Telecommunications and Networking PDF

CHAPTER 3 Telecommunications and Networking This chapter is the second of a trio of chapters devoted to the building blocks of information technology. So far we have considered hardware and software, but there are two more critical building blocks to go. The hardwar...

CHAPTER 3 Telecommunications and Networking This chapter is the second of a trio of chapters devoted to the building blocks of information technology. So far we have considered hardware and software, but there are two more critical building blocks to go. The hardware and software must have data to be processed to produce useful results, and the data resource is the focus of Chapter 4. If every computer were a stand-alone unit with no connection to other computers, then hardware and software and data would be the end of the story as far as computers are concerned. In fact, until about 40 years ago, that was the end of the story. Today, however, virtually all computers of all sizes communicate directly with other computers by means of an incredible variety of networks. For computers in organizations, these networks include intraorganizational local area networks (LANs), backbone networks, and wide area networks (WANs), as well as the worldwide Internet. For home computers, the most important network is the Internet. In addition to computer (or data) communications, today’s organizations also depend heavily on voice (telephone) and image (video and facsimile) communication. This chapter explores the increasingly important topic of telecommunications and networking. This chapter’s goal is to cover the telecommunications and networking technology that you as a business manager need to know. You need to understand the roles and general capabilities of various types of transmission media and networks, but you do not need to know all the technical details. You need to know the important terminology and concepts relating to telecommunications and networking. Most important, you need to understand the interrelationships between hardware, software, and telecommunications and networking so that you can use the full gamut of information technology to increase your productivity and your organization’s effectiveness. Change is everywhere in the information technology domain, but nowhere is this change more evident and more dramatic than in the realm of telecommunications and networking. A communications revolution is taking place that directly or indirectly affects the job of every manager, and the primary catalyst is the Internet and the World Wide Web (an application that runs on the Internet). The breakup of American Telephone & Telegraph (AT&T) in 1984 created an environment in which a large number of firms competed to develop and market telecommunications equipment and services. Partially because of this increased competition, innovation in the telecommunications and networking arena has been at an all-time high. Digital networks, fiber-optic cabling, cellular telephones, the ability to send both voice and data over the same wires at the same time, and wireless networks have contributed to the revolution. At the same time, most large U.S. businesses have restructured internally to reduce layers of middle management and create a leaner organization. They have also decentralized operations in order to respond more quickly to market opportunities and competitors’ actions and have created cross-functional teams to improve business processes and carry out projects. The net result of these internal changes is that communication has become more important than ever for the remaining, often geographically dispersed, managers. They need rapid, reliable voice and data communication with other parts of the company and with suppli-ers and customers. Small 60  Chapter 3 Telecommunications and Networking 61 Networks will Change Everything In the early 1990s, Paul Saffo, a fellow at the Institute for the Future, developed a fascinating set of forecasts about the effect of information technologies on the way we would work, play, and conduct business in the years to come. So far, his projections have been right on. “The short answer is that networks will change everything,” said Saffo. “In the next five years, networks will be supporting a shift to business teams from individuals as the basic unit of corporate productivity. In the 10-year time frame, we’ll see changing organizational structures. In 20 to 30 years, we’ll see a shift so fundamental, it will mean the end of the corporation as we know it.” According to Saffo, organizations have started down the path to a pervasive interconnectivity of workstations that will result in an entirely new “virtual” corporate structure. [Based on Wylie, 1993] businesses are also more dependent upon communication another under control of a LAN software package called than ever before, and developments such as LANs, cellular a server (or network) operating system. When a particular telephones, and increased functionality of the public wired user wants to print a color brochure or a color transparency, telephone network have helped fill this need. Internal it is sent electronically from the user’s machine to the needs and external competition and innovation combined network printer. to create a latter-twentieth-century communications Sharing resources is also important for larger revolution that is continuing into the new millennium. The computers. It is quite common for mainframes or aim of this chapter is to help you become a knowledgeable midrange computers to share magnetic disk devices participant in the communications revolution. and very high-speed printers. Further, wide area networks (WANs) permit the sharing of very expensive resources such as supercomputers. The National Science THE NEED FOR NETWORKING Foundation has funded several national supercomputer Let us be more precise in justifying the need for networking centers across the United States, and researchers from among computers and computer-related devices such as other universities and research laboratories are able to printers. Why do managers or other professionals working share these giant machines by going through their local at microcomputers need to be connected to a network? Why computer network into a national high-speed backbone are small computers often connected to larger machines? network such as Internet2 (more on this network later in Why are laser printers often attached to a LAN? Why is it the chapter). critical for most businesses to be connected to the Internet? In our judgment, there are five primary reasons for Sharing of Data networking. Even more important than the sharing of technology resources is the sharing of data. Either a LAN or a WAN Sharing of Technology Resources permits users on the network to get data (if they are author- Networking permits the sharing of critical (and often ized to do so) from other points, called nodes, on the expensive) technology resources among the various users network. It is very important, for example, for managers to (machines) on the network. For example, by putting all of be able to retrieve overall corporate sales forecasts from the PCs in an office on a LAN, the users can share a corporate databases to use in developing spreadsheets to variety of resources, such as a high-speed color printer project future activity in their departments. In order to that is a part of the network. The users can also share satisfy customers, automobile dealers need to be able software that is electronically stored on a file server to locate particular vehicle models and colors with specific (another computer designated for that particular equipment installed. Managers at various points in a purpose). All these devices are connected by wiring (or a supply chain need to have accurate, up-to-date data on wireless network) and are able to communicate with one inventory levels and locations. Accountants at corporate 62 Part I Information Technology headquarters need to be able to retrieve summary data on Enhanced Communications sales and expenses from each of the company’s divisional Networks enhance the communications process within an computer centers. The chief executive officer, using an organization (and between organizations) in many important executive information system (see Chapter 6), needs to be ways. The telephone network has long been a primary able to access up-to-the-minute data on business trends means of communication within and between organizations. from the corporate network. In some instances, data may Electronic mail over the corporate computer network has be retrieved from a commercial, public database external become a mainstay of communication in most major organ- to the firm, such as LexisNexis and Dow Jones Newswires. izations in the past two decades, and the development of the Of course, the ultimate sharing of data is now occur- Internet has extended the reach of these electronic mail sys- ring via the World Wide Web on the Internet. By conser- tems around the world. Electronic bulletin boards (including vative estimates, there are now at least 1.7 billion users internal, regional, and national bulletin boards), blogs, and of the Web at sites around the world, and this number mass electronic mailing lists for people with common inter- continues to grow rapidly. Each of these users has easy ests permit multiparty asynchronous communication on an (and often free) access to an incredible array of informa- incredible array of topics. Instant messaging permits syn- tion on any topic. The user begins by using a search engine chronous text communication over the Internet. And video such as Google or a favorite reference site and then communication, especially videoconferencing, provides a follows hypertext-based links to seek out the desired data. richer medium to permit more effective communication. In short, the Web has created a new and exciting way of Direct data communication links between a company sharing data. and its suppliers or customers, or both, have been successful- ly used to give the company a strategic advantage. The Distributed Data Processing and SABRE airline reservation system is a classic example of a Client/Server Systems strategic information system that depends upon communica- tion provided through a network. Recent developments to be With distributed data processing, the processing power discussed later in this chapter—such as DSL and IP is distributed to multiple computers at multiple sites, telephony (voice over IP)—permit both voice and data com- which are then tied together via telecommunications lines. munications to occur over the same telecommunications line Client/server systems are a variant of distributed systems at the same time. Starting with “plain old telephone service” in which the processing power is distributed between a (POTS) networks and continuing with today’s LANs, WANs, central server system, such as a midrange computer or a and the Internet, networks have enhanced the communication mainframe, and a number of client computers, which are process for individuals and organizations. usually desktop microcomputers. Distributed and client/server systems tend to reduce computing costs because of their reliance on more cost-effective microcomputers and Marketing Outreach workstations. In the last 20 years, the Internet has become an important There are many examples of distributed systems. marketing channel for a wide variety of businesses. One is the use of laptop computers by a company’s sales Marketing is communication, of course, but it is a very force, where orders and sales data are transmitted over the specialized type of communication. Most midsized and Internet (using a virtual private network, to be discussed larger business firms have a major presence on the World later in this chapter) to the corporate computer center. A Wide Web, with extensive Web sites providing information second example is the use of a client/server application for on the firms’ products and services and, in many cases, an general ledger accounting, with desktop microcomputers online ordering capability. Many smaller firms are also as the clients and a high-powered workstation as the server. using the Web for marketing outreach, perhaps by creating In most cases, such a package is implemented over a LAN a Yahoo! store. Chapter 7 will consider the wide variety of in a single building or a cluster of buildings (a campus). marketing activities undertaken on the World Wide Web. A third example, also a client/server system, involves the creation of a commercial real estate database on a server located at the real estate firm’s main office. The client AN OVERVIEW OF TELECOMMUNICATIONS machines are microcomputers located in the firm’s branch AND NETWORKING offices or customer offices, with the clients and server linked via the Internet. In any case, it is the existence of a Networking—the electronic linking of geographically telecommunications network that makes distributed data dispersed devices—is critical for modern organizations. processing a feasible and attractive arrangement. To participate effectively in the ongoing communications  Chapter 3 Telecommunications and Networking 63 revolution, managers need to have a rudimentary under- Editorial involves checking for errors and putting standing of the various telecommunications and network- the communication into a standardized format, and ing options available to their organizations. conversion includes any necessary changes in the coding The prefix tele- simply means operating at a distance. system or the transmission speed when moving from one Therefore telecommunications is communications at a device on the network to another. In networks where al- distance. There are a number of other terms or abbrevia- ternative paths are possible between the source and the tions that are used almost interchangeably with telecommu- destination of a communication (particularly WANs and nications: data communications, datacom, teleprocessing, the Internet), routing—choosing the most efficient telecom, and networking. We prefer telecommunications path—is an important task. Closely related to the pro- because it is the broadest of these similar terms. It includes cessing function is network control, which includes both voice (telephone) and data communications (including keeping track of the status of various elements of the text and image). Teleprocessing means the computer system (e.g., which elements are busy or out of service) processing is taking place at a distance from where the data and, for some types of networks, checking each worksta- originate, which obviously requires telecommunications. tion periodically to see if it has a communication to Networking is the electronic linking required to accomplish send. A not-so-obvious but critical function is the provi- telecommunications. sion of an interface between the network and the user; One might think that only a wire or a wireless sig- hopefully this interface will make it easy and efficient nal is needed for telecommunications, but it is much for a manager or any other network user to send a com- more complex than that! To begin a detailed considera- munication. The next major section explores the variety tion of telecommunications, first consider the primary of ways in which the functions listed in Table 3.1 can functions performed by a telecommunications network, be delivered. as listed in Table 3.1. The most obvious of these func- tions is the transmission of voice or data, or both, using the network and the underlying media. The processing KEY ELEMENTS OF TELECOMMUNICATIONS involves making sure that an error-free message or data AND NETWORKING packet gets to the right destination. Subfunctions of We believe that you as a business manager need to processing include editorial, conversion, and routing. understand certain key elements about telecommunica- tions and networking to participate effectively in the communications revolution—to know what the options are for the business systems you need. These key ele- TABLE 3.1 Functions of a Telecommunications ments include certain underlying basic ideas, such as Network analog versus digital signals and switched versus pri- Function Brief Description vate lines; the variety of transmission media available; the topology (or possible arrangements) of networks; Transmission Movement of voice and/or data using network and underlying media the various types of networks, including LANs and WANs; and the network protocols employed on these Processing Ensuring that error-free communication networks. This section will be somewhat technical, but gets to right destination we will do our best to make it understandable. Above all Editorial Checking for errors and putting the details, we want you to keep sight of the big picture communication into standardized of telecommunications. format Conversion Changing coding system or speed Analog and Digital Signals when moving from one device to another Perhaps the most basic idea about telecommunications is Routing Choosing most efficient path when that the electronic signals sent on a network may be either multiple paths are available analog or digital, depending on the type of network. Historically, the telephone network has been an analog Network control Keeping track of status of network elements and checking to see if network, with voice messages sent over the network by communications are ready to be sent having some physical quantity (e.g., voltage) continuously vary as a function of time. This analog signal worked fine Interface Handling interactions between users for voice transmission because it required the significant and the network variations provided by an analog signal (corresponding to 64 Part I Information Technology variations in human speech characteristics) and was insen- advantage to transmitting voice signals over a digital sitive to minor degradations in the signal quality. On the network. Digital voice transmission can provide higher- other hand, computer data consist of a string of binary quality transmission—less noise on the line—just as digi- digits, or bits—a string of zeros and ones—to represent the tal recording provides higher-fidelity music. Most of our desired characters. The form of computer data does not telephone instruments are still analog devices, so the signal mesh well with analog transmission. First, only two sent from the instrument to the nearest switching center distinct signals—representing zeros and one—need to be (which may be operated either by the telephone company sent, and second, the data are extremely sensitive to degra- or by your own organization) is still an analog signal. These dations in signal quality. Noise in a telephone line could telephone switches, however, are rapidly being converted easily cause a zero to be interpreted as a one or vice versa, from analog to digital switches. When the analog voice and the entire message might become garbled. Because of signal arrives at a digital switch, it is converted to a digital this problem with noise, data cannot be sent directly over voice signal for transmission to a digital switch somewhere the analog telephone network. else, which may be across town or across the country. Two solutions are possible to the problem of trans- Thus, an increasing proportion of the voice transmission mitting computer data. The original solution, and one that between switching centers is digitized. In the future, our is still used, is to convert the data from digital form to telephone instruments will also be digital devices, so the analog form before sending it over the analog telephone entire telephone network will eventually become digital. network. This conversion is accomplished by a device called a modem, an abbreviation for a modulator/ Speed of Transmission demodulator (see Figure 3.1). Of course, the data must be reconverted from analog form back to digital form at the Whether the signal is digital or analog, another basic other end of the transmission line, which requires a second question is the speed of transmission. Please note that by modem. The use of modems and the analog telephone speed we do not mean how fast the signal travels in terms like network is an acceptable way to transmit data for many miles per hour, but rather the volume of data that can be applications, but it is severely limited in terms of transmis- transmitted per unit of time. Terms such as bandwidth, baud, sion speeds and error rates. and Hertz (Hz) are used to describe transmission speeds, The second and longer-term solution to the problem whereas a measure such as bits transmitted per second (bits of transmitting computer data is to develop digital per second, or bps) would be more understandable. Happily, networks specifically designed to directly transmit a digi- the three terms mentioned previously are essentially the same tal signal consisting of zeros and ones. Digital networks as bits per second in many circumstances. In common have the advantages of potentially lower error rates and usage, bandwidth is just the circuit capacity. Hertz is higher transmission speeds, and modems are no longer cycles per second, and baud is the number of signals sent per necessary. Because of these advantages, the networks that second. If each cycle sends one signal that transmits exactly have been specifically created for the purpose of linking one bit of data, which is often the case, then all these terms computers and computer-related devices are digital. are identical. To minimize any possible confusion, we will Furthermore, the telephone network is gradually being talk about bits per second (bps) in this chapter. In information shifted from an analog to a digital network. Digital services technology publications, baud was formerly used for such as ISDN and DSL (to be explored later in this relatively slow speeds such as 2,400 baud (2,400 bps) or chapter) are now available in many parts of the United 14,400 baud (14,400 bps), while Hertz (with an appropriate States for users seeking higher-speed access to the Internet prefix) was used for higher speeds such as 500 megaHertz over the public telephone network. (500 million bps) or 2 gigaHertz (2 billion bps). More This shift of the telephone network from analog to recently, the term baud has fallen into disfavor, but Hertz is digital is due in part to the increasing volume of data being still widely used in PC advertisements. For clarity, we will transmitted over the network, but there is also a significant stick with bps in this chapter. Transmission Line MICROCOMPUTER MODEM MODEM SERVER Digital Analog Digital Signal Signal Signal FIGURE 3.1 The Use of Modems in an Analog Network  Chapter 3 Telecommunications and Networking 65 The notion of bandwidth, or capacity, is important from a common carrier company such as Verizon, Sprint for telecommunications. For example, approximately Nextel, or AT&T. A company might choose to lease a line 50,000 bits (0s and 1s) are required to represent one page between Minneapolis and Atlanta to ensure the quality of of data. To transmit 10 pages using a 56,000 bps (56 kbps) its data transmissions. Private lines also exist within a modem over an analog telephone line would take about building or a campus. These are lines owned by the organ- nine seconds. If one were transmitting a large data file ization for the purpose of transmitting its own voice and (e.g., customer accounts), that bandwidth or capacity data communications. Within-building or within-campus would be unacceptably slow. On the other hand, to trans- lines for computer telecommunications, for example, are mit these same 10 pages over a 1 million bps (1 mbps) usually private lines. DSL line would take only half of a second. Graphics The last basic idea we wish to introduce is the require approximately 1 million bits for one page. To difference among simplex, half-duplex, and full-duplex transmit 10 pages of graphics at 56 kbps over an analog transmission. With simplex transmission, data can travel telephone line would take a little under 3 minutes, while only in one direction. This one-way communication is transmitting these same 10 pages of graphics over a rarely useful, but it might be employed from a monitoring 1 mbps DSL line would take only 10 seconds. One hour of device at a remote site (e.g., monitoring power consump- high-definition video requires approximately 3 billion tion) back to a computer. With half-duplex transmission, bytes (3 gigabytes), which would take nearly five days to data can travel in both directions but not simultaneously. transmit over a 56 kbps analog telephone line—obviously Full-duplex transmission permits data to travel in no one would ever do this! This high-definition video both directions at once, and, therefore, provides greater would still take 6 2/3 hours over a 1 mbps DSL line, but capacity, but it costs more than half-duplex lines. Ordinary the time drops to a more reasonable 13 1/3 minutes over telephone service is full-duplex transmission, allowing a 30-mbps fiber-to-the-premises line. The bandwidth both parties to talk at once, while a Citizen’s Band (CB) determines what types of communication—voice, radio provides half-duplex transmission, allowing only one data, graphics, stop-frame video, full-motion video, high- party to transmit at a time. definition video—can reasonably be transmitted over a particular medium. Transmission Media A telecommunications network is made up of some physi- Types of Transmission Lines cal medium (or media) over which communications are Another basic distinction is between private (or dedicated) sent. Five primary media are in use today: twisted pair of communication lines and switched lines. The public wires, coaxial cable, wireless, satellite (which is a special telephone network, for example, is a switched-line system. form of wireless), and fiber-optic cable. When a communication of some sort (voice or data) is sent over the telephone network, the sender has no idea TWISTED PAIR When all uses are considered, the most what route the communication will take. The telephone common transmission medium is a twisted pair of wires. company’s (or companies’) computers make connections A twisted pair consists of two insulated copper wires, between switching centers to send the communication over typically about 1 millimeter thick, twisted together in a the lines they deem appropriate, based on such factors as long helix. The purpose for the twisting is to reduce the length of the path, the amount of traffic on the various electrical interference from similar twisted pairs nearby. routes, and the capacity of the various routes. This Most telephones are connected to the local telephone switched-line system usually works fine for voice commu- company office or the local private branch exchange nications. Data communications, however, are more sensi- (PBX) via a twisted pair. Similarly, many LANs have been tive to the differences in line quality over different routes implemented by using twisted pair wiring to connect the and to other local phenomena, such as electrical storms. various microcomputers and related devices. For example, Thus, a data communication sent from Minneapolis to Category 5e cabling—which consists of four twisted pairs Atlanta over the telephone network might be transmitted in a single cable jacket—is currently used for many new perfectly at 11 A.M., but another communication sent from high-speed LANs. Minneapolis to Atlanta 15 minutes later (a different The transmission speeds attainable with twisted connection) might be badly garbled because the communi- pairs vary considerably, depending upon such factors as cations were sent via different routes. the thickness of the wire, the distance traveled, and the One way to reduce the error rate is through private number of twisted pairs in the cable. On the analog voice lines. Most private lines are dedicated physical lines leased telephone network, speeds from 14,400 to 56,000 bps are 66 Part I Information Technology commonplace. When a digital service such as DSL is used Because of its construction, coaxial cable provides a on the telephone network, a speed of 256,000 bps is good combination of relatively high transmission speeds common, with outbound DSL speeds ranging up to 896 and low noise or interference. Two kinds of coaxial cable kbps and inbound DSL speeds up to 12 million bps are in widespread use—baseband coax, which is used for (12 mbps). Much higher speeds can be obtained when digital transmission, and broadband coax, which was twisted pairs are used in LANs. Multiple twisted pairs in a originally used for analog transmission but which is now single cable—such as Category 5e cabling—can support used for digital transmission as well. speeds up to 100 mbps when used in a Fast Ethernet LAN, Baseband coax is simple to use and inexpensive to or even up to 10 billion bps (10 gbps) with Gigabit install, and the required interfaces to microcomputers or Ethernet (more on these LAN types later). The speeds of other devices are relatively inexpensive. Baseband offers a twisted pair and other media are summarized in Table 3.2. single digital transmission channel with data transmission rates ranging from 10 million bits per second (10 mbps) up COAXIAL CABLE Coaxial cable (coax) is another common to perhaps 150 mbps, depending primarily on the transmission medium. A coaxial cable consists of a heavy distances involved (longer cables mean lower data rates). copper wire at the center, surrounded by insulating material. Baseband coax was widely used for LANs and for long- Around the insulating material is a cylindrical conductor, distance transmission within the telephone network, which is often a woven braided mesh. Then the cylindrical although much of this coax has now been replaced by conductor is covered by an outer protective plastic covering. fiber-optic cabling. Figure 3.2 illustrates the construction of a coaxial cable. Broadband coax, which uses standard cable television cabling, was originally installed for analog transmission of television signals, but it increasingly employs digital transmission. A single broadband coax can TABLE 3.2 Telecommunications Transmission Speeds be divided into multiple channels, so that a single cable can Transmission Medium Typical Speeds support simultaneous transmission of data, voice, and Twisted pair—voice telephone 14.4 kbps–56 kbps television. Broadband data transmission rates are similar to Twisted pair—digital telephone 128 kbps–24 mbps those for baseband coax, and high transmission speeds are Twisted pair—LAN 10 mbps–100 gbps possible over much longer distances than are feasible for baseband coax. Because of its multiple channels and addi- Coaxial cable 10 mbps–150 mbps tional capacity, broadband coax has been more enduring Wireless LAN 6 mbps–600 mbps than baseband. Broadband coax is still widely used for Microwave 256 kbps–1 gbps cable television and LANs that span a significant area, Satellite 256 kbps–1 gbps often called metropolitan area networks. Fiber-optic cable 100 mbps–3,200 gbps KEY: bps = bits per second WIRELESS Strictly speaking, wireless is not a transmis- kbps = thousand bits per second, or kilo bps sion medium. Wireless is broadcast technology in which mbps = million bits per second, or mega bps radio signals are sent out into the air. Wireless communi- gbps = billion bits per second, or giga bps cation is used in a variety of circumstances, including Copper Insulating Braided Protective Core Material Outer Plastic Conductor Covering FIGURE 3.2 Construction of a Coaxial Cable  Chapter 3 Telecommunications and Networking 67 cordless telephones, cellular telephones, wireless LANs, 3G (third generation) networks operated by the cellular and microwave transmission of voice and data. carriers. The 3G networks provide much faster data trans- A cordless telephone is a portable device which mission rates—typically from 500 kbps to 2 mbps—than may be used up to about 1,000 feet from its wired the 2G networks, which are often in the 20 kbps range. telephone base unit. This permits the user to carry the 4G networks, with even higher data transmission rates, are telephone to various rooms in a house or take it outdoors on the horizon. on the patio. By contrast, a cellular telephone may be Wireless LANs are growing in popularity. They used anywhere as long as it is within range—about eight to have the obvious advantage of being reasonably easy to ten miles—of a cellular switching station. At present, these plan and install. A wireless system provides networking cellular switching stations are available in all metropolitan where cable installation would be extremely expensive or areas of the United States and most rural areas. The impractical, such as in an old building. A wireless LAN switching stations are low-powered transmitter/receivers also permits users of mobile devices such as handheld or that are connected to a cellular telephone switching office laptop computers to connect to the LAN (and thus the by means of conventional telephone lines or microwave Internet) whenever they are within range of a wireless technology. The switching office, which is computer- access point (WAP), such as in a coffee shop or an airport controlled, coordinates the calls for its service area and terminal. A wireless LAN is less secure than a wired LAN links the cellular system into the local and long-distance and more susceptible to interference, which might increase telephone network. Today’s smartphones, such as a the error rate and force the wireless LAN to operate at a BlackBerry or an iPhone, combine a cell phone with a slower data rate. Most wireless LANs operate in the range handheld computer and have the ability to access the of 6 to 100 mbps, with some of the newer wireless LANs Internet over the so-called 2G (second generation) or operating at speeds up to 300 mbps. RFID Signals Growing Slowly Radio frequency identification (RFID) has been around since World War II, but it didn’t gain public attention until 2003 when Walmart announced that it would require its 100 top suppliers to begin using RFID for selected applications by January 2005. At the time, there was an expectation that RFID would change retailing forever, but that hasn’t happened. In fact, even Walmart has pulled back on its original mandate. While it still uses RFID in stores and distribution centers, Walmart does not require its suppliers to use RFID. Only about 600 of Walmart’s 20,000 suppliers have joined the movement to RFID. Nevertheless, RFID has blossomed in areas other than retailing, and the sales of RFID tags, readers, soft- ware, and services have grown to $6 billion a year. The biggest use of RFID is in smart cards, including passports, ID cards, and prepaid transportation cards. The military is using RFID to keep track of equip- ment and supplies, and governments are using RFID to tag livestock to help stop the spread of diseases such as mad cow disease. There are actually two flavors of RFID, passive and active. Both types of RFID are built around an RFID tag, which is a small piece of hardware—often the size of a postage stamp or smaller—that uses radio-transmitted codes to uniquely identify itself. A passive RFID tag combines a tiny chip with an antenna; it does not have an internal power source. Instead, a passive tag relies on the minute electrical current induced in the antenna by an incoming radio signal, providing enough power for the tag to send a brief response, typically just an ID number. An active RFID tag contains its own power supply and can transmit identifying information either continuously, on request, or on a predetermined schedule. Active tags are much more expensive, usually over $1.00, while the prices of passive tags have dropped to the $0.05 to $0.10 level. At present, most of the RFID action involves passive RFID tags because of their lower cost. Passive RFID tags include those used on smart cards and livestock, as well as on an increasing number of consumer items. Active RFID tags are commonly found in cell phones, aircraft transponders, and other specialized applications such as the tracking of medical equipment. RFID is not creating a retailing revolution, but it is slowly and surely becoming important for identifying people, products, equipment, and even livestock. [Based on Khermouch and Green, 2003; Ohlhorst, 2005; and Weier, 2009] 68 Part I Information Technology Microwave has been in widespread use for long- is widely used for long-distance telephone communication distance wireless communication for several decades. and, to a lesser extent, for corporate voice and data net- Microwave is line-of-sight transmission—there must be an works; transmission speeds up to 1 gbps are possible. unobstructed straight line between the microwave trans- Other line-of-sight transmission methods exist in mitter and the receiver. Because of the curvature of the addition to microwave. For short distances (e.g., from one earth, microwave towers have to be built, typically about building to another), laser or infrared transmitters and 25 to 50 miles apart, to relay signals over long distances receivers, mounted on the rooftops, are often an economical from the originating transmitter to the final receiver. These and easy way to transmit data. requirements for towers, transmitters, and receivers suggest that microwave transmission is expensive, and it is, but SATELLITE A special variation of wireless transmission long-distance microwave is less expensive than burying employs satellite communication to relay signals over fiber-optic cable in a very long trench, particularly if the very long distances. A communications satellite is simply right of way for that trench has to be obtained. Microwave a big microwave repeater in the sky; it contains one or Bluetooth is Here! Harald Bluetooth was a tenth-century Viking king in Denmark. Now a wireless technology named in his honor allows communication among a wide variety of devices, such as mobile telephones, desktop and notebook computers, palmtop computers, DVD players, and printers, eliminating cables and permitting communication where it used to be impossible. Bluetooth is short-range radio technology that has been built into a microchip, enabling data to be transmitted wirelessly at speeds up to 3 mbps (Version 2.1 + EDR)—and eventually at speeds up to 24 mbps if a Wi-Fi (wireless fidelity) connection is available (Version 3.0 + HS). The Bluetooth Special Interest Group’s founding members were two leading mobile phone manufacturers, Ericsson (Sweden)1 and Nokia (Finland); two leading notebook computer vendors, IBM (now Lenovo, based in China) and Toshiba (Japan); and Intel, the leading producer of microprocessor chips. They have been joined by many other companies, including Microsoft and Motorola, as promoter members. The Bluetooth Special Interest Group has developed Bluetooth technology standards that are available free of royalties to any company that wishes to use them. Products using Bluetooth technology have to pass interoperability testing prior to release. Thousands of Bluetooth products of all kinds are now available for purchase, and Bluetooth support is embedded in operating systems such as Microsoft Windows and Apple Computer’s Mac OS. The possibilities are endless for the use of Bluetooth. By adding Bluetooth cards (containing the microchip) to a notebook computer and a palmtop, a business traveler is able to synchronize the data in a notebook computer and palmtop simply by placing both devices in the same room. Bluetooth can eliminate the need to use cables to connect the mouse, keyboard, and printer to a desktop computer. With several states now prohibiting driving while talking on a handheld phone, there is growing demand for Bluetooth-equipped cell phones and hands-free headsets. An array of Bluetooth-equipped appliances, such as a television set, a stove, a thermostat, and a home computer, can be controlled from a cell phone—all from a remote location, if desired. The Bluetooth Special Interest Group has designed the microchips to include software controls and identity coding to ensure that only those units preset by their owners can communicate. As a specific example, UPS is using Bluetooth technology in the ring scanners used by package loaders. These ring scanners read bar-code data on packages and transfer it via Bluetooth to terminals they wear on their waists. Then, using wireless LAN access points deployed throughout all of UPS’s buildings, the data are sent from the LAN via a landline to a global scanning system—which stores all of the information on packages—at one of two UPS data centers. Watch out for the Viking king— Bluetooth is here! [Based on Bluetooth Web site, 2010; Malykhina, 2006; Perez, 2009a; and Wildstrom, 2008] 1 The companies listed in this chapter have their headquarters in the United States unless otherwise noted.  Chapter 3 Telecommunications and Networking 69 more transponders that listen to a particular portion of the satellites are close to the earth, the ground stations need electromagnetic spectrum, amplify the incoming signals, less power for communication and the round-trip delay is and retransmit back to earth. A modern satellite might have greatly reduced. Fifteen years ago it appeared as though around 40 transponders, each of which can handle an nearly 1,700 LEO satellites would be launched by 2006— 80-mbps data transmission, 1,250 digital voice channels of more than ten times the 150 commercial satellites in orbit 64 kbps each, or other combinations of data channels and at that time (Schine et al., 1997)—but that did not happen. voice channels. Transmission via satellite is still line- Let’s see why. of-sight transmission, so a communication would have to The first major LEO project was Iridium, which be relayed through several satellites to go halfway around launched 66 satellites to offer mobile telephony, paging, and the world (see Figure 3.3). data communication services. Investors in the $5 billion One interesting, but annoying, aspect of satellite Iridium project included Motorola, Lockheed Martin, and transmission is the substantial delay in receiving the signal Sprint; Motorola managed the project. The satellites were all because of the large distances involved in transmitting up flying, and the Iridium system went live in 1998, with to the satellite and then back down to earth. This is two-page advertisements splashed in major magazines such particularly true for the geostationary earth orbit (GEO) as BusinessWeek. The Iridium customer would have an satellites, which are positioned 22,000 miles above the individual telephone number that would go with him or her equator such that they appear stationary relative to the anywhere on earth, enabling the customer to make and earth’s surface. The minimum delay for GEO satellites is receive calls from even the most remote places on the globe. just under one-third of a second, which is an order of Unfortunately, the prices to use the Iridium service magnitude larger than on fiber-optic connections or earth- were too high, and it never caught on. Iridium filed for bound microwave covering the same ground distance. bankruptcy in 1999, and for a time it appeared likely that In the 1990s, a great deal of interest arose in low the satellites would be allowed to fall out of orbit. But earth orbit (LEO) satellites, orbiting at a distance of only Iridium got a second chance! A group of investors paid 400 to 1,000 miles above the earth—compared to 22,000 $25 million for the satellites and other assets of the original miles above the earth for GEO satellites. Because of their Iridium (quite a bargain!) and started satellite telephone rapid motion, it takes a large number of LEO satellites service again in 2001 (Jarman, 2009). The old Iridium for a complete system; on the other hand, because the needed 1 million customers to break even; the new Iridium needed only tens of thousands. Many of these customers came from the U.S. military, which signed a deal for Satellite unlimited use for up to 20,000 soldiers. The British A military is another customer, as are many news media rep- Microwave resentatives (Maney, 2003). Marketing of the new Iridium Station 1 is aimed at corporations, not individuals, with a focus on maritime, aviation, and mobile operations. Iridium now has a customer base of over 300,000 (Jarman, 2009). The cost is still substantial for the reborn Iridium but not nearly as high as before: The telephone, which weighs 9.4 ounces, costs about $1,500, and one sample calling plan costs $250 a month for 150 included minutes ($1.67 per minute), plus Earth $1.40 per minute for additional minutes. In addition, there is a $100 activation fee and a substantially higher rate to call a telephone operating on another satellite network (Blue Sky Network Web site, 2010). A competing LEO satellite system, Globalstar, has also had a troubled history. With its 48 LEO satellites, Globalstar does not provide complete coverage of the Microwave planet, but it does offer service in over 120 countries. The Station cost of Globalstar’s Freedom 150 plan (in the United 2 Satellite States) is $65 per month for 150 included minutes ($0.43 B per minute), plus $0.99 per minute for additional minutes. The telephone for Globalstar costs about $1,000 FIGURE 3.3 Satellite Communications (Globalstar Web site, 2010). 70 Part I Information Technology The plug was pulled on a third proposed LEO satellite fiber (50 to 100 micron2 core, which does not include system, named Teledesic, in October 2002. The original any protective covering) to as high as 3,200 gbps for plan for Teledesic, which was sponsored by Craig McCaw small-diameter fiber (10 microns or less). The fact that (who built McCaw Cellular before selling it to AT&T), Bill the smaller-diameter fiber has much larger capacity Gates (Microsoft), and Boeing, was to create a 288-satellite might be surprising, but light reflections are greatly network to provide low-cost, high-speed Internet access, reduced with a smaller fiber—the light ray bounces corporate networking, and desktop videoconferencing. The around less—permitting higher transmission speeds. The number of satellites was later reduced to 30, each with a large-diameter fiber is multimode, meaning that several larger “footprint” on the earth, but even that plan was light rays are traversing the fiber simultaneously, bounc- cancelled in 2002 before any Teledesic satellites were ing off the fiber walls, while the small-diameter fiber is launched. All these LEO satellite systems seemed like good single mode, with a single light ray at a time propagated ideas at the time they were planned, but the expenses essentially in a straight line without bouncing. Single- involved were massive. Furthermore, the LEO systems took mode fiber, unfortunately, requires higher-cost laser so long from concept to deployment that competing, less light sources and detectors than multimode fiber. In a expensive technologies—such as cell phones, DSL, and recent development, the light ray sent through a single- cable—had made massive inroads into the potential market mode fiber can be split into 80 or more different colors, before the satellites were launched. each carrying its own stream of data. In this process, called dense wave division multiplexing, prisms are used FIBER OPTICS The last and newest transmission medium— to send these multiple colors down a single fiber. For fiber-optic cabling—is a true medium, not broadcast tech- example, some of the fiber currently being installed by nology. Advances in optical technology have made it possi- telephone companies is 8-micron single-mode fiber with ble to transmit data by pulses of light through a thin fiber of a transmission speed, using wave division multiplexing, glass or fused silica. A light pulse can signal a 1 bit, while of 800 gbps. The outside diameter (including protective the absence of a pulse signals a 0 bit. An optical transmis- covering) of this single-mode fiber is only 125 microns, sion system requires three components: the light source, which is about one-fiftieth the outside diameter of a typ- either a light-emitting diode (LED) or a laser diode; the ical coaxial cable. Thus, both the speed and size advan- fiber-optic cable itself; and a detector (a photodiode). The tages of fiber optics are significant. light source emits light pulses when an electrical current is applied, and the detector generates an electrical current Topology of Networks when it is hit by light. Fiber optics are much faster than other media and The starting point for understanding networks is to require much less space because the fiber-optic cable is recognize that all telecommunications networks employ very small in diameter. Fiber-optic cables are more secure one or more of the transmission media discussed previ- because the cables do not emit radiation and, thus, are very ously. But what do the networks look like in terms of difficult to tap. They are also highly reliable because they their configuration or arrangement of devices and are not affected by power-line surges, electromagnetic media? The technical term for this configuration is the interference, or corrosive chemicals in the air. These topology of the network. There are five basic network benefits are leading telephone companies to use fiber topologies—bus, ring, star, hierarchical or tree, and optics in all their new long-distance telephone lines, lines mesh (see Figure 3.4)—plus an unlimited number of connecting central office sites, and most of their new local variations and combinations of these five basic forms. lines from central office sites to terminuses located in subdivisions. (The advantages of speed and security are BUS The simplest topology is the linear or bus topology. obvious; the size is important because many of the cable With the bus, a single length of cable (coax, fiber, or ducts already installed lack room for other media but can twisted pair) is shared by all network devices. One of the hold the thinner fiber-optic cabling.) The high cost of the network devices is usually a file server with a large data required equipment and the difficulty of dealing with the storage capacity. An obvious advantage of the bus is the tiny fibers make this an unattractive medium for most wiring simplicity. A disadvantage is its single-point LANs, except when it is used as a backbone to connect failure characteristic. If the bus fails, nodes on either side multiple LANs and where very high speeds or high-security of the failure point cannot communicate with one another. needs exist. Transmission speeds for fiber range up to 1 billion 2 bits per second (1 giga bps or 1 gbps) for large-diameter A micron is one-millionth of a meter or one-thousandth of a millimeter.  Chapter 3 Telecommunications and Networking 71 because with some rearrangement (spreading the branches Bus out around the central device), it looks like an extension of the star. The configuration of most large and very large computer networks is a tree, with the mainframe at the top of the tree connected to controllers such as a multiplexer3 and perhaps to other smaller computers. Then these Ring Star controllers, or smaller computers, are, in turn, connected to other devices such as terminals, microcomputers, and printers. Thus, the tree gets “bushy” as one traverses it from top to bottom. The tree has the same primary disadvantage as the Tree star. If the central device fails, the entire network goes down. On the other hand, the tree arrangement possesses a great deal of flexibility. The cost disadvantage of the star might not appear when devices are added to the network, for the use of intermediate devices (e.g., multiplexers, Mesh small computers) removes the necessity of connecting every device directly to the center. MESH In a mesh topology, most devices are connected FIGURE 3.4 Network Topologies to two, three, or more other devices in a seemingly irregular pattern that resembles a woven net or a mesh. A complete mesh would have every device connected to RING The ring topology is similar to the bus except that every other device, but this is seldom done because of the the two ends of the cable are connected. In this case, a cost. The public telephone network is an example of a single cable runs through every network device, including mesh topology; another example is the system of networks (usually) a file server. The wiring for the ring is slightly that makes up the Internet. more complicated than for the bus, but the ring is not as The ramifications of a failure in the mesh depend susceptible to failure. In particular, a single failure in the upon the alternative paths or routes available in the vicin- ring still permits each network device to communicate ity of the failure. In a complex mesh, like the telephone with every other device. network, a failure is likely to have little impact, except on the devices directly involved. STAR The star topology has a mainframe or midrange computer, a file server (usually a microcomputer), or a MORE COMPLEX NETWORKS Now the fun begins, networking device at its center, with cables (or media of because the previous five network topologies can be some type) radiating from the central device to all the other combined and modified in a bewildering assortment of network devices. This design is representative of many networks. For example, it is quite common to attach small-to-medium computer configurations, with all multiple bus or ring LANs to the tree mainframe computer workstations and peripherals attached to the single network. Multiple bus and ring LANs might be attached to midrange computer. Advantages of the star include ease of a high-speed backbone cable, which is in effect a bus identifying cable failure, because each device has its own network with the LANs as nodes on the network. cable; ease of installation for each device, which must only National and international networks—such as the be connected to the central device; and low cost for small telephone network and the Internet—are much more networks where all the devices are close together. The complex than those we have considered thus far, because star’s primary disadvantage is that if the central device the designers have intentionally built in a significant fails, the whole network fails. A cost disadvantage might amount of redundancy. In this way, if one transmission line also be encountered if the network grows, for a separate cable must be run to each individual device, even if several 3 devices are close together but far from the central device. A multiplexer is a device, usually located at a site remote from the mainframe or central device, whose function is to merge (“multiplex”) the data streams from multiple low-speed input devices, such as terminals TREE The fourth basic topology is the tree, or hierarchical. and microcomputers, so that the full capacity of the transmission line to This topology is sometimes called a hierarchical star the central device is utilized. 72 Part I Information Technology goes out, there are alternative routes to almost every node (a campus). In many organizations even today, the predom- or device on the network. We will take a closer look at an inant communication with the central computer is through extremely high-speed national research network named the computer telecommunications network. This type of Internet2 later in this chapter. network is controlled by the central computer, with all other devices (e.g., terminals, microcomputers, and printers) Types of Networks operating as subordinates or “slaves” on the network. IBM’s mainframe architecture was originally based on this Thus far we have considered two key elements of type of network, although LANs and other network types telecommunications networks: the transmission media may now be linked to a mainframe or large computer. used to send the communications and the arrangement or This is not a bad arrangement, but it puts a tremendous topology of the networks. Now we turn to the categoriza- communications control burden on the central computer. For tion of networks into basic types, including computer this reason, it is quite common to add a front-end processor telecommunications networks, LANs, backbone networks, or communications controller to the network—between the WANs, the Internet, and Internet2. central computer and the rest of the network—to offload the communications work from the central computer (see COMPUTER TELECOMMUNICATIONS NETWORKS It is Figure 3.5). A front-end processor or communications almost easier to describe this initial type of network by controller is another computer with specially designed what it is not. It is not a LAN, a backbone network, a hardware and software to handle all aspects of telecommuni- WAN, or the Internet. What we are calling a computer cations, including error control, editing, controlling, routing, telecommunications network is the network emanating and speed and signal conversion. from a single medium, large, or very large computer or a group of closely linked computers. This type of network LOCAL AREA NETWORKS A local area network (LAN) usually is arranged as a tree (see Figure 3.4) with coaxial is first and foremost a local network—it is completely cable and twisted pair as the media. Until the early 1980s, owned by a single organization and generally operates this was usually the only type of network (except for the within an area no more than 2 or 3 miles in diameter. telephone network) operated by an organization that did LANs are data networks that generally have a high data business in one building or a group of adjacent buildings rate of several million bps or more. Mainframe Computer Front-end Processor Controller Controller Controller Terminal Printer Fax Terminal Micro Terminal Printer Micro Micro Terminal FIGURE 3.5 Computer Telecommunications Network  Chapter 3 Telecommunications and Networking 73 A LAN differs from a computer telecommunica- but it is usually implemented as a physical star arrange- tions network in that a LAN contains a number of intelli- ment (see Figure 3.6). The usual way of creating a shared gent devices (usually microcomputers) capable of data Ethernet LAN is to plug the cables from all the devices on processing rather than being built around a central com- the LAN into a hub, which is a junction box containing puter that controls all processing. In other words, a LAN some number of ports (e.g., 12) into which cables can be is based on a peer-to-peer relationship, rather than a plugged. Embedded inside the hub is a linear bus connecting master–subordinate relationship. all the ports. Thus, shared Ethernet operates as a logical There are five types of LANS in use today—three bus but a physical star. types of wired LANs and two types of wireless LANs—for Switched Ethernet is a newer variation of Ethernet which standards have been developed by the Institute for providing better performance at a higher price. The design Electrical and Electronic Engineers (IEEE) and subse- is similar to shared Ethernet, but a switch is substituted for quently adopted by both national and international the hub and the LAN operates as a logical star as well as a standards organizations. These five LAN standards are physical star. The switch is smarter than a hub—rather officially designated as IEEE 802.3 (contention bus than passing all communications through to all devices on design); IEEE 802.4 (token bus design); IEEE 802.5 the LAN, which is what a hub does, the switch establishes (token ring design); IEEE 802.11, including 802.11a, separate point-to-point circuits to each device and then 802.11b, 802.11g, and 802.11n (Wi-Fi wireless design); forwards communications only to the appropriate device. and IEEE 802.16, including 802.16d and 802.16e This switched approach dramatically improves LAN (WiMAX wireless design). performance because each device has its own dedicated circuit, rather than sharing a single circuit with all devices Wired Local Area Networks. The contention bus on the network. Of course, a switch is more expensive than design was originally developed by Xerox and subsequent- a simple hub. ly adopted by Digital Equipment Corporation (now part of The token bus design employs a bus topology with Hewlett-Packard) and several other vendors. This design is coaxial cable or twisted pair wiring, but it does not rely on usually referred to as Ethernet, named after the original contention. Instead, a single token (a special communication Xerox version of the design. The contention bus is or message) is passed around the bus to all devices in a spec- obviously a bus topology (see Figure 3.4), usually ified order, and a device can only transmit when it has the implemented using coaxial cable or twisted pair wiring. token. Therefore, a microcomputer must wait until it receives Communication on an Ethernet LAN is usually half- the token before transmitting a message; when the message duplex—that is, communication in both directions is is sent, the device sends the token on to the next device. After possible, but not simultaneously. The interesting feature of some deterministic period of time based on messages sent by this design is its contention aspect—all devices must other devices, the device will receive the token again. contend for the use of the cable. With Ethernet, devices listen to the cable to pick off communications intended for the particular device and PC PC PC determine if the cable is busy. If the cable is idle, any device may transmit a message. Most of the time this works fine, but what happens if two devices start to transmit at the same time? A collision occurs, and the messages become garbled. The devices must recognize that this collision has occurred, stop transmitting, wait a random period of time, and try again. This method of PC Bus operation is called a CSMA/CD Protocol, an abbrevia- Hub tion for Carrier Sense Multiple Access with Collision Detection. In theory, collisions might continue to occur, and thus there is no upper bound on the time a device might wait to send a message. In practice, a contention bus design is simple to implement and works very well as long as traffic on the network is light or moderate (and, thus, PC PC PC there are few collisions). The original Ethernet design, now called shared FIGURE 3.6 Shared Ethernet Topology: Logical Bus, Ethernet, employs a contention bus as its logical topology, Physical Star 74 Part I Information Technology The token bus design is central to the appear (more on WiMAX later). Wi-Fi LANs are rapidly Manufacturing Automation Protocol (MAP), which was proliferating. Wi-Fi technology has obvious advantages for developed by General Motors and adopted by many people on the move who need access to the Internet in manufacturers. MAP is a factory automation protocol (or airports, restaurants, and hotels and on university campus- set of standards) designed to connect robots and other es. Wi-Fi is also gaining acceptance as a home or neigh- machines on the assembly line by a LAN. In designing borhood network, permitting an assortment of laptop and MAP, General Motors did not believe it could rely on a desktop computers to share a single broadband access contention-based LAN with a probabilistic delay time point to the Internet. Numerous city-wide Wi-Fi networks before a message could be sent. An automobile assembly were developed in the early twenty-first century, but many line moves at a fixed rate, and it cannot be held up because of these have been commercial failures; it now appears that a robot has not received the appropriate message from other technologies, such as WiMAX and other 4G the LAN. Therefore, General Motors and many other networks operated by the cellular carriers, will form the manufacturers have opted for the deterministic token basis of wide-area wireless networks (Martin, 2008; bus LAN design. Merron, 2007). Wireless LANs have also moved into the The third LAN standard is the token ring, originally corporate and commercial world, especially in older build- developed by IBM, which combines a ring topology ings and confined spaces where it would be difficult or (see Figure 3.4) with the use of a token as described for the impossible to establish a wired LAN or where mobility is token bus. A device attached to the ring must seize the paramount. Even in newer buildings, wireless LANs are token and remove it from the ring before transmitting a often being employed as overlay networks. In such cases message; when the device has completed transmitting, it Wi-Fi is installed in addition to wired LANs so that releases the token back into the ring. Thus, collisions can employees can easily move their laptops from office to never occur, and the maximum delay time before any office and can connect to the network in places such as station can transmit is deterministic. The usual implemen- lunchrooms, hallways, and patios (FitzGerald and Dennis, tation of a token ring involves the use of a wire center into 2009, p. 244). which cables from individual devices are plugged, creating Today’s Wi-Fi LANs use one of the standards incor- a physical star but a logical ring. porated in the IEEE 802.11 family of specifications. All All three types of wired LAN designs are in use these standards use the shared Ethernet design (logical bus, today. Token bus dominates the manufacturing scene, and physical star; see Figure 3.7) and the CSMA/CA Protocol, Ethernet leads token ring by a wide and growing margin in which is an abbreviation for Carrier Sense Multiple Access office applications. But the hottest type of LAN in the with Collision Avoidance. CSMA/CA is quite similar to early twenty-first century is the wireless LAN, to which CSMA/CD used in traditional Ethernet, but it makes we will now turn. greater efforts to avoid collisions. In one approach to collision avoidance, any computer wishing to transmit a Wireless Local Area Networks. Most wireless LANs message first sends a “request to transmit” to the wireless in use today are of the Wi-Fi (short for wireless fidelity) access point (WAP). If no other computer is transmitting, variety, although WiMAX networks are beginning to the WAP responds by sending a “clear to transmit” signal Laptop Laptop Laptop Laptop Laptop Switch to Bus rest of network Wireless Access Point Laptop FIGURE 3.7 Wi-Fi Local Area Network Topology  Chapter 3 Telecommunications and Networking 75 to all computers on the wireless LAN, specifying the 120 million people by the end of 2010 (Clearwire, 2010). amount of time for which the network is reserved for the In practical terms, WiMAX will operate very much like requesting computer. Wi-Fi but over greater distances and for a greater number In order to establish a wireless LAN, a wireless of users. network interface card (NIC) must be installed in each There are actually two types of WiMAX. The IEEE computer. The wireless NIC is a short-range radio trans- 802.16d standard covers fixed-point wireless access, and it ceiver that can send and receive radio signals. At the heart is used to connect a central access point to a set of fixed of a wireless LAN is the wireless access point (WAP), networks, such as from a branch office to a central office a which is a radio transceiver that plays the same role as a few miles away. Under ideal conditions, 802.16d provides hub in a wired Ethernet LAN. The WAP receives the a data rate of 40 mbps for up to 20 miles, but actual data signals of all computers within its range and repeats them rates and distances are much less. An important use of to ensure that all other computers within the range can hear 802.16d is to connect multiple Wi-Fi public access points them; it also forwards all messages for recipients not on (e.g., in a city-wide network) to a central switch so that this wireless LAN via the wired network. users can connect to the Internet. At the present time, there are four Wi-Fi LAN stan- By providing access to mobile users, the 802.16e dards in use, with the newest standard (802.11n) approved standard is designed to be direct competition for outdoor in 2009. The 802.11a Wi-Fi standard operates in the 5 GHz Wi-Fi networks. The network is expected to provide up to (gigaHertz) band at data rates up to 54 mbps. The problem 15 mbps of capacity and to have an effective range of up to with 802.11a is that the range is only about 150 feet; in 6 miles with a line of sight to the access point or 2 miles fact, the 54 mbps data rate can be sustained reliably only without a line of sight; in practice, an 802.16e network within about 50 feet of the WAP. The 802.11b standard operates at about 4 to 8 mbps. operates in the 2.4 GHz band at data rates of 5.5 to 11 mbps. The range of 802.11b LANs is typically 300 to 500 feet, Higher-Speed Wired Local Area Networks. LAN tech- which is greater than that of 802.11a. The most widely nology continues to advance as we move further into the used Wi-Fi LAN standard today is 802.11g, which uses a twenty-first century. The top speed of a traditional Ethernet different form of multiplexing in the 2.4 GHz band to LAN is 10 mbps, but Fast Ethernet, operating at 100 mbps, achieve the same range as 802.11b (300 to 500 feet) with is now the most common form of Ethernet in new LANs. data rates up to the 54 mbps of 802.11a. The recently Fast Ethernet uses the same CSMA/CD architecture and the approved 802.11n standard uses both the 2.4 GHz and same wiring as traditional Ethernet. The most popular imple- 5 GHz frequency ranges simultaneously (by using multiple mentations of Fast Ethernet are 100 Base-T, which runs at sets of antennas) in order to increase its data rate and its 100 mbps over Category 5 twisted-pair cabling (four pairs of range. The 802.11n standard provides five times the wires in each cable), and 100 Base-F, which runs at throughput of earlier Wi-Fi standards—up to 300 mbps— 100 mbps over multimode fiber-optic cable (usually two and twice the range—up to 1,000 feet. Happily, the strands of fiber joined in one cable). Although the wiring for 802.11n standard is backward compatible with the earlier Fast Ethernet could handle full-duplex communication, in Wi-Fi standards, so that it will co-exist with—and eventu- most cases only half-duplex is used. ally replace—all of the earlier standards. Many commenta- Even newer and faster than Fast Ethernet is Gigabit tors believe that the 802.11n standard, with its increased Ethernet, with speeds of 1 billion bps and higher. Gigabit speed, coverage, and reliability, will hasten the movement Ethernet is often used in backbone networks, to be to replace wired LANs with a wireless office environment discussed in the next section. There are two varieties of (Bulk, 2008; Moerschel, 2010). Gigabit Ethernet in use today: 1-gbps Ethernet, commonly The newest type of wireless network is WiMAX called 1 GbE; and 10-gbps Ethernet, or 10 GbE. A 1 GbE (short for worldwide interoperability for microwave running over twisted-pair cables is called 1000 Base-T, access), which is based on the IEEE 802.16 family of and it operates over one Category 5e cable (four pairs of specifications. At the beginning of 2010 there were about wires) by using an ingenious procedure to send streams of half a million WiMAX users in the United States, but bits in parallel. There are two versions of 1 GbE when Clearwire—the leading vendor of WiMAX service— running over fiber-optic cabling: 1000 Base-SX uses multi- expects that number to grow substantially over the next mode fiber and 1000 Base-LX uses either multimode fiber few years. In early 2010, the Clearwire WiMAX service, or single-mode fiber depending on the distances involved named Clear, was available in 27 markets across the (up to 1,800 feet with multimode fiber or over 16,000 feet United States, covering 34 million people; Clearwire with single-mode fiber). Currently, 10 GbE is being expects those numbers to grow to 42 markets covering deployed in some backbone networks or in circumstances 76 Part I Information Technology Clearwire and 4G Networks The Clearwire story is an interesting one that is still unfolding. Clearwire was founded by entrepreneur Craig McCaw, who in the 1980s built a cellular phone business that he sold to AT&T for $11.5 billion. In 2008, Clearwire obtained $3.2 billion in investment funding from Sprint Nextel, Intel, Comcast, and Google, among others. Another $1.6 billion in funding was obtained in 2009, with $1 billion of that from Sprint Nextel, which holds a 51 percent interest in Clearwire. With that funding, Clearwire continues to build its 4G WiMAX wireless network. However, there is competition in terms of the technology that will be used to build 4G networks. Both AT&T and Verizon Wireless—the two largest cellular providers in the United States—have decided to develop their 4G networks based on a cellular technology called Long Term Evolution (LTE). Some commentators think that there is room for both WiMAX and LTE technologies in the 4G world, with WiMAX favored for “local nomads” who want fast Internet access anywhere in their local region and LTE favored for national roamers who regularly travel around the country. Even if the dual 4G approach does not work, Clearwire does not see the 4G battle as a big issue—Clearwire insists that its equipment can be shifted from WiMAX to LTE technology relatively easily. The important point is that WiMAX and other 4G networks have the potential to do for broadband Internet access what cell phones have done to phone access. In the same way that many users have given up on wired phone service in favor of cell phones, the 4G networks could replace cable and DSL services and provide Internet access just about anywhere you go. [Based on Ante, 2009; Clearwire, 2010; Kapustka, 2008; and Perez, 2009b] where very high data rates are required. Moreover, 40-gbps The technology involved in backbone networks is Ethernet (40 GbE) and 100-gbps Ethernet (100 GbE) are essentially the same as that described for LANs but at just around the corner, with standards currently under the high end. The medium employed is either fiber-optic development by IEEE. These ultrahigh speed networks cabling or twisted-pair cabling, providing a high data have been designed to run over fiber-optic cables, but, transmission rate—100 mbps, 1 gbps, or more. The amazingly, they can also run over twisted-pair cables. topology is usually a bus (Fast Ethernet or Gigabit Most of these Gigabit Ethernet networks are configured to Ethernet). The only new terminology we need to intro- use full-duplex communication. Ethernet speeds keep duce relates to the hardware devices that connect net- going up, suggesting that Ethernet will continue to be the work pieces together or connect other networks to the preferred networking approach for high-speed LANs and backbone network. backbone networks for the foreseeable future. We have already introduced the hub, the switch, and the WAP. A hub, we know, is a simple device into BACKBONE NETWORKS Backbone networks are the which cables from computers are plugged; it can also be in-between networks—the middle distance networks that used to connect one section of a LAN to another. Hubs interconnect LANs in a single organization with each forward every message they receive to all devices or sec- other and with the organization’s WAN and the Internet. tions of the LAN attached to it, whether or not they need For example, the corporate headquarters of a large firm to go there. A wireless access point is the central device might have multiple buildings spread out over several city in a wireless LAN that connects the LAN to other net- blocks. Each floor of a large building might have its own works. A bridge connects two LANs, or LAN segments, LAN, or a LAN might cover an entire smaller building. when the LANs use the same protocols, or set of rules All these LANs must be interconnected to gain the (more on this later); a bridge is smart enough to forward benefits of networking—enhanced communications, the only messages that need to go to the other LAN. A sharing of resources and data, and distributed data router, or a gateway (a sophisticated router), connects processing. In addition, the LANs must also be connected two or more LANs and forwards only messages that need to the company’s WAN and, in most cases, to the Internet. to be forwarded but can connect LANs that use different A backbone network is the key to internetworking protocols. For example, a gateway is used to connect an (see Figure 3.8). organization’s backbone network to the Internet. A  Chapter 3 Telecommunications and Networking 77 To the Internet Gateway 16-mbps Gateway Token Ring LAN 100-mbps Fast Ethernet backbone network 10-mbps Ethernet LAN Router Wireless Access Point Switch 10-mbps Hub Ethernet LAN (hub contains internal bus) Bridge 10-mbps Ethernet LAN 10-mbps Ethernet LAN Hub Key: 10-mbps indicates microcomputer Ethernet LAN or other network device FIGURE 3.8 Sample Backbone Network switch connects more than two LANs, or LAN segments, usually owned by several organizations (including both that use the same protocols. Switches are very useful to common carriers and the user organization). In addition, a connect several low-speed LANs (e.g., a dozen Ethernet WAN employs point-to-point transmission (except for satel- LANs running at 10 mbps) into a single 100-mbps back- lites), whereas a LAN uses a multiaccess channel (e.g., the bone network (running Fast Ethernet). In this case, the bus and ring). We will note some exceptions, but for the switch operates very much like a multiplexer. The top most part WANs rely on the public telephone network. vendors of these hardware devices include Cisco Systems, Juniper Networks, 3Com (now owned by DDD and WATS. The easiest way to set up a WAN is Hewlett-Packard), and Alcatel-Lucent (France). to rely on ordinary public telephone service. Direct Distance Dialing (DDD) is available through a telephone WIDE AREA NETWORKS Today’s more complex, more company such as AT&T, Sprint Nextel, or Verizon and can widely dispersed organizations need wide area networks be used for voice and data communications between any (WANs), also called long-haul networks, to communicate two spots served by the telephone network. Of course, the both voice and data across their far-flung operations. A speed for data transmission is quite limited (up to 56 kbps), WAN differs from a LAN in that a WAN spans much greater data error rates are relatively high, and the cost per hour is distances (often entire countries or even the globe) and is very expensive. Wide Area Telephone Service (WATS) is 78 Part I Information Technology also available, in which the organization pays a monthly TABLE 3.3 SONET Circuits fee for (typically) unlimited long-distance telephone serv- ice using the ordinary voice circuits. WATS has the same SONET Level Data Transmission Rate advanta