InformIT

Introduction to Digital Transport Systems

By and

Date: Aug 16, 2002

Sample Chapter is provided courtesy of Prentice Hall Professional.

Return to the article

The next generation of digital communication is here and taking over a network near you. Explore the history, the benefits, and the future of digital communication by examining the Synchronous Optical Network (SONET).

This chapter introduces the Synchronous Optical Network (SONET) and T1 technology. A brief history of SONET is provided, as well as the reasons that SONET came into existence. We also provide a brief history of the T1 system and make a general comparison of T1 and SONET. As a prelude to later chapters, a general description is provided of the major features of SONET.

What are SONET and T1

Digital carrier systems, such as the well-known T1 technology, have served the telecommunications industry well for over 40 years, and they shall continue to do so for quite some time. T1 was first installed in 1962 to provide a high-speed (1.544 Mbit/s) digital carrier system for voice traffic. It was modified later to support data and video applications.

In terms of communications technology, 1962 is a very long time ago. Since this date, extraordinary progress has been made in the fields of computers and communications. Many of the technical underpinnings of SONET exploit this new technology.

We have cited T1 in a previous paragraph to help explain the nature of SONET. T1 and associated systems (such as T3 and similar technology in other parts of the world) are first-generation digital transport systems. SONET is a second-generation digital transport system. Like T1, its purpose is to transport, multiplex, and switch digital signals representing voice, video, and data traffic to and from users' applications.

However, T1 and SONET differ significantly in how they accomplish these functions. Many facets of the T1 architecture are based on technology that is over four decades old. In contrast to T1, the SONET architecture is based on the technology of today. With this brief comparison in mind, let us take a look at how SONET came into existence and then examine some of the major features of SONET.

The Development of SONET

Some people view SONET as a new technology, and it is only in the last decade that SONET has been deployed extensively. However, SONET did not just appear suddenly on the scene. Extensive research has been underway for well over a decade on many of the features that are found in SONET. One notable achievement began in 1984. It focused on the efforts of several standards groups and vendors to develop optical transmission standards for what is known as the mid-span meet (also known as transverse compatibility). The goal was to publish a specification that would allow different vendors' equipment to interoperate with each other at the fiber level.

In addition, due to the breakup of the Bell System in 1984, there were no standards developed beyond T3 technology. Prior to the divestiture, all equipment was built by AT&T's manufacturing arm, Western Electric (WECO), which ensured that there would be no compatibility problems in any network components.

After the breakup, there was little incentive for the other carriers (such as MCI and Sprint) to purchase AT&T-based equipment. Indeed, there was no incentive to purchase AT&T equipment, since AT&T, MCI, and Sprint had begun competing with each other for long distance services. This situation led to the rapid growth of alternate equipment vendors (such as Nortel Networks), who were developing advanced digital switching technologies.

The 1984 divestiture paved the way for alternate long distance carriers through the equal access ruling. The alternate carriers were given equal access to the local exchange carrier (LEC) infrastructure and connections to AT&T for end-to-end long distance service. The LEC could connect to MCI, Sprint, and others through their switching facilities at an interface in the LEC or long distance carrier offices called the point of presence (POP).

During this time, higher capacity schemes beyond T3 became proprietary, creating serious compatibility problems for network operators who purchased equipment from different manufacturers. In addition, the early 1980s witnessed the proliferation of incompatible and competing optical fiber specifications.

Precursors to SONET

We interrupt the discussion on divestiture to explain some of the technology that was being developed during the early 1980s. A landmark project that contributed to SONET was Metrobus, an optical communications system developed at AT&T's Bell Labs in the early 1980s. Its name was derived from its purpose: It was situated in a metropolitan area to serve as a high-speed optical transport network.

Metrobus demonstrated the feasibility of several new techniques that found their way into SONET. (They are explained in this chap- ter and subsequent chapters.) Among the more notable features were (a) single-step multiplexing, (b) synchronous timing, (c) extensive overhead for network management, (d) accessing low level signals directly, (e) point-to-multipoint multiplexing, and (f) the employment of multimegabit media for achieving high bandwidth network transmission capacity (of approximately 150 Mbit/s).

This latter decision along with the ensuing research and testing was important, because a 150 Mbit/s signal rate can accommodate voice, video, and data signals, as well as compressed high definition television (HDTV). Moreover, these techniques permitted the use of relatively inexpensive graded-index multimode fibers instead of the more expensive single mode fibers, although single mode fiber is now the preferred media for SONET.

The various standards groups began the work on SONET after MCI send a request to them to establish standards for the mid-span meet. The SONET specifications were developed in the early 1980s, and Bellcore submitted its proposals to the American National Standards Institute (ANSI) T1X1 Committee in early 1985,1 based on a 50.688 Mbit/s transfer rate. The initial SONET work did not arouse much interest until the Metrobus activity became recognized.

Later, using the innovative features of Metrobus, the SONET designers made modifications to the original SONET proposal, principally in the size of the frame and the manner in which T1 signals were mapped into the SONET frame.

From 1984 to 1986, various alternatives were considered by the ANSI T1 Committee, who settled on what became known as the synchronous transport signal number one (STS-1) rate as a base standard. Finally, in 1987, the ANSI T1X1 committee published a draft document on SONET.

Participation by ITU-T

During this time, the international standards body now known as the International Telecommunication Union-Telecommunication Standardization Sector (ITU-T) had rejected the STS-1 rate as a base rate in favor of a base rate of 155.520 Mbit/s. For a while, it appeared that the North American and European approaches might not converge, but the SONET frame syntax and structure were altered one more time to a rate of 51.84 Mbit/s which permitted this rate to be multiplexed (concatenated) by an integer of three to the European preference of 155.52 Mbit/s. This work has resulted in almost complete compatibility between the North American and ITU-T approaches. The ITU-T Recommendations are now considered the "official" standards and are collectively called the Synchronous Digital Hierarchy (SDH).

Once the major aspects of the standards were in place, vendors and manufacturers began to develop SONET and SDH equipment and software. These efforts came to fruition in the early 1990s and, as of this writing, SONET and SDH have been deployed throughout the United States and other parts of the world.

Key ITU-T Documents

Listed below are some of the most commonly cited SDH standards available from the ITU-T.2

Role of ANSI and Key Standards documents

Today, ANSI coordinates and approves the SONET standards. The standards are developed by the T1 committee. T1X1 and T1M1 are the primary T1 Technical Subcommittees responsible for SONET. T1X1 deals with the digital hierarchy (shown in Figure 1–5) and synchronization. T1M1 deals with internetworking operations, administration, maintenance, and provisioning (OAM&P). Listed below are some of the most commonly cited SONET standards available from ANSI.3

The Network and Services Integration Forum (NSIF)

In order to assist in the SONET standards process, the Network and Services Integration Forum (NSIF) was formed to provide an open industry forum for the discussion and resolution of multiple technology integration and SONET interoperability issues.4 Its goal is to enable the delivery of services across a set of networks from different vendors and different network operators. NSIF coordinates with the appropriate standards groups, such as ANSI and the ITU-T, as required by a specific issue.

NSIF is a nonprofit membership organization comprised of equipment vendors, service providers, and other industry players who cooperatively develop end-to-end multitechnology service delivery capabilities based on industry and international standards.

SONET and T1

While the primary focus of this book is on SONET, we devote considerable coverage to T1 technology. With this in mind, we will use this section to introduce T1 and compare the T1 and SONET technologies.

Comparison of SONET and T1

Figure 1–1 shows the placement of T1 and SONET in a communications network. Currently, several configurations exist in commercial systems. SONET may operate end-to-end or it may act as the transport system within the network carrying T1 traffic through the network between customer premises equipment (CPEs). Of course, in some networks SONET has not been implemented so T1/T3 may operate within the network. Another widely used placement is the installation of fractional T1 (FT1) at the CPE switch, which then multiplexes user payloads into a T1 or T3 transport frame. As you can see from Figure 1–1, SONET can operate end-to-end or can function as a backbone technology, operating within the network.

Figure 6-1Figure 1–1 T1 and SONET.

Figure 1–2 shows the relationship of the Open System Interconnection Model (OSI) layers to the layers of T1 and SONET. The first observation is that T1 and SONET operate at the physical layer of the OSI Model. As such, they are concerned with the physical generation of the signals at a transmitting machine and their correct reception and detection at a receiving machine. While this statement may imply that these technologies are rather simple, we shall see that the opposite is true, especially regarding SONET.

Figure 6-2Figure 1–2 Comparison of the layers in regard to OSI.

The layers operating above the physical layer are quite varied. Some products run the Internet protocols in these layers, others run SS7, still others run vendor-specific protocols, such as IBM's Systems Network Architecture (SNA). Later chapters explore the relationships of these layers and SONET.

Features of SONET and T1

This section explains why SONET differs from other digital transport systems. First, SONET is an optical-based carrier (transport) network utilizing synchronous operations between the network components/nodes, such as multiplexers, terminals, and switches. SONET's high speeds (some systems operate at the gigabit rate) rely on high-capacity fiber. Much of the T1 technology is geared toward the use of copper (twisted pairs) media and operates at more modest transmission rates.

As just stated, the SONET network nodes are synchronized with each other through very accurate clocking operations, ensuring that traffic is not "damaged," or lost due to clocking inaccuracies. T1 clocking systems are very accurate, but are not as reliable as SONET systems. Later discussions explain how synchronous networks experience fewer errors than the older asynchronous networks and provide much better techniques for managing payloads.

SONET is quite robust and provides high availability with self-healing topologies. In the event a SONET link is lost due to node or fiber failure, SONET can recover by diverting the traffic to back-up facilities. Most T1 systems can be configured for backup, but "robustness" is not an inherent part of the original design of the T1 architecture.

SONET and its ITU-T counterpart SDH are international standards. As such, SONET paves the way for heterogeneous, multivendor systems to operate without conversions between them (with some exceptions).5

SONET uses powerful, yet simple, multiplexing and demultiplexing capabilities. Unlike T1, SONET gives the network node a direct access to low-rate multiplexed signals, without the need to demultiplex the signals back to the original form. In other words, the payloads residing inside a SONET signal are directly available to a SONET node.

Because of these capabilities, SONET efficiently combines, consolidates, and segregates traffic from different locations through one machine. This concept, known as grooming, eliminates back hauling6 and other inefficient techniques currently being used in transport networks. In older transport network configurations, grooming can eliminate back hauling, but it requires expensive configurations (such as back-to-back multiplexers that are connected with cables, panels, or electronic cross-connect equipment).

SONET provides extensive operations, administration, maintenance, and provisioning (OAM&P) services to the network user and administrator. Indeed, about 4% of the bandwidth in a SONET network is reserved for OAM&P. A T1 system only allows one bit per 193 bits for OAM&P. With this comparison in mind, it is easy to conclude that SONET has the capability for more extensive and powerful network management operations than T1.

Finally, like T1, SONET employs digital transmission schemes. Thus, the traffic is relatively immune to noise and other impairments on the communications channel. Of course, with the use of optical fibers, random errors on the channel are very rare.

Synchronous Networks

One of the most attractive aspects of SONET deals with how network components send and receive traffic to and from each other. The original, first-generation digital transport networks were designed to work as asynchronous (or more accurately, nearly synchronous) systems. With this approach, each device in the network runs with its own clock, or devices may be clocked from more than one source. That is, the clocks are not synchronized from a central source.6

The purpose of the terminal clock is to locate precisely the digital 1s and 0s in the incoming data stream on the link attached to the terminal—a very important operation in a digital network. Obviously, if bits are lost in certain user traffic (user traffic is called payload in this book) then the traffic may be unintelligible to the receiver. Equally important, the loss of bits or the inability to locate them accurately can cause further synchronization problems. When this situation occurs, the receiver usually does not deliver the traffic to the end user because it is simpler to discard the traffic than to initiate retransmission efforts.

To give the reader an idea of how precise the timing must be, consider a T1 system that operates at a modest 1.544 Mbit/s. Obviously, a receiver must be able to detect each bit as it "shows itself" at the link interface at the receiving machine. Each bit is only 648 ns in duration (1 sec/1544000 = .000000648). This means that the receiver's clock must be aligned accurately with the transmitter's clock.

Because a sender's clock may run independently of the receiver's clock in an asynchronous network, large variations can occur between the sender's clock (machine 1) and the rate at which the bits are received and then transmitted by the receiver's clock (machine 2). The problem is not at the receiving link at machine 2, since machine 2 can "lock" onto machine 1's incoming signal and accept the traffic. In this regard, machine 2 extracts the clock from machine 1's signal.

The problem occurs when machine 2 then prepares that traffic for transmission onto the next outgoing link. If it is using its own clock, it usually varies from the rate that was received from machine 1. These different timing operations can create a big headache for the network administrator. For example, experience has demonstrated that a T3 signal may experience a variation of up to 1789 bit/s for a 44.736 Mbit/s signal in a network that does not have precise and accurate timing.

The Perils of Bit Stuffing

Moreover, T1 signals such as DS1s are multiplexed in stages up to DS2, DS3, etc., and extra bits are added to the stream of data to account for timing variations in each stream. The process is called bit stuffing. The lower level signals, such as DS1, are not accessible nor visible at the higher rates. Consequently, the original stream of traffic must be demultiplexed if these signals are to be accessed. The demultiplexing process is very expensive and adds delay and overhead to the network.

SONET Timing

SONET is based on synchronous transmission, meaning the average frequency of all the clocks in the network are the same (synchronous) or nearly the same (plesiochronous). As a result of this approach, the clocks are referenced to a stable reference point. Therefore, the need to align the data streams directly is less necessary. As we stated earlier, the user payloads, such as DS1, are directly accessible so demultiplexing is not necessary to access the bit streams. Also, the synchronous signals can be stacked together without bit stuffing.

The Benefits of Byte Alignment

Byte multiplexing (also called octet multiplexing) is more efficient and less error-prone than the bit multiplexing operation explained earlier. Most hardware and software today are designed to process data in chunks of eight bits, often called byte-aligned processing. In addition, bit aligned processing is more error-prone than its byte-aligned counterpart, because of the use of smaller buffers and shorter timing increments. There is less tolerance for errors in the bit aligned operation (and we show examples of bit processing in Chapters 3 and 4).

SONET requires byte alignment operations, and any timing adjustments that are performed in the SONET network are done on a byte basis, not on a bit basis.

Floating Payloads

Another major aspect of synchronous systems (in general), and SONET (specifically), pertains to how payload, such as DS1 or DS3 signals, is inserted into the SONET channel. For those situations in which the reference clocking signal may vary (even if only slightly), SONET uses pointers to allow the payload streams to "float" within the payload envelope (the term envelope is used to describe the SONET signal on the channel; the term frame is also used). Indeed, synchronous clocking is the key to pointers; it allows a flexible allocation and alignment of the payload within the transmission envelope. Thus, SONET's payload is called a synchronous payload envelope (SPE).

The concept of a synchronous system is elegantly simple. By holding specific bits in a silicon memory buffer for a defined and predictable period of time, it is possible to move information from one part of a payload envelope to another part. It also allows a system to know where the bits are located at all times. Of course, this idea is "old hat" to software engineers, but it is a different way of thinking for other designers. As one person has put it, "Since the bits are lined up in time, we now know where they are in both time and space. So, in a sense, we can now move information in four dimensions, instead of the usual three."

The U.S. implementation of SONET uses a central clocking source—for example, from a telephone company's end office. This office must use an accurate clocking source known as stratum 3. Stratum 3 clocking requires an accuracy of 1.6 parts in 1 billion elements. Chapter 3 provides more detailed information on synchronization and clocking operations as well as the accuracy levels of the stratum 1, 2, 3, and 4 clocks.

Table 1–1 Typical SONET payloads.

Type

Digital Bit Rate

Voice Circuits

T–1

DS3

System Name

North American multiplexing hierarchy

DS1

1.544 Mbit/s

24

1

 

DS1C

3.152 Mbit/s

48

2

 

DS2

6.312 Mbit/s

96

4

 

DS3

44.736 Mbit/s

672

28

1

 

DS4

274.176 Mbit/s

4032

168

6

 

European multiplexing hierarchy

E1

2.048 Mbit/s

30

 

 

M1

E2

8.448 Mbit/s

120

 

 

M2

E3

34.368 Mbit/s

480

 

 

M3

E4

139.264 Mbit/s

1920

 

 

M4

E5

565.148 Mbit/s

7680

 

 

M5

Japanese multiplexing hierarchy

1

1.544 Mbit/s

24

 

 

F1

2

6.312 Mbit/s

96

 

 

F6M

3

34.064 Mbit/s

480

 

 

F32M

4

97.728 Mbit/s

1440

 

 

F100M

5

397.20 Mbit/s

5760

 

 

F400M

6

1588.80 Mbit/s

23040

 

 

F4.6G


Payloads and Envelopes

SONET is designed to support a wide variety of payloads. Table 1–1. summarizes some typical payloads of existing technologies. The SONET node accepts these payloads and multiplexes them into a SONET envelope. These payloads are called virtual tributaries (VTs) in North America and virtual containers (VCs) in SDH. Later chapters will explain how SONET manages these payloads.

As you can see, the first-generation digital carrier systems are not the same across different geographical regions of the world. Shortly after the inception of T1 in North America, the ITU-T published the E1 standards, which were implemented in Europe, and Japan followed with a hierarchy that was similar to the North American specifications at the lower speeds (but not at the higher speeds).

Optical Fiber—The Bedrock for SONET

It is likely you understand and appreciate the advantages of using optical fiber as the transmission medium for a telecommunications system. This section summarizes the major aspects of optical fiber, and Appendix A explains these aspects in more detail. Optical fiber is widely used as the physical medium in SONET for the following reasons:

Figure 1–3 shows the progress made in transmission capacity of single-mode fibers since 1980 [CHAR99]. The top line represents experimental systems, and the bottom line represents commercial systems. The commercial results have lagged behind experimental results by about six years. The dramatic growth in the experimental capacity was due to improved laboratory techniques and the progress made in optical signal management.

Figure 6-3Figure 1–3 Progress in optical fiber technology.

While optical fiber is an excellent medium for the transport of SONET traffic, it is by no means the only alternative. Other cable-based systems, such as twisted pair, and wireless systems, such as microwave, are applicable as well.

Typical SONET Topology

Figure 1–4 shows a typical topology for a SONET network. This topology is a dual ring. Each ring is an optical fiber cable. One ring is the working facility. The other ring is the protection facility, which acts as a standby in the event of fiber or system failure on the working facility.

Figure 6-4Figure 1–4 SONET topology.

End-user devices operating on LANs and digital transport systems (such as DS1, E1, etc.) are attached to the network through a SONET service adapter. This service adapter is also called an access node, a terminal, or a terminal multiplexer. This machine is responsible for supporting the end-user interface by sending and receiving traffic from LANs, DS1, DS3, E1, ATM nodes, etc. It is really a concentrator at the sending site because it consolidates multiple user traffic into a payload envelope for transport onto the SONET network. It performs a complementary, yet opposite, service at the receiving site.

The user signals (such as T1, E1, and ATM cells) are converted (mapped) into a standard format called the synchronous transport signal (STS), which is the basic building block of the SONET multiplexing hierarchy. The STS signal is an electrical signal. The notation STS-n means that the service adapter can multiplex the STS signal into higher integer multiples of the base rate. The base rate is 51.84 Mbit/s in North America and 155.520 Mbit/s in Europe. Therefore, from the perspective of a SONET terminal, the SDH base rate in Europe is an STS-3 multiplexed signal (51.84 x 3 = 155.520 Mbit/s).

The terminal/service adapter (access node) shown in Figure 1–4 is implemented as the end-user interface machine, or as an add-drop multiplexer (ADM). The ADM implementation multiplexes various STS input streams onto optical fiber channels. The optical fiber channels are now called the optical carrier signal and designated with the notation OC-n, where n represents the multiplexing integer. OC-n streams are demultiplexed as well as multiplexed with the ADM.

The term add-drop means that the machine can add or drop payload onto one of the fiber links. Remaining traffic that was not dropped passes straight through the multiplexer without additional processing.

The digital cross-connect (DCS) machine usually acts as a hub in the SONET network. It can not only add and drop payload, but it can also operate with different carrier rates, such as DS1, OC-n, E1, etc. The DCS can make two-way cross-connections between the payload and can consolidate and separate different types of payloads.

The DCS is designed to eliminate devices called back-to-back multiplexers. As we learned earlier, these devices contain a plethora of cables, jumpers, and intermediate distribution frames. SONET does not need all these physical components because cross-connection operations are performed by hardware and software.

The topology can be set up either as a ring or as a point-to-point system. In most networks, the ring is a dual ring, operating with two or more optical fibers. As noted, the structure of the dual ring topology permits the network to recover automatically from failures on the channels and in the channel/machine interfaces. This is known as a self-healing ring and is explained in later chapters.

Present Transport Systems and SONET

The present digital transport carrier system varies in the different geographical regions of the world. The structure is different in Japan than it is in North America, which is itself different than the structure in Europe. This disparate approach is complex and makes the interworking of the systems difficult and expensive. Moreover, it means that companies that build hardware and software for carrier systems must implement multiple commercial platforms for what could be one technology.

While SONET does not ensure equipment compatibility, it does provide a basis for vendors to build worldwide standards. Moreover, as shown with the shaded area in Figure 1–5, SONET is backwards compatible, in that it supports the current transport carriers' asynchronous systems in North America, Europe, and Japan. This feature is quite important because it allows different digital signals and hierarchies to operate with a common transport system, SONET. By the way, don't be concerned with all the details shown in Figure 1–5; they are explained later.

Figure 6-5Figure 1–5 SONET support for current technologies. Note: Unless noted otherwise, speeds in Mbit/s.

Clarification of Terms

Before we proceed into a more detailed discussion of the subject matter, it is appropriate to pause and define some terms that will be used in subsequent chapters. In the early 1980s, AT&T introduced the digital access and cross-connect system (DACS) as a major enhancement to its digital transport system products. It is also called the digital cross-connect system, or DCS. The subject of DCS is introduced here in order to explain several concepts that are central to subsequent chapters.

The original DACSs were considered complex implementations of microelectronics technology. They had a RAM (random access memory) of 256 words! At that time, they were considered very esoteric machines.

Figure 1–6 shows some of the major operations of a DCS, which uses a combination of time-division and space-division switching techniques. The original DACS terminated up to 127 DS1 signals (3048 DS0 channels) and provided up to 1524 cross-connections.

Figure 6-6Figure 1–6 Terms and concepts.

The DCS permits the assignment and redistribution of DS0 channels for drop-and-insert (also known as ADM in today's technology) services. As Figure 1–6(a) shows, the operation provides for the distribution of traffic to nodes reachable by the DCS (the drop operation). It also allows a node to send traffic to the DCS for delivery to another node in the network (the insert operation). Figure 1–6(b) shows how the DCS performs back hauling (sending traffic downstream and returning it back), which does not require back-to-back channel banks (or the conversion of the digital signals to analog and back to digital again). A DCS also performs groom-and-fill operations (Figure 1–6(c)), a technique in which the machine accepts input from two or more low-speed lines and multiplexes these signals into a higher speed line.

Figure 1–7 shows some examples of equipment and configurations that exist at the customer premises equipment and the telephone central office. Be aware that a wide variety of options are available and these examples are only a sampling of possible services and arrangements.

Figure 6-7Figure 1–7 Examples of configurations.

Figure 1–7(a) shows a voice interface into a channel bank, which converts the analog signal into a digital signal with a CODEC (a coder/decoder). The CO then multiplexes multiple digital signals together for transmission into the network.

Figures 1–7(b) through 1–7(d) show how channel service units (CSUs), digital service units (DSUs), and office channel units (OCUs) may be employed. Before the divestiture of AT&T in 1984, the CSU and and DSU were in separate boxes. They are responsible for coding the user traffic into self-clocking signals and performing a variety of testing and loop-back functions.

In the 1970s and early 1980s, the customer was provided an interface into a digital channel with a Western Electric 500A, a CSU and DSU, or a combination of the two (CSU/DSU). The DSU converts the user equipment signals into signals that are more efficient for use in a digital network. The DSU also performs clocking and signal regeneration on the channel. The CSU performs functions such as line conditioning (equalization), which keeps the signal's performance consistent across the channel bandwidth; signal reshaping, which reconstitutes the binary pulse stream; and loop-back testing, which entails the transmission of test signals between the CSU and the network carrier's office channel unit (OCU).

Summary

Modern telecommunications and applications need increased carrier capacity for wide area transport service. In the past, the T1 system has provided this service and will continue to do so for many years. SONET represents second-generation digital carrier technology, which will eventually supplant the T1 technology. Fortunately, SONET is designed to support the T1 technology, which facilitates its placement into the pervasive and ubiquitous infrastructure.

800 East 96th Street, Indianapolis, Indiana 46240