UNDERSTANDING TELECOM INFRASTRUCTURE

Much like broadcast systems and facilities, telecom systems and facilities have their unique symbols, labels, and documentation conventions. The key to understanding the technology and its application is getting a grip on the language. The keys to designing, specifying, and getting what you want include a basic knowledge of the subject matter and some design tools. The design tools include a computer with word processor, spreadsheet, and drawing capability. For large complex systems and networks, a database manager becomes an important tool for thorough and accurate cost and operations analysis. In addition to these tools, design reference material in the form of recommended practices, standards, product specifications, and data sheets are a must.

Another way of defining telecom infrastructure is a technique called layering. Layering, as used in computer and communications, is best explained as a technique whereby software and digital networks are designed and built in layers. The notion is that if the operational routine or software system is built in layers and each layer interacts or interoperates with the one above and below it, then the entire system is more likely to achieve its overall goals and effectiveness.

Examples of layering include the ISO Open Systems Interconnect model, the four-layer Internet model, and the synchronous optical network/synchronous digital hierarchy (SONET/SDH) four-layer model. These models are three entirely separate models, and while they may not have been created without awareness of each other, taken literally, they don’t appear to have any relationship. Figure 1 shows these three models side by side and how they relate to each other.

Figure 1: Open Systems Interconnect Stack

It’s almost impossible to design, build, and operate a network without some form of integration of all three. For now, take notice of the matching shades of gray—dark, medium, and white. The intention is to use the darker levels to represent the bottom layers and the lighter layers to represent intermediate and top layers in each stack.

THE NATURE OF TELECOM NETWORKS

Telecom or common carrier networks have evolved over many years. The basic elements include switching, transmission, and network management. Telecom network operations run 24/7 and require very similar levels of flexibility and reliability as broadcast operations. Telecom networks are said to be ubiquitous and are built to reach the widest possible market for their services.

A key characteristic of communications networks is the concept of channelization or use of one or a group of channels between two points to support movement of information. Channelization first appeared when Edison and a now nameless technician accidentally discovered movement of sound waves over a pair of wires. That first pair of wires eventually turned into two pair to facilitate a two-way talk path. Over time through the magic of technological evolution ways and means to enable multiple channels on a single talk path, or four-wire facility were realized. For many years the jellybean of telephone technology was and still is the voice grade channel. The channel, be it wire or a virtual channel has certain capabilities and limitations bound by the laws of physics. Attempting to send a 30-Mbs payload through a 64-Kbs channel doesn’t result in any more success than an attempt to pump 30 barrels of oil per hour through a 1-inch pipe. Expecting networks to carry voice, data, and video without some way to match each with a unique part of a communications channel is the equivalent of mixing oil, water, and orange juice into the same pipe and expecting each to arrive intact at the other end. The challenge is not so much in the mixing as in the separation at the receive end.

Telecom network architecture is standards based. Most telecom network operators standardize and base their designs on a limited number of manufacturers and suppliers, generally constrained by maintaining sound, competitive procurement practices. Equipment and software suppliers active in the market generally participate in standards development and tend to comply in most, if not all, aspects of applicable standards.

Telecom networks interoperate across business entity and physical boundaries that are local, regional, national, and international.

Telecom networks are built using a layered architecture. This architecture is based on international standards and consists of a physical layer, facilities, and service layers. Although this layering characteristic is related, it should not be confused with the seven-layer ISO model commonly used in data communications and information technology documentation. It is also related and should not be confused with the four-layer approach sometimes seen or referred to in Internet or Internet protocol (IP) documentation.

One of the main objectives of this book is to bring clarity to this picture for the reader. Clarity comes from understanding a greater level of detail of the telecom entity. For purposes here, the Internet is another facilities or service layer in the overall architecture. The Internet relies on classical telecom physical or transmission layer infrastructure. It is likely to remain that way for the foreseeable future.

Many, if not most Internet people, netizens as they sometimes refer to themselves in a third-person way, lack appreciation and understanding of the physical layer or the entire classical telecom infrastructure. To many it’s old, outdated, obsolete, and subject to complete disregard, yet it is critical to successful operation of the Internet. That’s because the original Department of Defense Advanced Research Projects Agency project included an assumption that the physical layer would always be 100% available. That general approach and attitude remains today and is reflected in such things as contemporary certification training and testing by router equipment manufacturers. There has always been a similar attitude and approach in classical data communications network design and operations. The simple message in this point is ‘‘don’t forget the layer 1 and 2.’’

Another area of misunderstanding and common misuse is the term private network. Networks are made up of individual elements of hardware and software, supported and managed by a sophisticated network and equipment management infrastructure capable of monitoring, detecting, and reporting network behavior and performance outside predetermined limits. Strictly speaking, a private network is built using the same or similar kinds of equipment, software, and supporting infrastructure. It requires capital investment and ongoing operations expense. In the same context, private networks are used exclusively by their owners and don’t provide services or facilities to others. Satellite facilities are generally thought of as private networks; however, regardless of the term, they are built using shared and nonshared equipment.

Public networks, also called common carrier networks, are shared among a community of enterprise and residential customers. They operate under rules and regulations promulgated by the Federal Communications Commission (FCC) and state Public Utilities Commissions (PUCs), based on state and federal law in the United States, and similar government bodies in other countries. This includes physical aspects of orbital platforms, radio frequency spectrum, and, to a lesser degree, the equipment and services in ground station facilities.

Virtual private network (VPN) is a term used to describe a network designed and created from public network facilities and services. In some designs, the customer purchases or otherwise acquires the right to use equipment and software making up a VPN. Typically, this equipment is located on customer premises, although it may be colocated in service provider facilities with other customers or carrier owned equipment. VPNs didn’t just show up yesterday. The incarnation in the mid-1980s was software-defined network (SDN). These terms are another source of confusion. At a high level, both mean the same thing. Both evolved from private networks, which are built using point-to-point private line facilities and premises based multiplexing and switching equipment. The use of the terms virtual private and software-defined came into vogue around the time of the divestiture of AT&T and the deregulation of the long distance industry. Early examples of SDNs include AT&T and the Bell Operating Companies, who shared central office switching as an alternative to a private branch exchange and enhanced private switched communications services.

VPN is often used to describe a data network, but it is nothing more than a combination of point-to-point private leased lines and shared core routers or, in the old days, multiplexers. Some will claim that VPNs include encryption, firewalls, and so-called tunneling and that they are capable of carrying voice traffic. These claims and attributes are all true; however, the concept of shared faculties is the same as used in the SDNs that arose in the late 1970s and early 1980s to support multiple locations private dial plans for telephone service.

SATELLITE SYSTEMS AND TECHNOLOGY

A background summary on communications wouldn’t be complete without including satellite systems and technology, another unique segment of the field. The global satellite system today has evolved over more than 40 years. Satellite services are grouped into fixed satellite service (FSS) and broadcast satellite service (BSS) by the ITU. The more common informal reference for the BSS is DBS, meaning direct broadcast service. In addition to communications, satellite systems and technology provide vital weather information, mapping, location information through the global positioning system, plus many valuable services to the military.

The current DBS system is conceptually very similar to one ArthurC. Clarke described in an article written in the fall of 1945 for Wireless World. In this article, he foresaw 24-hour manned satellites being used to distribute television programs. Despite a repeated version of the concept in another publication, The Exploration of Space written in the early 1950s, the idea never gained much interest or attention.

John Pierce of AT&T Bell Labs is credited with being the first to take serious technical and financial interest in the idea. Pierce elaborated on the basic idea to the extent that the space-based platforms would perform much like a mirror and be located in medium and 24-hour orbits. He estimated the capacity of the satellite to be equivalent to 1000 simultaneous telephone calls and comparing it to the first trans-Atlantic telephone cable with a capacity of 36 simultaneous calls, arrived at a conclusion that it would cost 36 million dollars and be worth a billion.

AT&T caught the FCC by surprise in 1960 when it requested permission to launch an experimental satellite. At the time, the commission and other parts of the government simply weren’t equipped with policy and rules covering satellite communications. RCA was awarded a contract to build a medium-orbit satellite in mid-1961. Around the same time, Hughes was awarded a contract to build a high orbit, 24-hour satellite. By 1964, four medium-orbit and two high-orbit satellites had operated successfully. The Communications Satellite Act of 1962 formed the basis for Communications Satellite Corporation with an initial capitalization of 200 million dollars to build a system of several dozen medium-orbit satellites. Ultimately, COMSAT decided to build satellites for the higher geosynchronous orbit, the first of which was launched from Cape Canaveral in April 1965.

A key early broadcast event was televising part of the 1964 Tokyo Olympics. At the same time the United States was gaining this initial expertise and capability, other countries had been involved from the beginning. American companies built COMSAT’s initial satellites and launch vehicles. AT&T negotiated with Foreign PTT organizations to build earth stations and began tests and experiments aimed at providing telephone service. By the time COMSAT’s first satellite was launched and ready for service, France, the United Kingdom, Germany, Italy, Brazil, and Japan had operational earth stations. In August 1964, agreements were signed to create the International Telecommunications Satellite Organization (Intelsat).

By 1969, when Apollo 11 landed on the moon, half a billion people watched the event all over the globe through Intelsat transmission facilities. The last facilities making up the first global network were placed in service over the Indian Ocean just days before the moon landing occurred on July 20, 1969.

ABC proposed a domestic satellite system to distribute television signals in 1965, but it never gained traction. In 1972, ANIK was placed in service by Telesat Canada to serve the vast regions of the country. RCA and Western Union both launched the first domestic satellites in 1974 and 1975. AT&T launched its first domestic satellite in 1976. Satellites were intended to provide voice and data service; however, television quickly became a major user. By the end of 1976, 120 transponders were in service, each capable of 1500 telephone conversations or one TV program. Movie channels and super stations were made available to cable head ends, driving the growth in cable TV demand. During this same period, the major radio and television networks began using satellites to distribute programming to their affiliates. Satellite distribution would prove far more reliable and less expensive than terrestrial networks.

Arthur Clark’s vision of watching television from a satellite would be realized in the fall of 1994 when Hughes, RCA, and Hubbard Broadcasting launched the DirecTV transmission system. The first serious competition for cable got off the ground. A few years later, Echostar would launch its Dish Network.

A key component of satellite technology, the traveling-wave tube (TWT) was invented in England and perfected at Bell Labs. It is used to generate the signal transmitted from the ground to the satellite and back from the satellite-to-ground station receivers. Achieving adequate power level for the signal to be received by the satellite and re-transmitted back to earth required very large (100-foot diameter) dish antennas in the early uplink transmission systems. Early TWT power output levels were only approximately 1 W, but they have grown to more than 300 W. Uplink antennas approaching one tenth the size of early versions now cost around 30,000. Receiving antennas that are the size of a large pizza now enable reception of several hundred TV programs and data links to millions of businesses requiring credit card authorizations and accurate inventory tracking.

When COMSAT launched its first satellite in 1965, it provided almost 10 times the capacity of the submarine telephone cables for almost one tenth the price. Telephone service on a satellite facility suffers from the long path it must travel. In the early days, the availability of the service was its key selling point. Satellite telephone service is still the service to and between many countries today. The first fiber cable, TAT-8, was laid in the Atlantic Ocean in the mid-1980s and provided competition. Satellites are still competitive in many applications, especially point-to-multipoint service such as DBS and network distribution to affiliates.

CONSOLIDATION REDUX

From approximately 1999 through the end of 2002, telecom employees and executives saw option packages worth tens of millions of dollars, and their jobs, evaporate. Some of the most prestigious companies in the business were thrown into turmoil. Lucent Technologies, parent of Bell Labs, shrunk its global workforce from 153,000 to 35,000.

The nation’s top telecom clusters in New Jersey, San Francisco, Boston, Dallas, Atlanta, and Washington suffered as well. High-flying startups initially flush with venture capital ran out of money and simply ceased operations or declared bankruptcy.

In early 2003, a deeply divided Federal Communications Commission considered and left in place rules passed in the wake of the 1996 Telecom Legislation that were intended to foster local telephone competition. Seven original RBOCs had consolidated into four. Ameritech and Pacific Telesis became children of SBC. Bell Atlantic merged with Nynex. GTE, once independent, that is, not part of the original AT&T, or a Regional Bell, merged with Bell Atlantic to form Verizon. Ostensibly, this action was to free the Bells so they could foster growth in broadband services. Freeing the Bells ideally would have allowed them to stop making their network elements available to resellers, allow them to compete in the long distance market as well as offer Internet services, the new name for the old ‘‘data’’ services. Part of the rationalization for allowing the Bells to compete in broadband services, allows them to compete with cable companies’ Internet service offerings. However, this ideal was not to be.

The structure of the new rules removes restrictions on broadband, also known as high-speed Internet access, but not traditional telephone service, part victory and part setback because the Bells’ traditional telephone service will remain subject to regulation of state PUCs, and thereby likely to continue to require the Bells to make their network elements available for resale by competitive carriers. So it seems like the world, at least in the United States, the communications landscape will continue to include the courts, various regulatory bodies, and to quote FCC Chairman Michael Powell, ‘‘Picassoesque,’’ predicting it will lead to ‘‘legal and regulatory chaos.’’

Where is all this turmoil leading? It’s difficult as always to predict, but it’s not difficult to get an indication from a simple look at the status of the local exchange business from the perspective of the dominant and not so dominant players.

First, the dominant players are generally considered to be the ILECs. The former RBOCs have for many years prospered from voice services sold to enterprise and residential customers. For the most part, these services are delivered via single-pair copper wire. Even where the services are delivered via fiber, the nature, character, and performance of the facilities delivering the service is bound by the constraints inherent in channelized, 64-Kbs T-carrier or PDH access, switching and transport hierarchy. Unchannelized facilities are available, but are simple point-to-point private lines. Digital interface to, and transport through, ILEC network facilities is limited to T1 and DS3.

A financial glance will show suffering from declining revenue and high debt. Declining revenue from wired voice services—their bread-and-butter—is being made up by mobile or wireless voice services. Overall, opportunities for growth are constrained geographically and technically. Relief from the geographic restraint provides opportunity in the form of incremental revenue from long distance services, but it is in exchange for unbundling network elements for sale to competitive local exchange carriers (CLECs), who turn right around and bundle and sell these same network elements in the same market, depriving them of pieces of their most profitable revenue streams.

Many of the ILECs profess to be interested in broadband equipment and facilities. Establishing significant levels of service requires significant investment in new equipment. There’s been an initial ‘‘toe in the water’’ in the form of digital subscriber line (DSL) equipment acquisition and service offerings. But for anything significant such as would allow them to take market share from broadcasting, cable television, or satellite operators, the investment would have to be fiber replacement of copper wire, with IP access, switching and transport, in essence a wholesale replacement of existing infrastructure over time and that wouldn’t necessarily relieve them of their regulatory requirement to continue selling unbundled network elements (UNE).

On the less than major player side of the market, there are the CLECs mentioned previously. CLEC business strategy comes in two forms: one sells Internet access and the other buys and sells unbundled network elements mentioned previously. Others, there are a few who have invested in their own facilities in large markets where the investment level is potentially rewarding because of short transmission distance and access to the more lucrative enterprise customer. One or two of the cable television operators have installed SONET/SDH equipment on common fiber with TV service and offer voice service in the residential market. They have been moderately successful, but real returns are only incremental mostly because of the limitations imposed by voice service pricing in a monopoly market dominated by incumbents.

Many are betting voice-over IP (VOIP) will be at the roots of the next major wave of change in the industry. Some maintain that this could be the next utopian storm. And it may come at faster speed than the classical telephone world is used to enjoying. Remember the game you learned as a child whereby a series of dots were placed on paper in a matrix form and then you and another player started connecting two dots at a time with a single mark? The objective of the game was to reach a point where the last line drawn created a box into which the player could put their initial and claim ownership. The end of the game came when all boxes had been claimed, and the winner was the player who had acquired the most boxes. In the VOIP game, there’s a lot of dots out there now, including laptop personal computers (PCs) with a microphone and speaker, to say nothing of the number of desktops with circuitry and plugs, even microphones and speakers thrown away or languishing in a desk drawer. Microsoft started shipping operating system (OS) software with QOS capabilities in 2000. The next generation OS includes what they called a real-time communications client.

PC technology is pervasive or becoming that way in enterprise and consumer market places. Just connecting these dots alone could make a devastating impact on ILEC voice service business. Sooner or later, the owner of a PC will discover that many or most of the other people they talk to on the telephone can be reached by an alternate route through their PC and Internet access and wonder what else they could buy with the $30 or $40 they spend for local and long distance telephone service.

Local area network (LAN) technology and products have been as much a business necessity as telephone products and services for many years. Their utility started out as a way to share printers and later files. Somewhere along the path, the MIS department figured out they were a cheap replacement for the coaxial cable used to connect expensive terminals to mainframe computers. When the Internet came along, what did the business connect it to? For sure they didn’t connect it to the telephone system.

Cable television operators discovered cable modems could provide Internet access. This immediately made a significantly positive impact on revenue from an incremental investment, far less than the incremental investment required to offer telephone service using SONET/SDH equipment. These devices connect to a PC, not a telephone.

THE INTERNET

It’s hard to view the Internet as having been in an experimental stage. Technologically, what’s now known as the Internet is a collection of myriad techniques from other communications methods and networks. Some might contend Internet experimentation started in the 1970s and continued into the 1980s. Initially the Internet was a US Department of Defense (DOD) research project aimed at developing survivable networks. Over time, participation in the research was expanded to include academic and private research institutions. By the mid-1990s, the DOD had its survivable network. Responsibility for various administrative and technical functions previously held by the military was turned over to non-government entities.

Historically, and by example, telegrams—maybe even smoke signals or pounding on drums to send and receive coded messages—could be considered ‘‘data’’ communications. Telex service remains in use in some parts of the world. It had a counterpart in the United States called TWX. Telex service is an international service, whereas TWX is or was restricted to domestic US locations. Both services are enabled with a teletype, a machine created by combining telegraph key and typewriter technology. Telegrams are a service offering used when a teletype machine is used to create, transmit, and print a message for a third party. Telegrams ceased to be used in the early 1990s.

Telegrams, Telex, and TWX effectively died out in the 1990s. AT&T is said to have transmitted its last telegram in 1991. So the cannibalization of the telegraph by the telephone took almost a century.

It’s almost too complicated to go into, but along the way, AT&T and the RBOCs had, and continue to have, great challenges with anything outside telephone or voice services. The combined entities had deep and credible resources and capabilities without equal, except perhaps IBM. However, getting these resources to market was impossible because restrictions on AT&T equipment manufacture prohibited sale of their products to any other than the former BOCs. One result of divestiture and deregulation freed AT&T to sell computer equipment in competition with IBM and others. But coincidentally with the historic event in the telecom industry, Intel and Microsoft blindsided and blew away the dreams of the white-coat, MIS department mainframe computer industry, and those who would see AT&T as a possible competitor to IBM.

Setting all the issues of selling computer or computer-like equipment aside, the former AT&T and new entities had been and would continue to be dominant in a small and growing market for what started out as computer communications or data services. After all, AT&T long lines were first to install digital transmission and switching equipment in their network. Earlier FCC decisions had clearly and cleanly marked a dividing line between equipment and services. On the communications side a piece of equipment specifically defined as a customer services unit (CSU) faced another piece of equipment called a data services unit (DSU). The CSU was part of the computer system. The telephone company installed the DSU and included an amount in the monthly service charge. These functions are now built into equipment the customer buys and the telephone company installs at the customer’s location or nearby equipment cabinet.

Data communications has become a matter of buying compatible equipment and services. In a few short years, the Internet has taken hold across business and consumer or residential services like nothing in the past. Along with that change has come tough times for investors and employees, and opportunities for users to avail themselves of new capabilities. Now, and for the foreseeable future it’s ‘‘voice, data, and video.’’

LOCAL SERVICES DEREGULATION AND CHANGE AFTER 1996

With the breakup of the Bell System, long distance services were deregulated and began a path to a competitive marketplace. AT&T remained under greater restrictions than other common carriers. Regulations prohibited them from providing local service of any kind and they continued to be subject to tariff processes administered by state PUCs and the FCC.

The RBOCs were restricted as well. They were prohibited from offering any kind of inter-LATA service with three exceptions, the North Jersey LATA, an area on the border between Newark and New York; the South Jersey LATA between Philadelphia and Camden; and the DC LATA covering DC, and immediate surrounding areas in Maryland and Virginia.

With passage of the 1996 Telecom Legislation, the RBOCs were given a path to compete in the long distance market in return for opening up their networks to re-sale by third parties. This legislation defined and named entities in the local exchange business. The RBOCs were branded incumbent local exchange carrier (ILEC), and all others were named competitive local exchange carrier (CLEC).

LONG DISTANCE DEREGULATION AND GROWTH AFTER 1984

The battle for the long distance market really started on August 13, 1969, when the FCC voted 4 to 3 in favor of granting MCI’s application to construct a microwave radio system between Chicago and St. Louis. At the time of the vote, AT&T had a de facto monopoly on all long distance service in the United States. The vote to grant MCI a construction permit was the first signal of a breach in that monopoly. For the first time since the eve of World War I, AT&T had competition.

The breakup of the Bell System occurred on January 1, 1984, when it entered into a consent decree with the US Department of Justice. The terms of the consent decree prevented the RBOCs from entering the long distance or equipment manufacturing business. The decree also established equal access to the local exchange networks for AT&T and other common carriers such as MCI.

Before the breakup, state PUCs and the FCC controlled pricing for services based on return on invested capital. Prices for services were highly controlled and generally increased only when justified by a new or improved service resulting from a capital investment. The regulatory framework was based on intrastate services, the domain of the PUC and interstate services under the watchful eye of the FCC. Revenue funneled through the local exchange business just like the telephone calls. Intrastate revenue and interstate revenue funded the monopoly.

Part of the breakup was the establishment of more than 200 LATAs to implement the local access provisions of the consent decree. The local exchange companies could only carry traffic within the LATA, effectively maintaining their monopoly. Inter-exchange carriers carried traffic between the LATAs. Intrastate matters remained the domain of the state PUC, while interstate services continued under FCC jurisdiction.

From a practical standpoint, this new arrangement took time to understand and get used to. Each inter-exchange carrier established one or sometimes more points of presence (POP) in each LATA. These POPs continue to exist and provide access points in the LATA where they terminate and pick up traffic.

Ordering and provisioning service under the new way of doing business with the telephone company became something of a nightmare. If you were a consumer or small business requiring a single telephone line all of a sudden you were required to order local service from one source—the local exchange service provider who could not offer long distance service—and a long distance service provider who could not offer local service. No more one-stop service and, worse yet, two bills to pay at the end of the month where there used to be only one. Worse yet, those who had never experienced the joy of shopping for a telephone instrument could delay that for a while; however, avoiding paying exorbitant prices in a third monthly bill for equipment was impossible.

The other extreme represented by large businesses with multiple sites requiring service had the opposite kind of challenge. Before the break-up, many large businesses were served by a centralized group of people, typically located physically close to corporate headquarters and in many cases fully dedicated to serving the account. All of a sudden the former account management group had broken up into multiple groups, one for long distance and anywhere from one to seven or more for local service. Worse yet billing turned into a nightmare. Billing was site-specific. Corporations formerly paying a single bill for combined long distance and local service, covering hundreds and in some cases thousands of locations, had a blizzard of paper to deal with.

As an example, billing for private line services used in host-to-host or terminal-to-host data communications resulted in three bills: two for the local access or tail circuits and a third for the interexchange facility. Ordering and provisioning private line services now turned into a coordination exercise for anyone requiring such service. The model for present-day service was set to the extent that local service is ordered and provisioned separate from long distance.

A local telephone call originates and terminates in the same LATA. A private line between two points in the same LATA is local service. The same local facility or service configured to terminate in the inter-exchange POP provides access to the inter-exchange carrier. The inter-exchange carrier provides service to the distant or foreign LATA. There it is terminated in a POP and passed to the local exchange carrier providing far end access, sometimes called egress.

The LATA scheme became the basis for widespread use of the terms access and transport. Functionally, the BOCs provide access facilities, and long distance carriers, also called inter-exchange carriers, provide transport.