T1: A Survival Guide | O'Reilly Media

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

All is flux, nothing is stationary.

—Heraclitus

We live in a digital world.

By digital, I don't mean the overused "digital economy" or anything remotely related to punditry. In the beginning, there was analog. Transmitted signals were subject to fading, interference, and noise. All that changed with digital encoding and transmission. Digital signals offer superior properties for information. Perfect regeneration is possible with sophisticated error detection and correction methods. Signals can be sent without loss or distortion over arbitrary distances. Think of all the digital systems people use on a daily basis: CDs are digitized audio, many cable systems are now digital, and the Internet cannot be overlooked. Then there is the telephone, which was one of the earliest experiments in digital technology.

Zeros and ones are not just the building blocks of the future—they are also the building blocks of the present.

It was not always so, of course. Telephony was initially all analog, and signal processing was not even a discipline. Social changes and increased mobility in the 1950s threatened to break the telephone system by swamping its capacity. Rescue came by way of digital technology and its increased capacity and quality. Telephone companies invested in the new digital technology to transmit digitized voices over long distances and, in doing so, laid the foundation for the long-distance data transmission so critical to the Internet. T1, though now used overwhelmingly for data transmission, found its roots in the U.S. telephone network in the digital upheaval of the 1960s.

When Alexander Graham Bell invented the telephone in 1876, he followed a course of action familiar to many inventors today: he filed patent applications. His patent applications were approved in 1876-77, and he promptly sought financial backing to build a company to collect patent license fees. In 1877, he formed the Bell Telephone Company with two investors.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

When Alexander Graham Bell invented the telephone in 1876, he followed a course of action familiar to many inventors today: he filed patent applications. His patent applications were approved in 1876-77, and he promptly sought financial backing to build a company to collect patent license fees. In 1877, he formed the Bell Telephone Company with two investors.

The year 1878 saw the construction of the first telephone exchange in New Haven, Connecticut. In the first exchanges, telephone calls were switched manually by operators using plugboards—a far cry from the modern central offices with computerized switching equipment. In the three years that followed, Bell Telephone licensed many local operators and constructed exchanges quickly in major cities.

Western Union attempted to enter the infant telephony market in 1861, after securing some patent ammunition of its own with the construction of the first transcontinental telegraph line. Bell Telephone, by this point renamed American Bell, fought Western Union's entry tooth and nail in court until, eventually, Western Union agreed to stay out of telephony, provided that Bell did not compete in the telegraphy market. With the collapse of Western Union's ambitions in the telephone market, its major supplier, the Western Electric Manufacturing Company, was financially hobbled. Western Electric was an engineering firm with a record of successfully manufacturing novel inventions (including the incandescent light bulb), but was forced to sell a controlling stake to American Bell in 1881 to survive. Western Electric subsequently became the manufacturing arm of American Bell. In 1883, Western Electric's Mechanical department was founded. Ultimately, the Mechanical department grew into Bell Telephone Laboratories, a name interwoven with much of the technological developments later in this story.

In 1885, American Telephone and Telegraph (AT&T), founded to build and manage the long-distance network, started operations in New York; the long-distance network reached Chicago by 1892. Because electrical impulses degrade with distance, newer technologies were needed in order to reach farther. Loading coils, which reduce the decay of signal strength in the voice band (frequencies less than 3.5 KHz), enabled the long-distance network to reach Denver. Transcontinental service needed one more major enhancement, the addition of practical electronic amplifiers, which allowed AT&T to extend the long-distance network to San Francisco by 1915. In the midst of all this, AT&T became the parent company of American Bell in 1899.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

It may not have been obvious at the time, but the telephone network was on the cusp of a fundamental change driven by several factors. Early long-distance connections, especially those across an ocean, were full of static and could fade quite abruptly. Development of appropriate cabling would eliminate the problems caused by the use of radio waves, but noise and distortion of the analog signal across such great distances posed a large problem. Postwar prosperity in the U.S. dramatically increased the demand for telephone service, bringing with it millions of new telephone users and requiring ever more operators to connect calls. Better means of providing service had to be found. Fortunately, two new technologies came to the rescue.

The transistor, invented in 1947 at Bell Labs, made it possible to construct sophisticated electronics. Today's semiconductor and computer industries owe their existences to the transistor. Transistors also made AT&T's "Electronic Switching System" possible. In the early 1960s, AT&T built and field-tested specialized, programmable devices for controlling the telephone network. These devices, of course, were computers, but the terms of the 1956 consent decree prohibited AT&T from manufacturing computers. AT&T followed the time-honored tradition of redefining terms in the argument. AT&T deployed the first such device, the No. 1 ESS, in New Jersey in 1962 following a field trial. Electronic switching made it possible to run the telephone network without having to dramatically increase the number of operators because the previous mundane tasks of taking numbers and connecting calls could be relegated to the phone switches. Electronic switching also made it feasible to build a telephone network with enough call-routing capability to eliminate operator intervention in connecting long-distance calls; by 1965, 90% of telephones in the U.S. were directly dialing long-distance numbers.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Digitalization of the telephone network sowed the seeds for a second government action against AT&T. The appearance of standardized digital carrier systems in the 1960s had blown apart Vail's justification of a natural monopoly based on incompatibility.

Round two of the antitrust wrangling with AT&T commenced in 1974. Although distracted by legal battles with the Department of Justice, AT&T remained the dominant telecommunications carrier. In the 1970s, AT&T pioneered fiber-optic transmission, beginning with an experimental system in Chicago that could carry 672 voice channels on a single strand of glass.

Legal struggles concluded in 1982. AT&T retained businesses that operated in competitive marketplaces, namely long distance, Bell Labs and other R&D centers, and Western Electric. The Regional Bell Operating Companies (RBOCs), which were formed from AT&T's local exchange assets, continued to regulate local telephone service as a natural monopoly. After a long preparation period, AT&T split off from the RBOCs on January 1, 1984, in a momentous event known as divestiture. T1 had been used as a trunk line in the AT&T network for many years, but it was only after divestiture that T1 became a tariffed service that could be ordered by customers.

Problems with the T-carrier hierarchy stem from one simple fact: T-carrier systems were designed for AT&T's voice network. They can be used to transport data, but the adaptation to this use is not always a clean one.

Voice systems do not have the same requirements for operations, administration, maintenance, and provisioning (OAM&P) as data applications do, and the T-carrier hierarchy does not have enough overhead for even simple management tasks. For signaling purposes, the T-carrier system initially depended on robbed-bit signaling, which significantly diminishes data throughput, as we will see later.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

The world hates change, yet it is the only thing that has brought progress.

—C. F. Kettering

One of the complaints that many data-networking veterans have when venturing into the telecommunications world is the bewildering number of acronyms and strange terminology that await them. Before diving into small details about different components of a T1, some background with the technology is essential. This chapter introduces the terms and basic structure of a T1 circuit so that successive chapters can delve into detail on the most important components.

Figure 2-1 shows a high-level diagram of the link between an Internet service provider (ISP) and a customer, delivered over an archetypal T1 circuit. Due to the number of components required to build a T1, Figure 2-1 is divided into four parts. In the classic case, two wire pairs, one each for reception and transmission, combine to form a T1. Devices called repeaters are used to allow for transmission of the signal over long distances. Signals degrade as they travel along the path, but repeaters recover the digital input data and retransmit the digital data at full strength. In theory, this allows for perfect transmission of data because a digital signal can be perfectly recovered and an exact bit-for-bit copy sent on to the destination. Repeaters divide the circuit into a series of spans, or paths between repeaters.

As technology marched on, the technology used to deliver T1 circuits changed. In most locations, only the end span connecting subscribers to the central office (CO) is delivered over the four-wire copper interface. Fiber-optic transport systems such as SONET are usually used to move data between the two COs at each end of the circuit. Among its many benefits, SONET provides protection switching. If a link in a SONET network fails, protection switching diverts connections onto a backup path within milliseconds.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Figure 2-1 shows a high-level diagram of the link between an Internet service provider (ISP) and a customer, delivered over an archetypal T1 circuit. Due to the number of components required to build a T1, Figure 2-1 is divided into four parts. In the classic case, two wire pairs, one each for reception and transmission, combine to form a T1. Devices called repeaters are used to allow for transmission of the signal over long distances. Signals degrade as they travel along the path, but repeaters recover the digital input data and retransmit the digital data at full strength. In theory, this allows for perfect transmission of data because a digital signal can be perfectly recovered and an exact bit-for-bit copy sent on to the destination. Repeaters divide the circuit into a series of spans, or paths between repeaters.

As technology marched on, the technology used to deliver T1 circuits changed. In most locations, only the end span connecting subscribers to the central office (CO) is delivered over the four-wire copper interface. Fiber-optic transport systems such as SONET are usually used to move data between the two COs at each end of the circuit. Among its many benefits, SONET provides protection switching. If a link in a SONET network fails, protection switching diverts connections onto a backup path within milliseconds.

Figure 2-1: Typical T1 span

Once the span reaches the near-end CO, it is routed over the telco's network to the far-end CO. Between all the COs, redundancy and spare capacity is provided by the telco through SONET protection switching. Network outages due to telco equipment failure are far more common on the repeatered copper spans than on the SONET rings. It is not uncommon for these spans to be interrupted by construction equipment. When construction activity breaks up the end spans, protection switching is not available because only one path exists from the customer location to the CO. This type of interruption is frequently called

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

The dead govern the living.

—Auguste Comte

As mentioned in Chapter 1, T1 is a time-division multiplexed stream of 24 telephone calls. Each call is carried by a 64-kbps digital stream called a DS0. Several meanings are ascribed to the acronym DS; you may hear any combination of the words data, digital, service, stream, speed, and signal. DS0 is the bottom rung of the T-carrier hierarchy. Higher levels of the hierarchy are built by multiplexing lower levels together. Understanding the T-carrier hierarchy starts with understanding DS0 transmissions.

AT&T's initial digital leased-line offering was called the Digital Dataphone Service (DDS). DDS was offered at several different speeds, ranging from 2,400 bps to 56 kbps. Service initially topped out at 56 kbps because a portion of the signal is required for timing overhead. DDS circuits formed the Internet backbone in December 1969. Traffic growth eventually overwhelmed the limited-circuit capacity, and the Internet backbone was upgraded to T1 circuits in the late 1980s. Mushrooming traffic led to a further network upgrade to T3 in the early 1990s. Eventually, economics conspired to kill DDS as a standalone service. T1 is not much more expensive, so companies that required more throughput, higher reliability, and guaranteed service levels shifted to T1, while budget-conscious users migrated to cheaper technologies such as DSL.

The telephone network is a circuit-switched network. Each telephone call is assigned to a dedicated path through the network for its duration. Telephone calls require a 64-kbps path through the network. At each hop between switching offices, the trunk lines are divided into 64-kbps channels, which are called DS0s. These individual 64-kbps channels are the building blocks of the telephone network because a DS0 has sufficient capacity for a one-voice telephone call.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

The telephone network is a circuit-switched network. Each telephone call is assigned to a dedicated path through the network for its duration. Telephone calls require a 64-kbps path through the network. At each hop between switching offices, the trunk lines are divided into 64-kbps channels, which are called DS0s. These individual 64-kbps channels are the building blocks of the telephone network because a DS0 has sufficient capacity for a one-voice telephone call.

To the phone company, 64 kbps means 64,000 bits per second, not the 65,536 bits per second a computer engineer might expect. However, this basic quantity is used by all telephone companies throughout the world. Bundling is different: in the U.S., the T1 standard bundles 24 DS0 channels, plus framing overhead, into a single T1. In Europe, an E1 circuit is composed of 30 DS0 channels, plus framing and signaling overhead. Given the differences in telecommunications between the U.S. and the rest of the world, it should be obvious that since we all use a given standard, there must be something special about 64,000 bits per second.

Transmitting understandable human speech requires using frequencies up to 4,000 kHz. To adequately represent an analog signal in digital format, the analog signal must be sampled at twice the maximum frequency in the signal, a result known as the Nyquist criterion or sampling theorem. To adequately digitize voice, therefore, requires a sampling rate of 8 kHz. Each sample is represented by 8 bits using a technique called Pulse Code Modulation (PCM), where the value of each sample is transmitted as an 8-bit code. In one second, 8,000 samples are transmitted. With a sample size of 8 bits, the resulting data rate is 64,000 bits per second.

DS0s do not impose any higher-level structure on the data stream—they are simply an unframed, raw sequence of bits. The telephone company takes the bits from one location and moves them to another location; to the telephone company's equipment, it does not matter whether the bits are a voice telephone call or a data circuit, as long as the stream of bits obeys the rules.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

In common encoding schemes, ones are represented by voltage pulses and zeros are represented by the lack of a voltage pulse. Each pulse is approximately 3 volts in amplitude and has a 50% duty cycle, meaning it takes up half of the time slot for pulse transmission. Pulses have a tendency to spread out in the time domain as they travel down a line, as illustrated in Figure 3-1. Restricting the initial transmission to occupy half the time slot helps the repeaters, and the receiving end, find the middle of each time slot and stay synchronized.

Figure 3-1: Time domain spreading

Commonly, a scheme called bipolar return to zero or Alternate Mark Inversion (AMI) is used. One and zero are sometimes referred to as mark and space, respectively, in communications jargon. AMI gets its name from the fact that only ones, or marks, result in pulses on the line. Successive pulses are encoded as positive and negative voltages, as shown in Figure 3-2.

Figure 3-2: AMI encoding

Alternating pulse polarity enables a quick-and-dirty form of error detection. AMI specifies that polarity must alternate, so two successive pulses of the same polarity, such as the sequence in Figure 3-3, might be an error. (Bipolar schemes can catch many errors, but not all of them. Errors in voice transmission result in odd sounds and can be shrugged off. Data transmission is far less forgiving, which is why far better error-detection techniques were developed.)

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Throughput suffers when the eighth bit is blindly stuffed with a one. Maintaining synchronization requires only that enough pulses are sent down the line. An alternative to straight AMI encoding is to use a scheme based on code word substitution. A second encoding method, called Bipolar with Eight Zero Substitution (B8ZS), is able to transmit an arbitrary bit sequence. When eight consecutive zero bits are scheduled for transmission, a B8ZS transmitter replaces the eight-zero sequence with a code word that contains intentional bipolar violations (BPVs), as shown in Figure 3-4. The code word takes the form of 000VP0VP, where V is a pulse of the same polarity as the previously transmitted pulse and P is a pulse with the opposite polarity as the previous pulse.

Figure 3-4: B8ZS intentional bipolar violation

When B8ZS-capable CSU/DSUs receive the B8ZS code word containing bipolar violations, the code word is replaced with eight consecutive zeros before passing the data on to the user. Transmission at the full line rate is possible with B8ZS because it does not require a portion of the circuit capacity to be used for synchronization. Because the full line rate is available, the telco may refer to the circuit as one that has Clear Channel Capability (CCC), or some similar term. Widespread use of B8ZS is a relatively recent development. Most early digital CO equipment was designed to catch and flag bipolar violations. Moving to B8ZS required massive upgrades to remove all of these "helpful" pieces of equipment. In the late 1980s, less than 1% of telco equipment was B8ZS-capable. Extensive equipment upgrades in the past 10 years have made widespread use of B8ZS possible, so virtually all new T1s are deployed using B8ZS.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

The tyranny of the multitude is a multiplied tyranny.

—Edmund Burke

To move beyond the DS0 into higher-bandwidth realms, additional layers of multiplexing are needed. This chapter describes how DS0s are bundled into DS1. DS1 refers to a digital signal operating at 1.544 Mbps; T1 refers specifically to a DS1 delivered over a four-wire interface. Most people simply use the term T1 to refer to any digital signal at that speed and, to avoid breaking the common convention, so does this book.

Higher levels of multiplexing are used to generate further levels of the T-carrier hierarchy, such as DS3. DS3 is different from T1, though, because the much higher speed requires different encoding methods, far more precise timing, and new network-to-router interfaces. To avoid getting lost in DS3 details, therefore, this chapter details only the DS0 to DS1 multiplexing process.

Assembling higher-speed links in the T-carrier system is conceptually easy. Take a collection of lower-speed links and bundle them together as channels in a TDM framework. When 24 DS0 streams are bundled together, the result is a higher-level digital stream: DS1. Multiple DS1s are bundled together to form DS2s, and DS2s are tied together into DS3s. Table 4-1 shows the standardized data rates in the T-carrier system. While the data rate for DS4 was standardized, most of the network interface details were not.

Table 4-1: T-carrier comparison
Stream type	Speed (Mbps)	Equivalent T1s	Equivalent voice channels

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Assembling higher-speed links in the T-carrier system is conceptually easy. Take a collection of lower-speed links and bundle them together as channels in a TDM framework. When 24 DS0 streams are bundled together, the result is a higher-level digital stream: DS1. Multiple DS1s are bundled together to form DS2s, and DS2s are tied together into DS3s. Table 4-1 shows the standardized data rates in the T-carrier system. While the data rate for DS4 was standardized, most of the network interface details were not.

Table 4-1: T-carrier comparison
Stream type	Speed (Mbps)	Equivalent T1s	Equivalent voice channels
DS0	0.064	1/24	1
DS1	1.544	1	24
DS1C	3.152	2	48
DS2	6.176	4	96
DS3	44.736	28	672

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

In the beginning, T1s were deployed to users as high-capacity voice trunks and were terminated by devices called channel banks. Western Electric called the first channel bank, which was first placed in service in 1962, the D1. Later channel banks received successively higher numbers: the D2 channel bank was introduced in 1969, the D3 was completed in 1972, and the D4 made its debut in 1976. Each new product introduced new features. With the D2 channel bank, Western Electric introduced the superframe (SF) standard. Some manuals may refer to the SF format as the D4 format, which is somewhat erroneous—the D2 channel bank was sold in the late 1960s, long before the D4 channel bank became commonplace nearly twenty years later. Superframe formatting groups 12 channels together into a single structure called a superframe, as shown in Figure 4-2.

Figure 4-2: The superframe

24 DS0 channels are bundled into frames. Twelve frames are then put together to form one superframe. The framing bits from the 12 constituent frames form a frame bit sequence, which is always 100011011100, as Figure 4-2 shows. Odd bit positions in the superframe are terminal framing bits, abbreviated F_t(101010), and even bit positions are signaling framing bits, abbreviated F_s(001110).

AT&T's introduction of the D2 channel bank brought link monitoring into existence—a major step forward. Alarm signals indicated severe error conditions. Table 4-2 summarizes the alarm conditions available on superframe links. Alarms initially had color names because of the corresponding color of the indicator lights on channel banks.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

SF-framed links have two notable drawbacks. First, a yellow alarm is transmitted by setting the second bit to zero in all of the time slots in a frame. When the yellow alarm is present, no data is received. Unfortunately, setting the second bit position to zero is something that can happen frequently in user data. Altering the bits is not acceptable with data transmission, so the only solution is to use one time slot for yellow-alarm prevention and to set the second bit to one. Although this prevents a false yellow alarm, it sacrifices one DS0 worth of bandwidth. Secondly, the error-detection mechanism with SF links is quite limited. Bipolar violations are a line error check, which means that they can flag potential problems in the local copper portion of the T1 span. Errors may be introduced anywhere along the span, however. Any corruption introduced at the central office, or in the high-speed optical components, cannot be detected by T1 equipment and must be detected by higher-layer protocols. What is needed is a path error check, which verifies data integrity across its entire path from one end to another, no matter what type of transport is used.

In response to these limitations, AT&T developed the extended superframe (ESF), which was introduced on the D5 channel bank in 1982. Advances in electronics made it possible to use a smaller proportion of the frame bit sequence for synchronization and devote it to solving the problems of the SF framing format. Figure 4-3 shows the ESF superframe. As with the SF superframe, it begins with frames, each of which is made up of 24 8-bit time slots with a single frame bit at the beginning. The 24 frames are put together into a single superframe.

Figure 4-3: The ESF superframe

Of the 24 framing bits, only six are needed for synchronization. Every fourth frame contributes a bit to the synchronization pattern, 001011. CSU/DSUs can easily identify the synchronization pattern because it cannot shift onto itself. Synchronization requires 2 kbps from the aggregate T1 capacity, which is only half of the bandwidth required by SF framing.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

On circuit-switched networks, nodes must occasionally send administrative messages to each other to establish, reroute, and tear down circuits. Call-control signaling is incorporated into T1 links by a technique known as bit robbing, illustrated in Figure 4-4. Every sixth frame steals the least-significant bit from each channel to create an auxiliary channel for network messages. When the least-significant bit is altered on a PCM word, the auditory effect on a modulated voice signal is negligible.

Figure 4-4: Bit robbing

Changing bits of data, though, is disastrous. When using robbed-bit signaling on a T1, data communications equipment is usually set to ignore all least-significant bits in case a signaling bit has clobbered a data bit. Ignoring the eighth bit limits the throughput of each channel to 56 kbps. Robbing is an appropriate term indeed, given that available capacity falls from 1.536 Mbps to 1.344 Mbps!

Each robbed bit in a frame is used to create a signaling channel. SF-framed links have two channels, which are called the A channel and the B channel. ESF-framed links have four signaling channels: the A channel, the B channel, the C channel, and the D channel. (The last channel should not be confused with the D channel in ISDN.) Use of the robbed-bit signaling channels is discussed in Appendix A.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Time is the extension of motion.

—Zeno

Faster networks depend on accurate timing. As the number of bits per second increases, the time in which to look for any particular bit decreases. Getting both sides to agree on timing becomes more difficult at higher speeds. Synchronous networking is largely about distribution of accurate timing relationships.

Synchronous communications do not depend on start and stop flags to mark the beginning and end of meaningful data. Instead, the network constantly transmits data and uses a separate clock signal to determine when to examine the incoming stream to extract a bit. Distributing clock information to network nodes is one of the major challenges for synchronous network designers. Three major types of timing are used on networks: asynchronous, synchronous, and plesiochronous. All three terms derive from the Greek word kronos, meaning time. The three differ in how they distribute timing information through the network.

Asynchronous systems do not share or exchange timing information. Each network element is timed from its own free-running clock. Analog modems are asynchronous because timing is derived from start and stop bits in the data stream. Free-running clocks are adequate for dial-up communications because the time slots are much longer than on higher-speed digital networks.

Synchronous systems distribute timing information from an extremely accurate primary system clock. Each network element inherits its timing from the primary clock and can trace its lineage to the common shared clock. When AT&T operated the U.S. telephone network, the system derived its timing from the primary reference source (PRS), a cluster of cesium clocks located in Hillsboro, Missouri.

Synchronous networks may have several layers of accuracy, but the important feature is that each clock can trace timing to a single reference source. In the case of the Bell system, the primary source was labeled Stratum 1. Less-accurate devices were in higher-numbered strata. Tandem offices, also called toll offices, serviced the long-haul portions of the telephone network and were located in Stratum 2. Local switching offices were located in Stratum 3, with end-user devices such as CSU/DSUs in Stratum 4.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Synchronous communications do not depend on start and stop flags to mark the beginning and end of meaningful data. Instead, the network constantly transmits data and uses a separate clock signal to determine when to examine the incoming stream to extract a bit. Distributing clock information to network nodes is one of the major challenges for synchronous network designers. Three major types of timing are used on networks: asynchronous, synchronous, and plesiochronous. All three terms derive from the Greek word kronos, meaning time. The three differ in how they distribute timing information through the network.

Asynchronous systems do not share or exchange timing information. Each network element is timed from its own free-running clock. Analog modems are asynchronous because timing is derived from start and stop bits in the data stream. Free-running clocks are adequate for dial-up communications because the time slots are much longer than on higher-speed digital networks.

Synchronous systems distribute timing information from an extremely accurate primary system clock. Each network element inherits its timing from the primary clock and can trace its lineage to the common shared clock. When AT&T operated the U.S. telephone network, the system derived its timing from the primary reference source (PRS), a cluster of cesium clocks located in Hillsboro, Missouri.

Synchronous networks may have several layers of accuracy, but the important feature is that each clock can trace timing to a single reference source. In the case of the Bell system, the primary source was labeled Stratum 1. Less-accurate devices were in higher-numbered strata. Tandem offices, also called toll offices, serviced the long-haul portions of the telephone network and were located in Stratum 2. Local switching offices were located in Stratum 3, with end-user devices such as CSU/DSUs in Stratum 4.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

CSU/DSUs are like bridges. They have one interface in telco territory and one interface in data-communications territory. Both are serial interfaces that make use of tight timing tolerances. Appropriate configuration of the CSU/DSU to work within the timing straitjacket is essential.

In the T1 world, clock signals are not transmitted separately from the data stream. Instead, receivers must extract the clock from the data signal based on the stream itself. Each bit time slot is 648 nanoseconds. Pulses are transmitted with a 50% duty cycle, meaning that for the middle half of the time slot, the voltage is at its peak. Based on these characteristics, the receiving CSU/DSU infers time slot boundaries from incoming pulses. Ideally, each pulse comes in the middle of a time slot, so finding time-slot boundaries is simply a matter of going 324 ns in each direction. Figure 5-3 illustrates clock inference from pulse reception.

Figure 5-3: T1 clock inference from pulse reception

In practice, of course, things are never quite as simple, and CSU/DSUs must compensate for a variety of non-ideal conditions. Clock signals may exhibit both short-term and long-term irregularities in their timing intervals. Short-term deviation is called jitter, and long-term deviation is referred to as wander.

Timing on the T1 network interface from the telco is implicit and based on the content of the pulse stream. On the other hand, the serial circuit that connects the CSU/DSU to the router makes use of explicit timing. V.35, for example, includes two pairs (four leads) for sending timing signals and one pair (two leads) for receiving timing.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

T1 equipment employs a variety of techniques to compensate for variations in timing signals. Intermediate network equipment may buffer the 192-bit frames to ensure that frames are complete before forwarding them on to their destinations. CSU/DSUs are equipped with phase lock loop (PLL) circuitry to track with the more accurate clocks at the local exchange office. Occasionally, though, these measures are not enough, and timing problems occur.

Imperfect timing conditions may force network equipment to replicate or delete data in a process called a frame slip. Slips are divided into two categories. Controlled slips replicate or delete a complete 192-bit frame of data, but do not cause any problems with the T1 path. Uncontrolled slips, which are also called change of frame alignment (COFA) events, are much more severe because they disrupt the framing pattern. Controlled slips are the more benign of the two because the path remains available. Uncontrolled slips indicate more severe problems with the circuit.

Controlled slips always involve complete frames, and can be the result of either a buffer overflow or underflow. Both conditions are illustrated in Figure 5-12. In the overflow case, the second frame is lost in time unit 1 because the buffer overflows and replaces it with the third frame. Both the second frame and its framing bit are lost. Receivers use the disruption in the framing bit sequence to detect controlled slips. Controlled slips may also occur because of a buffer underflow, which causes frames to be repeated. In Figure 5-12, the buffer underflow in time unit 1 means that no fresh data is available for transmission in the second time unit.

Figure 5-12: Controlled slip operations

Uncontrolled slips are far more severe. If a buffer overflow or underflow causes a partial frame to be lost, an uncontrolled slip will occur. Framing bits shift within the bit stream, as illustrated in Figure 5-13.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

A man is as old as his arteries.

—Pierre J. G. Cabanis,Epigrams

When bringing up a T1, a CSU/DSU is required by FCC part 68 to protect the telco network from your equipment. If the CSU/DSU does not function correctly, whether due to component failure or misconfiguration, the line will not come up. Unfortunately, producing clear, understandable printed documentation has never been a goal of data communications equipment vendors. CSU/DSU manuals have improved to the point of basic intelligibility, but even the best manuals require a solid understanding of telco networking.

In a perfect theoretical world, wires have no resistance and voltage pulses can travel forever. The real world, however, is rarely anything like the world of theory. (There is a reason that physicists joke about spherical frictionless cows, after all.) Repeaters and signal regenerators along T1 spans are the first admission of a practical world. Line build out (LBO) is the second concession to practicality.

Two different types of connections are used on T1 lines. A long-haul connection is made to the telco network. The first repeater may be up to 3,000 feet away from the end-user location. CSUs must be able to drive the pulse up to 3,000 feet, so the arriving pulse at the first repeater is still strong enough to be regenerated. Long-haul line build out is used to avoid problems that arise from having multiple T1 customers in an area. It works by reducing the signal to the level required by the telco network. Short-haul connections are digital cross-connect links up to 655 feet that are made between user-owned devices such as PBXs. Short-haul line build out is also called line equalization. Equalization increases the signal strength above the reference level. For historical reasons, a network interface is abbreviated DS1, and the local side is often labeled DSX or DSX-1. Figure 6-1 contrasts short- and long-haul line build out.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

In a perfect theoretical world, wires have no resistance and voltage pulses can travel forever. The real world, however, is rarely anything like the world of theory. (There is a reason that physicists joke about spherical frictionless cows, after all.) Repeaters and signal regenerators along T1 spans are the first admission of a practical world. Line build out (LBO) is the second concession to practicality.

Two different types of connections are used on T1 lines. A long-haul connection is made to the telco network. The first repeater may be up to 3,000 feet away from the end-user location. CSUs must be able to drive the pulse up to 3,000 feet, so the arriving pulse at the first repeater is still strong enough to be regenerated. Long-haul line build out is used to avoid problems that arise from having multiple T1 customers in an area. It works by reducing the signal to the level required by the telco network. Short-haul connections are digital cross-connect links up to 655 feet that are made between user-owned devices such as PBXs. Short-haul line build out is also called line equalization. Equalization increases the signal strength above the reference level. For historical reasons, a network interface is abbreviated DS1, and the local side is often labeled DSX or DSX-1. Figure 6-1 contrasts short- and long-haul line build out.

Figure 6-1: Long-haul LBO and short-haul LBO

All T1 connections to a telco network must be made with long-haul settings. Two main reasons motivate the use of long-haul LBO. The first relates to T1 repeaters, which are designed to detect incoming pulses at a wide variety of voltages. Some repeaters, however, cannot sense pulses unless they exhibit a loss of at least 7.5 dB from the nominal transmission level. Such repeaters will not register one bits from pulses that are not sufficiently attenuated and will therefore reconstruct a stream of continuous zeros. To work with these older repeaters, some CSU/DSUs may apply a default LBO setting of -7.5 dB.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Figure 6-4 shows a typical CSU/DSU. It has a series of lights and a switch that is used to control the loopback test mode. The meanings of various common indicator lights are discussed in Table 6-2.

Figure 6-4: Typical CSU/DSU

Table 6-2: Common T1 CSU/DSU indicators and their meanings
Indicator	Common labels	Meaning
Clear to send	CTS, CS	The CSU/DSU is ready to receive data.
Request to send	RTS, RS	The DTE is ready to send data.
Carrier detect	DCD, CD	The CSU/DSU is generating carrier signal.
Send data	TX, TXD, SD	The CSU/DSU is transmitting pulses to the telconetwork.
Receive data	RX, RXD, RD	The CSU/DSU is receiving pulses from the telconetwork.
Loss of signal	LOS	When no pulses arrive within 100 to 250 bit times, LOS is declared. Under normal conditions, at least a few pulses would be present in that interval.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

CSU/DSU configuration can be a frustrating affair for many reasons. Each vendor has a slightly different way of describing features, and terms in the manual may differ from the terms your telco uses. In many cases, your service provider or telco will include a CSU/DSU in the startup charges for a new line and will even pre-configure the CSU/DSU to work with the new line. Depending on the model, configuration can be frustrating. Some CSU/DSUs are configured with jumpers or DIP switches, and the cheapest ones even require a power-cycle to reread the jumper or switch settings. Higher-end CSU/DSUs have menu-driven configuration and are far easier to use.

While the array of configuration options may be bewildering, the following options are critically important and account for most of the problems you may see:

Framing: The choice is between SF and ESF, and the former is rare and getting rarer. This setting is supplied by the telco because it must match the telco equipment.

Line code/DS0 channel speed: These two settings are related. AMI encoding is almost always associated with 56k component DS0 channels, and B8ZS is always associated with 64k component DS0s. Depending on your CSU/DSU, configuration may be required for both line code and DS0 speed or just one of the two. Like framing, these settings are supplied by the telco.

Line build out: By the time a pulse reaches a T1 repeater, it is assumed that a certain amount of attenuation will have already occurred. T1 uses LBO to adjust the outgoing transmitted pulses so that they will be attenuated to the normal amount by the time the pulses arrive at the first repeater. Naturally, LBO depends on the distance from the first repeater. If the CSU/DSU is connected to a smart jack, the LBO can usually be set to 0 dB. Further distances from the first repeater require higher build out levels. The telco must supply this setting. Some CSU/DSUs have an automatic LBO mode that sets the build out based on the attenuation of received pulses.

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·Buy PDF of this chapter|Buy reprint rights for this chapter

Table 6-4 summarizes the most common CSU/DSU configuration options, along with available settings, common industry defaults, and additional remarks.

Table 6-4: T1 CSU/DSU configuration options
Option name	Possible settings	Common CSU/DSU default	Who sets this option?
Framing	Superframe (also SF or D4) Extended superframe (ESF)	Equipment defaults to SF; most lines are ESF Some equipment can autoselect	Telco
Line code	AMI B8ZS	Equipment defaults to AMI; lines are usually B8ZS	Telco
Transmit clock source	Network (also looped timed): transmit clock derived from receive clock DTE Internal CSU/DSU oscillator	Varies	DTE vendor; the network setting is generally safest

Additional content appearing in this section has been removed.
Purchase this book now or read it online at Safari to get the whole thing!

Content preview·

Table of Contents