel. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer. |
| Capability Package L | Alternate Voice/Circuit Mode Data on one B-channel, and Circuit Mode Data on the other B-channel. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer, and uses CACH EKTS. |
| Capability Package M | Alternate Voice/Circuit Mode Data on both B-channels. Only basic voice capabilities (no features) are provided. |
| Capability Package N | Alternate Voice/Circuit Mode Data on one B-channel, Circuit Mode Data on the other B-channel, and D-channel Packet. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer. |
| Capability Package O | Alternate Voice/Circuit Mode Data on one B-channel, Circuit Mode Data on the other B-channel, and D-channel Packet. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer, and uses CACH EKTS. |
| Capability Package P | Alternate Voice/Circuit Mode Data on both B-channels, and D-channel Packet. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer. |
| Capability Package Q | Alternate Voice/Circuit Mode Data on both B-channels, and D-channel Packet. Voice capabilities include Three-way (Conference) Calling, Call Hold, Call Drop, and Call Transfer, and uses CACH EKTS. |
| Capability Package R | Circuit-Switched Data on two B-channels. Data capabilities include Calling Number Identification. No voice capabilities are provided. |
| Capability Package S | Alternate Voice/Circuit-Switched Data on two B-channels. Data and voice capabilities include Calling Number Identification. |
| Capability Package T | Voice on two B-channels and basic D-channel packet. Only basic voice capabilities are provided, with no features. |
| Capability Package U | Alternate Voice/Circuit-Switched Data on both B-channels. Voice capabilities include non-EKTS voice features including Flexible Calling, Call Forwarding Variable, Additional Call Offering, and Calling Number Identification (which includes Redirecting Number Delivery). Data capabilities include Calling Number Identification (which includes Redirecting Number Delivery). |
| Capability Package V | Alternate Voice/Circuit-Switched Data on two B-channels. Voice capabilities include non-EKTS voice features including Flexible Calling, Advanced Call Forwarding (such as Call Forwarding Variable, Call Forwarding Interface Busy, Call Forwarding Don't Answer, and Message Waiting Indicator), Additional Call Offering, and Calling Number Identification (which includes Redirecting Number Delivery). Data capabilities include Calling Number Identification (which includes Redirecting Number Delivery). |
TIP: In the manual for your Terminal Equipment, the manufacturer of your equipment may suggest which capability packages will work best with your specific piece of hardware. If not, you can call the vendor or ask your retailer for advice.
ISDN is just now coming into its own, with advances in hardware, wider availability, and stronger consumer acceptance. The next section explains what you can expect from ISDN in the near future.
Since 1994, Bellcore has accepted enhancements to the National ISDN standards from the switch manufacturers. Enhancement features are included in the NI for a given year (NI-XX) when at least two participating switch suppliers have the capability available in the first quarter of that year. NI-1995, for example, provided more standardization for SPID format and call control.
We will see more standardization on services between the regional telephone companies. Also, ISDN hardware and software will become easier to install and use.
As bandwidth needs rise, the standard aggregation techniques for ISDN will not be adequate. Applications which require large amounts of bandwidth for short periods of time are inefficiently served by today's ISDN implementation.
Enter Broadband ISDN, which does away with the idea of "channels" of ISDN. With B-ISDN, the end-user gets true bandwidth on demand, only paying for the bandwidth used.
B-ISDN supplies bandwidth in excess of 1.544Mbps, or faster than a T-1. B-ISDN will be found in three common forms:
The Digital Subscriber Line is the basic underlying technology behind the T-I and the Switched 56 services in common use today. DSL systems come in three basic types:
Industry observers are questioning if the DSL technologies might make ISDN obsolete over the next few years. The answer depends on the speed with which the telephone companies can deploy DSL switches and hardware, and the growth of high-bandwidth consumer markets.
As a WAN technology, ISDN is becoming more affordable, easier to use, and widespread. A wide range of vendors now make interoperable ISDN equipment, which has driven hardware prices down. Most local carriers are offering both residential and office ISDN at prices that make the technology a viable competitor for T-1, 56K, and even POTS connectivity.
The metered nature of ISDN and the wide variety of equipment now available makes it best suited for WAN applications that
After reading this chapter, and the other chapters in this book, you should be able to determine if ISDN is the right WAN technology for you.
© Copyright, Macmillan Computer Publishing. All rights reserved.
){kind=link}
){kind=link}
){kind=link}
){kind=link}