3eEJNwHa

1457
2S U P P L E M E N T A R Y
3C H A P T E R 1 5
4Wireless
5What is essential is invisible to the eye.
6Antoine de Saint-Exupéry, Le petit prince, Ch. 21
7Overview
8Wireless networking has gained rapid popularity since its introduction in the late 1990s.
9Virtually every laptop computer sold since the early 2000s is capable of wireless networking.
10There are various kinds of wireless networks available today, each serving a specific
11user need. This chapter introduces the enabling legal provisions that make free wireless
12networking possible, and the different kinds of wireless networks. At the end of this chapter
13you should know:
14• the business impact of wireless networking
15• the special features of frequency bands that are used for wireless networking
16• how wireless local-area networks work
17• the types of wireless local-area networks
18• how wireless personal-area networks work
19• how wireless metropolitan-area networks work
20Introduction
21Wireless networks are computer networks that use the ISM wireless frequency bands for signal transmission.
22Wireless networks have become enormously popular among both computer users and
23businesses since the early years of the 21st century. Many cities have experimented with citywide
24wireless networks to provide free or inexpensive Internet access to citizens. Reasons
25for the popularity of wireless networks include their convenience and ease of deployment.
26On battery-powered laptops, wireless networking allows users to compute and communicate
27without any power or network cords. Businesses like wireless networking because setting up
28a basic wireless network in a small office requires nothing more than an inexpensive wireless
29router. By comparison, wired networking requires cables to be drawn through ceilings,
30floors, and walls. Wireless networking is becoming so popular that many organizations are 
31458 • Supplementary Chapter 15 / Wireless
32finding that more than half of the Ethernet ports in the organization are unused because
33users prefer wireless networks over wired networks.1
34 Wireless networking may be one of
35those rare services loved by both businesses and employees.
36Wireless networking introduces some important concerns and limitations that users and
37businesses should be aware of. The most visible concern is information security. Wired
38networks have wall outlets in specific locations that can only be reached by users with access
39to the building. By contrast, wireless signals spread out in all directions and can easily bleed
40outside the organization’s boundaries. Without adequate security, malicious users can easily
41access the organization’s computer network through an improperly secured wireless access
42point.
43In a well-publicized example, in 2006, the retail chain T.J. Maxx became the target of a
44hack when attackers were able to drive to the parking lot of a Marshalls store in Minnesota
45and sniff the passwords of store managers as they logged into the network. Because of other
46weaknesses in T.J. Maxx’s network, these hackers were able to retrieve most of the credit
47card information stored on T.J. Maxx’s computers. More than 45 million credit card records
48were stolen. The breach is estimated to have cost the company $250 million, including costs
49to settle lawsuits resulting from the breach. All this began with just an improperly secured
50wireless LAN at one of its stores.
51Another potentially important issue is related to health. As we will see in our discussion
52of the frequency bands at which wireless networks operate, the 2.4 GHz frequency used
53by most wireless LANs is absorbed very efficiently by water. Water is one of the biggest
54components of human bodies. Though the signals generated by wireless access points
55have very low energy and there is no evidence yet of health hazards from these signals, our
56knowledge of the health hazards of exposure to wireless signals is limited. Wireless LANs
57have only been around for about a decade. It is possible that exposure to wireless signals
58over longer durations could have adverse health effects.
59One final point is the limitation of wireless networks. Wireless networks are generally
60slower and less reliable than wired networks. Most wireless networks share bandwidth
61with other applications, such as cordless telephones, and are affected by environmental
62conditions. Connection drop-offs are common with wireless networks. This is not a major
63concern for browsing, e-mail, and other light applications. However, when continuity or
64speed of the connection is essential, wired networks are still greatly preferable to wireless
65networks.
661 J. Cox, “Is It Time to Cut Back on Now-Idle Ethernet?” Network World, 26 (2009): 1.
67Wi-Fi in stadiums
68Wi-Fi is becoming the standard mechanism to supplement cell phone capacity in dense
69areas. At Super Bowl 50 in 2016, the 10 terabyte Wi-Fi data-transfer mark in a single
70game was crossed for the first time. 70,000 fans used more than 1,300 Wi-Fi access
71points and more than 1,200 Bluetooth beacons to send selfies and messages to friends
72around the world.
73ISM Frequency Bands • 459
74ISM Frequency Bands
75Free or inexpensive wireless networking is possible because of the existence of a very
76special category of wireless frequencies. Before we look at wireless technologies, it is useful
77to become aware of these enabling frequencies.
78The special signal frequencies that enable wireless networking are called ISM frequencies.
79ISM frequencies or ISM bands are radio frequencies available internationally for free use for industrial,
80scientific, and medical applications. The terms industrial and scientific are interpreted very broadly,
81and ISM frequencies may be put to almost any use by anybody without permission from
82anyone or payments of license fees to anyone. These frequencies are therefore also called
83unregulated frequencies. Cordless phones, remote-controlled cars, microwave ovens,
84wireless keyboards, and mice are other applications that use ISM frequencies.
85Wireless frequencies have become big business for government. We know from the
86physical layer discussion that for distinct separation at the receiving end, there must be
87only one sine wave at a specific frequency in any given location. Since there are many users
88who would like to use sine waves for wireless transmissions, but only one user can transmit
89at any given frequency, some coordination and allocation is necessary to determine who
90can transmit at a specific frequency. In the US, this coordination is done by the Federal
91Communications Commission (FCC). In the early days, the FCC did not charge fees for
92the privilege of using specific frequencies for transmission. Instead, frequency bands were
93allocated based on technological requirements. However, beginning in 1994, the FCC
94realized that operators of cell phones and other services would be willing to pay for access
95to specific frequencies. Accordingly, the FCC began spectrum auctions to allocate frequency
96bands for specific commercial services to the highest bidders offering these services.
97Why are ISM frequencies available for free use when cell phone operators pay billions
98of dollars to use other frequencies? One reason is that regulators recognize the need for
99wireless frequencies for experimentation and amateur use. The specific frequencies that
100have been selected for ISM use are generally not very useful for commercial use. ISM
101frequencies generally have poor transmission properties and are unlikely to fetch meaningful
102Bill: What did the Vikings use to communicate secretly?
103Jill: I don’t know.
104Bill: The Norse code!
105Source: Boys’ Life magazine, 2013
106Italian inventor Guglielmo Marconi started wireless communications in 1895 by sending
107a Morse message over a distance of a mile. Marconi was awarded the Nobel Prize in
108physics in 1909 for his contributions towards the development of wireless telegraphy.
109460 • Supplementary Chapter 15 / Wireless
110prices at auctions. For example, signals at the 2.45 GHz band are strongly absorbed by
111water. Microwave ovens operate at this frequency since almost 75% of food mass is made
112of water. By quickly transferring energy to the water in food, microwave ovens are able to
113heat and cook quickly and with very high efficiency. Similarly, water vapor in the atmosphere
114absorbs signals in the 2.45 GHz band, resulting in a very short range for these signals. ISM
115frequencies are also absorbed by walls and foliage. Commercial operators are unlikely to pay
116for signals that have poor transmission properties.
117Safety of wireless signals
118The best available research suggests that the radiation from wireless devices and cell
119phones is harmless to humans. However, concerns remain and research into the effects
120of long-term sustained exposure to wireless signals is ongoing. A study at USF on the
121possible adverse effects of exposure to cell phone radiation attracted a lot of attention.2
122Another paper identified an approximately 8% increase in brain activity in regions closer
123to the antenna,3
124 though the clinical significance of this finding is yet unknown. In 2013,
125a student experiment suggested that Wi-Fi routers stinted plant growth in their vicinity.4
126In recent history, however, many of the best minds the world has ever known
127have been lost to radiation exposure because the scientists working with radioactive
128materials were unaware of the dangers at the time. The casualty list includes Prof.
129Richard Feynman, Nobel Prize-winner in physics, whose lectures are extremely
130famous5
131 and who helped identify the cause of the Challenger disaster; Marie Curie,
132who discovered radium and its medical use for taking x-rays; and Rosalind Franklin,
133who took the x-rays of DNA that led to the discovery of the double-helix model of
134DNA. Prof. Feynman is believed to have been exposed to fatal radiation from a ball
135of uranium he kept on his desk while working on the Manhattan project; Marie Curie
136carried radium in her pockets while traveling and working; and Rosalind Franklin
137spent hundreds of hours working with x-rays to take the sharp pictures that led to
138the DNA model.6
139 Until the end of her days, Marie Curie refused to believe that
140x-rays, which were otherwise so beneficial, could cause death. Her papers are so toxic
141with radioactive smudges that they are kept in special containers and require special
142permissions and protections to access.7
1432 http://electromagnetichealth.org/pdf/FINAL-Alzheimers-Mouse-Study-Do-We-Smell-A-Rat.pdf
1443 N.D. Volkow, D. Tomasi, G. Wang, et al., “Effects of Cell Phone Radiofrequency Signal Exposure on Brain
145Glucose Metabolism,” JAMA, 305(8) (2011): 808–813.
1464 Jenn Savedge, “Student science experiment finds plants won’t grow near wifi router,” Mother Nature Network,
147May 23, 2013.
1485 http://www.feynmanlectures.info/.
1496 https://en.wikipedia.org/wiki/Photo_51.
1507 http://www.openculture.com/2015/07/marie-curies-research-papers-are-still-radioactive-100-years-later.html.
151ISM Frequency Bands • 461
152Percy Spencer, the inventor of microwave ovens8
153Percy Spencer was a self-educated inventor, who supported himself and his aunt from
154the age of seven. In the Navy, he became fascinated with wireless signals after learning
155about the wireless operators aboard the Titanic. While working at Raytheon on a radar
156project for the US Department of Defense, he noticed that the chocolate bar in his
157pocket melted when he got close to the radar equipment. Investigating this further led
158to the commercial development of the microwave oven.
159Table 15.1: ISM frequency bands in the US
160ISM frequency Bandwidth
1616.78 MHz ± 15.0 kHz
16213.56 MHz ± 7.0 kHz
16327.12 MHz ± 163.0 kHz
16440.68 MHz ± 20.0 kHz
165915 MHz ± 13.0 MHz
1662.45 GHz ± 50.0 MHz
1675.8 GHz ± 75.0 MHz
16824.125 GHz ± 125.0 MHz
16961.25 GHz ± 250.0 MHz
170122.5 GHz ± 500.0 MHz
171245 GHz ± 1.0 GHz
172The frequencies defined for ISM use in the US are shown in Table 15.1. Fortunately, ISM
173applications make excellent use of these otherwise useless frequencies. You may recognize
174some of these frequencies. Remote controls for radio-controlled cars often use the
17540.68 MHz band. Older cordless phones used the 915 MHz band, while most current
176cordless phones use the 2.45 GHz band. Most of the popular wireless LANs also use the
1772.45 GHz band. The highest-frequency ISM bands will become useful when electronic
178devices operating at these extremely high frequencies can be built at more affordable prices.
179Wireless Network Categories
180There are three primary categories of wireless computer networks. All these categories of
181wireless networks use ISM frequencies. The most familiar are wireless LANs, which go by
182names such as 802.11b and 802.11g. These networks have a range of about 100 feet, which
1838 https://en.wikipedia.org/wiki/Percy_Spencer.
184462 • Supplementary Chapter 15 / Wireless
185is enough to cover an average suburban home or small office. The second category of
186wireless networks is Bluetooth, which is called a personal-area network. This technology is
187used for connectivity within about 10 feet, which is ideal for connecting peripheral devices
188such as wireless keyboards and cameras in the immediate vicinity of a computer. Finally, we
189have an emerging category of wireless networks called metropolitan-area networks, which
190can provide coverage over a range of about 20 miles, enough to cover many metropolitan
191towns. In the rest of this chapter, we will look at the technologies and features of each of
192these categories of wireless networks.
193Wireless Local-area Networks (the 802.11 series)
194Wireless local-area networks are the most familiar of the three categories of wireless
195networks. Most college campuses now have blanket wireless LAN coverage, and many
196college students use laptops with built-in wireless LAN capability to access the Internet. The
197technologies used in wireless LANs are specified in the 802.11 series of IEEE standards
198such as 802.11b, 802.11g, 802.11a, and 802.11n. Wireless LANs are often known as Wi-Fi.
199Technologically, wireless LANs share many similarities with Ethernet, which is a wired
200network and was discussed in detail in the context of the data-link layer. For example, the
201frame structure of wireless LANs is almost identical to the frame structure of Ethernet.
202Also, wireless LANs use the 48-bit MAC addresses discussed with Ethernet.
203There are, however, some important differences between Ethernet and wireless LANs.
204The most important difference is that wireless LANs have no defined boundaries. Wall jacks
205define the end-points for Ethernet. An Ethernet wall jack is hardwired to a specific port
206on a specific switch. As a result, a network administrator can control every aspect of the
207network traffic that flows through the wall jack and to the computer connected to the jack.
208When you connect to the network through a wall jack, you become part of a well-defined
209network. Typically one area of an office is served by one switch, and most users have no
210choice but to become a member of the Ethernet network that is closest to them.
211Figure 15.1: Wireless networks can overlap
212Wireless Local-area Networks (the 802.11 series) • 463
213On the other hand, wireless networks can overlap and they often do. At home, if you
214open up your “connect to network” dialog (right-click on the wireless icon in your system
215tray → connect to a network), you are likely to see wireless networks from many of your
216neighbors, as shown in Figure 15.1. If any of these wireless networks is not security enabled,
217you can use it to connect to the Internet. At the airport, you are likely to see overlapping
218wireless networks from the airport operator, Starbucks, mobile-phone companies, etc.
219Again, if any of these is not security enabled, you can use it to get Internet access.
220Therefore, whereas geographical location uniquely defines network membership in
221Ethernet, it does not define network membership in wireless LANs. The technical implication
222is that, whereas the signal strength of a wired connection always meets Ethernet standards,
223the signal strength, and hence the network experience of a wireless connection, cannot be
224specified. The network performance of a wireless connection depends upon the distance of
225the host from the access point. A user who is very far from an access point will get very weak
226signals. To best serve users at different distances, wireless LAN standards specify different
227signal-modulation schemes for users at different distances from access points. Users who
228are close to an access point are served by faster-changing signals that can carry higher data
229rates but need strong signal strength for reliable detection. Users who are farther away from
230access points are served by signals that can only provide lower data rates but are easier to
231detect in the presence of noise. (You may be able to relate this to the discussion on signal
232detection, especially relating to signal reception in the presence of noise.)
233Another difference between wireless LANs and wired LANs is that, whereas wired
234networks are extremely reliable, wireless is an inherently unreliable medium. Wireless
235networks are hurt by adverse weather, humidity, temperature, and other environmental
236conditions. As a result, the boundaries of a wireless network are not stable and keep
237shifting as environmental conditions change. Also, wireless networks are unprotected from
238competing signals from other devices such as cordless phones, walkie-talkies, fluorescent
239lamps, car ignitions, etc. By contrast, Ethernet cables do not carry any signals besides data,
240thus providing excellent signal-transmission properties.
241Yet another difference concerns multiplexing. Ethernet does not use multiplexing
242because it uses all the available bandwidth in the medium to transmit signals. This is possible
243because Ethernet cables are not used for other applications. But wireless LANs share the
244bandwidth in the air with other users and have to send signals in specified signal bands.
245Therefore, wireless LANs use multiplexing. To use the available bandwidth efficiently,
246multiple channels have been defined within the 2.4 and 5.8 GHz bands. Since stations may
247be transmitting on any of these channels, wireless stations have to scan all the available
248channels to locate transmissions.
249One last factor that makes wireless networks different from Ethernet is that, whereas
250all stations on an Ethernet can hear every transmission, stations at two opposite ends of
251a wireless LAN may not be able to hear each other. As a result, collision detection may
252be unsuccessful in wireless LANs. Wireless LANs therefore do not use CSMA/CD for
253medium-access control. Instead, wireless LANs use collision avoidance, and the mediumaccess
254control (MAC) mechanism used in wireless networks is called carrier sense multiple
255access with collision avoidance (CSMA/CA). What this means is that a waiting wireless
256station does not start transmitting immediately after a previous transmission ends. This is
257because the station knows that this is the time when other waiting stations are also likely 
258464 • Supplementary Chapter 15 / Wireless
259to try to transmit and therefore the chances of a collision are highest at this time. Wireless
260stations wait for a certain time after a transmission ends before attempting to transmit data.
261The primary implication of all these differences between wireless LANs and Ethernet
262is that wireless LANs require far greater error-detection capabilities than Ethernet.
263We will see that this is manifested in the physical layer of wireless LANs, which adds error
264protection over and above the CRC error detection introduced in Ethernet. This relates
265to the discussion about the physical layer and the impact of the transmission medium
266on data communication technologies. The primary challenge in wireless networks is the
267increased noise level. The technology response is to increase error protection in wireless
268LANs compared to Ethernet.
269Wireless LAN Architecture
270Like Ethernet, wireless LANs are a data-link layer technology. Technology standards for
271both Ethernet and wireless LANs are defined by the 802 group at IEEE. The IEEE 802
272group defines standards for local-area networks. As a result of this common origin, the
273frame structure of wireless LANs is almost identical to the frame structure of Ethernet. The
274differences between wireless and wired media discussed in the previous section are handled
275by differences in the physical layer. To account for the greater need for error detection in
276wireless media, the physical layer in wireless LANs adds header fields that help the receiver
277in error detection. We will see these fields later in this section.
278To facilitate mobility, the designers of wireless LAN technologies planned wireless LANs
279in such a way that larger wireless LANs can be built from smaller wireless LANs. The smallest
280component unit of a wireless LAN is the area covered by a single access point. A wireless access
281point is a device that allows wireless hosts to connect to a wired network using wireless LAN technologies such
282as Wi-Fi. The area covered by an access point is called a basic service area (BSA). The basic service area
283and the access point covering that area together are called a basic service set (BSS).
284To create a larger network, such as a campus-wide wireless LAN, basic service sets can
285be connected to each other using any suitable networking technology. This connecting
286technology is called a distribution system. Thus, a campus-wide wireless LAN consists of
287many basic service sets connected to each other through a distribution system. The portal
288is the connection point where the entire wireless LAN is connected to the rest of the wired Internet. This
289structure is shown in Figure 15.2. The larger campus-wide wireless LAN is called an extended
290service set. The 802.11 standard does not consider the distribution system to be a part of the
291extended service set because end users cannot directly connect to the distribution system for
292wireless access. End users have to use a basic service set for wireless access.
293In Figure 15.2, when the first laptop (station 1) wants to send a message to the second
294laptop (station 4), it creates a MAC frame with the MAC address of the second laptop 2 as
295the destination MAC address, and sends the message to its access point (station 2). Station
2962 will forward the message to station 3 over the distribution system, and finally station 3
297will send the message to the laptop at station 4.
298The advantage of composing large wireless LANs from multiple basic service areas is
299that it facilitates mobility. The extended service set appears to end users as one large LAN.
300Users can move anywhere within an extended service set and still retain the same IP address
301and subnet membership. If wireless LANs were not designed as an extended service set,
302and each access point served as a router, each BSS would become an independent subnet. 
303Wireless Local-area Networks (the 802.11 series) • 465
304Each time a user moved from one access point to the next, he would connect to a different
305subnet. This would potentially give him a different IP address and gateway router address.
306This address reallocation would stop any ongoing transfers and could also potentially disturb
307the network connectivity of some applications. With the concept of an extended service
308set, when users move from one access point to another, there is no change to any network
309setting, and ongoing network transfers can continue without interruption.
310Within an organization, basic service sets may be placed as appropriate to deliver the
311required coverage and reliability. For example, in high-traffic areas, basic service sets may
312overlap to provide redundancy and to share traffic. If areas requiring network coverage are
313far from each other, basic service sets may be organized as in Figure 15.2, where they are
314separated from each other.
315On one extended service set, a host needs to get associated with one access point through
316which it will send and receive messages. The distribution system uses this association
317information to deliver messages for a host to the correct access point.
318The 802.11 standard does not specify how the distribution system should send messages
319between access points. Any local-area network technology can be used for the purpose. It is
320common for network administrators to use Ethernet for the distribution system.
321The portal acts as the gateway between the extended service set and the rest of the
322Internet. When a message is sent to a host that is not in the extended service set, the
323distribution system sends the message to the portal. The portal performs all necessary
324packet format changes required for the message to be transported on the neighboring
325network. For example, in Figure 15.2, the portal transforms the outgoing message from the
326wireless 802.11 frame format to the 802.3 Ethernet frame format.
327Figure 15.2: Structure of a campus-wide wireless LAN
328466 • Supplementary Chapter 15 / Wireless
329Most likely, your home network is built using one wireless router. This router acts as the
330access point as well as the portal. Depending upon the technologies used by your Internet
331service provider, this router may transform packets from 802.11 format to 802.3 format,
332or from 802.11 to the WAN frame format used by the ISP.
333802.11 Frame MAC Layer Frame Format
334The general MAC layer frame format for wireless LANs is shown in Figure 15.3. You may
335note that it has many similarities with the Ethernet frame format. However, the wireless
336frame is more complex than Ethernet. Extra fields are necessary to identify the basic service
337set and access points, and to provide reliable transmission in the presence of noise. Also,
338most wireless LAN packets do not have all the fields in the general frame structure shown
339in Figure 15.3.
340As in Ethernet, the wireless LAN frame format includes the source and destination MAC
341addresses, a frame check sequence, and data from the IP layer. These fields perform the
342same functions as in Ethernet—addressing, error detection, and data transfer. However,
343we see that wireless LANs also have some additional fields such as frame control, sequence
344control, QoS control, and duration/ID.
345There are also four possible address fields (recall that Ethernet frames only have two
346address fields). We also see that the preamble and SFD fields of the Ethernet MAC frame
347are missing in the wireless LAN frame. The missing fields are the simplest to understand.
348These are moved to the physical layer header and retain their positions as the earliest fields
349of incoming frames. It is the additional MAC-header fields that are more complex to
350describe and to understand. These additional fields help in identifying the access points
351and in improving reliability. Their functions are described below:
352Frame control This field describes attributes of the frame. For example, does the frame
353carry data, or does it report the status of the network? Is the frame going toward an AP?
354Is it being sent by an AP?
355Duration/ID This field announces the expected amount of time required to transmit
356this frame. All listening stations will wait for this duration before attempting to send data.
357Figure 15.3: 802.11 frame format
3582 bytes
3592 bytes
3606 bytes 6 bytes 6 bytes 2 bytes 6 bytes 2 bytes 4 bytes
3610-2304 bytes
362Direction of data flow
3636 bytes
364Wireless Local-area Networks (the 802.11 series) • 467
365Sending/receiving access-point addresses (address 1–address 4) Since wireless LAN
366packets need to pass through access points, the MAC addresses of the sending or receiving
367access points are added to the frame as required. Packets leaving an access point have the
368sending AP address, and packets sent to an AP have the receiving access point address.
369QoS control This field specifies the desired type of service. The available types of service
370include best effort, voice, and video.
371802.11 Frame Physical layer Format
372Recall that the physical layer in Ethernet added no header fields to the frame. It simply
373converted the frame to a signal. However, the wireless LAN physical layer does add fields
374to the frame header. The wireless LAN physical layer header is shown in Figure 15.5. The
375primary function of the physical layer header is to add error protection to the frame header.
376It also specifies the data rate being used in the transmission.
377Figure 15.4 shows the header fields in a captured wireless packet.9
3789 At this point, it is highly recommended to check out the quick technical tutorial on 802.11 by Pablo Brenner, “A
379technical tutorial on the IEEE 802.11 protocol,” http://www.sss-mag.com/pdf/802_11tut.pdf (accessed Dec. 2015).
380Figure 15.4: Header fields in a captured wireless frame
381468 • Supplementary Chapter 15 / Wireless
382Popular 802.11 Technologies
383Three technologies are currently specified for wireless LANs—802.11a, 802.11b, and 802.11g.
384802.11a and 802.11b were specified in 1999, and 802.11g was specified in 2003. A fourth
385technology, 802.11n was recently specified to increase the speed and range of wireless networks.
386802.11b and 802.11g operate in the 2.4 GHz ISM band. 802.11b is the simpler technology
387and can support a data rate of up to 11 Mbps. The signal modulation techniques used in
388802.11b are defined in Chapters 14, 15, and 18 of the 802.11 standard.10 802.11b uses
389the direct sequence spread spectrum (DSSS) modulation technique. 802.11g adds the
390orthogonal frequency division multiplexing (OFDM) modulation technique to get higher
391data rates than 802.11b. 802.11g can support data rates up to 54.11 Mbps.
392802.11a operates in the 5.5 GHz ISM band. It was the first 802.11 technology to use
393OFDM to achieve data rates of up to 54 Mbps. OFDM is defined in Chapter 17 of the
394802.11 standard. 802.11a has many features such as a greater number of channels that
395should make it the wireless technology of choice. Unfortunately, the signals in the 5.5 GHz
396band do not travel as far as signals at the same power in the 2.4 GHz band. As a result,
397each 802.11a access point covers a slightly smaller area than a 802.11b/g access point.
398802.11b/g are, therefore, economically more efficient than 802.11a for wireless coverage.
399Hence 802.11b/g technologies are more popular than 802.11a.
400802.11n
401The 802.11n standard was finalized in September 2009. The primary goal of 802.11n is
402to provide a data rate of up to 600 Mbps. 802.11n also aims to provide wider coverage
403than 802.11b/g, so that a single access point can cover more than twice the area covered
404by a 802.11b/g access point. 802.11n operates on both the 2.4 and 5 GHz bands and uses
405OFDM to maximize data rates.
406The primary innovation of 802.11n is multiple-input, multiple-output signal transmission.
407802.11n access points and receivers use multiple antennas. Each antenna pair creates an
408independent data channel. Each 802.11n transmission may therefore be seen as multiple
409802.11a/b/g transmissions.
410Apart from the technical innovations introduced in 802.11n, the technology is also very
411interesting because it allows us to peek into the process by which new data communication
412technologies are developed and standardized. The IEEE website maintains a log of the
413standardization process for 802.11n.11 According to the logs, four candidate technologies
41410 The 802.11 standard is accessible from the IEEE website.
41511 http://grouper.ieee.org/groups/802/11/Reports/tgn_update.htm.
416Figure 15.5: Wireless LAN physical layer header
417Wireless Local-area Networks (the 802.11 series) • 469
418were proposed for 802.11n in November 2004. The TGn technology that will be used in
419802.11n obtained the required level of support from the standards committee in March
4202007. In late 2008–early 2009, the technical and editorial issues in the standard were being
421fixed. Since technology vendors already knew most details of the 802.11n technology, and
422only editorial changes were expected in the standards document, draft 802.11n products
423became available in the market in 2008, even before the final standards document had been
424published.
425802.11ac
426The 802.11ac standard was designed to improve upon 802.11n technologies to support
427transmission of multiple channels of HD video streams in homes and support many users
428per access point in enterprises. The standard was finalized in January 2014. 802.11ac achieves
429these higher data rate design objectives primarily by making three refinements to 802.11n:
430(1) moving from the 2.4 GHz band to the 5 GHz band; (2) refining technologies such
431as MIMO introduced with 802.11n; and (3) capitalizing on the improved sensitivity of
432affordable electronics by packing in more information over the same bandwidth.
433Moving from the 2.4 GHz band to the 5 GHz band is one of the primary drivers
434of improved 802.11ac performance. You may recall from the discussion on the physical
435layer that data rates are directly proportional to bandwidth. The 2.4 Ghz band, used by
436802.11n, includes frequencies in the range 2.4 GHz–2.5 GHz, for a total of 100 MHz. In
437conventional use, these are split into three non-overlapping channels of 22 MHz each. The
4385 GHz band includes frequencies ranging from 5.15 GHz to 5.35 GHz, for a total of 200
439MHz, or twice the bandwidth of the 2.4 GHz band. 802.11ac can split this band into two
440non-overlapping bands of 80 MHz each, providing almost four times the bandwidth per
441channel compared to 802.11n.
442Multiple-input, multiple-output (MIMO) is a very recent development (in the 2000s)
443that allows transmitters to send multiple streams of data at the same frequency to the same
444receiver, but through different paths so that the conflicting signals never overlap. 802.11n
445allowed up to 4 MIMO paths, and 802.11ac allows up to 8 MIMO paths. By doubling the
446number of channels, where such MIMO benefits are possible, 802.11ac can provide twice
447the overall data capacity as 802.11n.
448Finally, 802.11ac tries to talk faster than 802.11n (256 QAM instead of 64 QAM), packing
449more information over a single transmission than 802.11n.
450The benefits of 802.11ac come with trade-offs. The total transmission power in ISM
451bands is regulated. When an 802.11ac transmitter uses the 80 MHz channel, the available
452power is distributed over four times the bandwidth compared to a 22 MHz channel. Reduced
453power means that the signal covers a shorter distance before it becomes indistinguishable
454from noise. Reduced power also means that the fast-talking technique (256 QAM) can
455only be used when the transmitter and receiver are in fairly close proximity (typically about
45615 feet), and without any obstacles. To visualize this, imagine talking to a friend. As you try
457talking faster, your friend will have greater difficulty keeping up, and the same level of noise
458will be more disruptive than when you were speaking slowly.
459For these reasons, 802.11ac is most useful in high-density areas such as lecture halls
460and cafeterias, where users can have a clear line-of-sight to the 802.11ac access point. 
461470 • Supplementary Chapter 15 / Wireless
462These locations were very challenging to 802.11a/b/g technologies anyway because the
4632.4 GHz band used by these technologies only supported three slower non-overlapping
464channels in a given space. By supporting two high-speed, non-overlapping channels in
465the same space, 802.11ac can allow network designers to serve high-density areas more
466conveniently.
467Most data communication technologies go through this process of standardization. The
468standardization process begins with the recognition of a need. In the case of 802.11n
469and 802.11ac, it was the need for a high-speed, long-range wireless technology. A neutral
470standards-making body takes the lead in organizing an expert group to identify technical
471solutions that meet the need. In the case of 802.11, this body was the IEEE, which has
472led the standardization of all local-area network standards. Any interested persons from
473universities, technical companies, and even members of the public can become members of
474the group. The group develops proposals and votes on them until one technology solution
475receives overwhelming support. Finally, the technology is described in adequate detail in
476a standards document so that any interested vendor can use the standards document to
477create an implementation of the standard. The standardization process assures users that
478equipment they buy from one vendor will always be compatible with equipment sold by
479other vendors. In the case of 802.11, for example, you can buy a network interface card from
480Broadcom and be assured that it will work with access points sold by Linksys.
481It is also hoped that the formal standardization process will lead to the adoption of the
482best possible technical solutions to meet requirements.
483Personal-area Networks (the 802.15 series)
484The previous section described wireless local-area networks. These networks use an access
485point and provide high-speed connectivity to any hosts within a radius of about 100 feet.
486Hosts typically use wireless LANs for Internet access.
487While wireless LANs are extremely useful, there are many connectivity applications
488where wireless access would be very useful but where Internet access is not necessary. An
489important example is replacing on desktops the short wires that are used for data transfer
490at low speeds, for example, to connect keyboards and mice to the desktop. This is where
491personal-area networks (PANs) come in. PANs, specified by the IEEE 802.15 standard, are
492designed to remove these wires that clutter desktops and make many devices cumbersome
493to use. Personal-area networks are computer networks designed for data transmission among devices in
494close proximity, typically owned by the same individual.
495DLNA
496The ubiquity of wireless LANs has created a strong desire among users to exchange
497content among home devices, for example, to stream music from smart phones to
498the home stereo, watch home security-camera footage on smart phones, etc. The
499digital living network alliance (DLNA) is an industry consortium aiming to achieve this
500interoperability.
501Personal-area Networks (the 802.15 series) • 471
502Personal-area networks like Bluetooth have been developed to provide communication
503over short distances, usually within 30 feet. This distance is sometimes called the personal
504operating space because people and devices within this range are usually in visual range. By
505limiting itself to this range, Bluetooth is designed to serve a small group of participating
506devices, usually carried by one person. Apart from keyboards and mice, other devices
507that use Bluetooth include cell phone headsets and digital cameras (to transfer pictures to
508computers).
509The focus of Bluetooth is to develop extremely small, inexpensive, and low-power
510connectivity solutions. This makes it easy for electronics manufacturers to add Bluetooth
511capabilities to virtually any electronic device, even headsets and telephones, which are not
512traditionally considered computers. Since many of these devices are very small and can
513carry only a limited amount of battery power, power efficiency is an extremely important
514requirement for Bluetooth. Since signal transmission requires power, and more power is
515needed to send signals to greater distances, power efficiency concerns are the reason why
516Bluetooth devices have very short communication ranges.
517Bluetooth operates in the 2.4 GHz ISM band, the same as 802.11b/g wireless LANs.
518Bluetooth is designed to offer data rates of up to 1 Mbps. This is much slower than the
51911/54 Mbps offered by 802.11 LANs. But 1 Mbps is adequate for applications such as
520keyboards and headsets that use Bluetooth. Bluetooth uses frequency-hopping spread
521spectrum (FHSS) modulation for signal transmission. This is shown in Figure 15.6. The
522sender and receiver communicate at a predefined sequence of frequencies. Any channel
523conflicts only disturb individual transmission blocks, and the bulk of the communication
524generally proceeds without interruption. This makes Bluetooth especially suitable for voice,
525since we generally do not notice brief interruptions.
526Figure 15.6: Bluetooth frequency-hopping transmission
527Each block represents signal transmission at that frequency.
528TIME
5292.480 GHz
5302.402 GHzFrequenc
531y
532472 • Supplementary Chapter 15 / Wireless
533Though Bluetooth has some similarities with wireless LANs, there are some important
534distinctions between the two, such as:
5351. Wireless LANs are largely used by computing devices such as laptops. Bluetooth is
536designed to be used by any electronic device to communicate with any other Bluetoothcapable
537electronic device.
5382. Wireless LANs are typically used to obtain Internet connectivity. Bluetooth is typically
539used to connect to other nearby devices, for example a keyboard to a desktop, or a
540headset to a cell phone. As a result, while high data rate is a very important requirement
541for wireless LANs, it is less important for Bluetooth.
5423. Wireless LANs require an infrastructure of access points. Bluetooth requires no such
543infrastructure. In fact, each Bluetooth device is capable of acting as a Bluetooth access
544point. Bluetooth devices automatically locate other Bluetooth devices in their vicinity.
5454. Devices using wireless LANs are typically connected to power outlets and there are
546no special power-efficiency concerns in wireless LANs. Bluetooth devices are almost
547always driven by battery power and long battery life is an important concern for
548Bluetooth.
5495. Finally, since Bluetooth is often used by devices that are relatively inexpensive, it is
550important for Bluetooth solutions to be extremely inexpensive, generally costing less
551than $10.
552Swedish origins of the term “Bluetooth”
553The development of Bluetooth was initiated in 1994 by the Swedish mobile phone
554company Ericsson, to help laptops make calls using cell phones. The rather unusual
555name comes from King Harald “Bluetooth” Blaatand II of Denmark (940–981 a.d.).
556The nickname came from the king’s love for blueberries, which eventually stained
557his teeth. King Bluetooth unified Denmark and Norway during his reign. Ericsson
558hoped that the technology would similarly unite the communication and computing
559industries.12
560Ericsson also created another well-known technology product in widespread use
561today—the MySQL database. The motivation was to create a small database engine to
562store contact information on cell phones.
563Bluetooth Architecture
564The basic unit of a Bluetooth network is the piconet. A piconet is a collection of devices connected
565to each other using Bluetooth. On the piconet, one master device connects with up to seven
566other active slave devices. A Bluetooth piconet serves a function similar to the basic service
567set (BSS) in 802.11 LANs. However, whereas an 802.11 BSS has a dedicated device called
568an access point that performs various management functions in the BSS, any device in a
569Bluetooth piconet can perform the management functions of a piconet. The device that
57012 Elias M. Awad, Electronic Commerce: From Vision to Fulfillment, 3rd ed. (Pearson, 2007).
571Personal-area Networks (the 802.15 series) • 473
572performs this function is called the master. All other devices in the piconet are called slaves.
573The master provides a synchronization clock that helps all other devices in the piconet
574remain in sync with each other. Whereas a device may be a slave on multiple piconets at the
575same time, it can only be a master on one piconet at a time.
576Many piconets may coexist in the same location. All the co-located piconets are called a
577scatternet. Think of a gathering of tech-savvy students in a classroom, with many students
578carrying Bluetooth-capable cell phones and music players. Each such student forms a
579piconet and the entire classroom becomes a scatternet. Devices connected to two different
580piconets in a scatternet do not have to route packets between the piconets.
581Piconets are the personal-area equivalent of the basic service set. However, there are
582some major differences between basic service sets and piconets. The basic service set has a
583fixed location defined by the geographic area covered by the signals from the access point.
584The piconet, on the other hand, has no defined location. The piconet exists wherever the
585Bluetooth devices go. For example, the Bluetooth devices in a car form a piconet. As the
586car hurtles down the highway, the piconet moves along with it. Also, whereas the basic
587service set can support tens or even hundreds of devices, a piconet can connect at most
588eight devices.
589To enable interference-free communication within co-located piconets, Bluetooth has
590mechanisms that enable each piconet to operate on a different physical channel. Recall from
591Chapter 2 that only one signal may be transmitted at a given frequency at a given location.
592Since all Bluetooth transmissions are in the 2.4 GHz ISM band, Bluetooth needs to create
593mechanisms whereby multiple transmissions can occur at the same frequency.
594Bluetooth creates multiple communication channels at the same frequency by enabling
595devices to transmit at different time slots. Though only one device may transmit at a given
596time in a given location at a specified frequency, different devices may transmit at different
597time slots on the same frequency. Stations keep hopping from frequency to frequency in a
598systematic manner within the 2.4 GHz ISM band. This is called frequency hopping. Devices
599in each piconet use a different hopping sequence, thereby reducing chances of collisions.
600Finally, to maintain confidentiality, devices in each piconet use a different access code and
601header encoding to ensure that even if their signals are received by devices in other piconets,
602the data is unreadable.
603Bluetooth Frame Structure
604The structure of Bluetooth frames is shown in Figure 15.7. The payload header is similar
605to the MAC header of 802.11 wireless LANs. The packet header is analogous to the 802.11
606physical layer header. The channel access code is unique to Bluetooth.
607Figure 15.7: Bluetooth frame structure
608474 • Supplementary Chapter 15 / Wireless
609You may note that the Bluetooth frame has some fields such as flow control, sequence
610number, and channel access code that are absent in wireless LANs. These fields help
611Bluetooth devices operate within a scatternet without interfering with each other. These
612fields also help Bluetooth provide reliable signal transmission for voice applications.
613Bluetooth Device Discovery
614A very special capability of Bluetooth is device discovery. Two Bluetooth devices in close
615proximity to each other will automatically discover each other. If you have used a Bluetoothenabled
616keyboard, you may have noticed this behavior when bringing the keyboard near
617your desktop. Your computer becomes aware of the presence of the keyboard and instantly
618pairs up with it. Device discovery makes Bluetooth extremely user-friendly and eliminates
619configuration-related problems for end users. The devices seem to become aware of each
620other as if by magic.
621To enable device discovery, Bluetooth defines a special channel for inquiry requests and
622responses. Devices that are looking for nearby devices are called inquiring devices. Inquiring
623devices send out inquiry requests on the special inquiry channel. Devices willing to be
624found are called discoverable devices. Discoverable devices listen on the inquiry channel
625for inquiry requests and respond to these requests. Once the two devices become aware of
626each other, the inquiry procedure ends and the connection procedure begins.
627In the connection procedure, one of the devices must be willing to receive a connection
628request from the other device. This device is called the connectable device. The connecting
629device sends a connection request to the connectable device on a connection channel
630specified by the connectable device. According to the Bluetooth standard, the device
631initiating the connection becomes the master for the connection.
632If you think about it, you may notice that the Bluetooth device discovery and connection
633procedure is almost identical to the connection procedures used on social networks such as
634LinkedIn or Facebook. The websites act as the inquiry channel. Without websites such as
635Facebook or LinkedIn, you would not know where to search for your old friends. People
636willing to be found create profiles. People with profiles become discoverable. People
637searching for friends inquire of (search) the social-network site to see if their friend has a
638profile on the site. If the profile is found, the inquiry procedure is over.
639Device discovery—LANs vs. PANs
640Device discovery is not needed in wireless LANs because in most cases user intervention
641is necessary to determine the LAN to connect to. There are also security issues associated
642with wireless LAN membership, as a result of which, network administrators like to
643have control over the users who have access to the LAN. However, once a laptop
644successfully joins a wireless LAN, most laptops offer to join the network in the future
645without user intervention. Therefore, subsequent wireless LAN connections do operate
646in a manner similar to the device discovery procedure.
647Personal-area Networks (the 802.15 series) • 475
648Device discovery in other contexts—the MH 370 disaster13
649This principle of requiring a discoverable device sending out a specific signal whose
650properties are known in advance is common across most contexts. A notable incident
651where this procedure received worldwide attention was associated with the discovery
652of the flight data recorder of the doomed flight MH 370, which vanished without a
653trace on March 8, 2014. Immediately following the accident, search efforts focused on
654detecting the signals emanating from the flight data recorders (black boxes) at 37.5 kHz.
655The black boxes are designed to send out this signal for 30 days at enough strength so
656they can be detected from a distance of up to a mile. Unfortunately, the devices could
657not be located within the appointed time, after which the batteries eventually would
658have died, ending the signal transmission.
659For the connection procedure, you need to send a special connection-request message
660(friend request or invitation) to your friend. If the friend is connectable (responds favorably)
661and accepts your invitation, the two of you become connected.
662WLAN and WPAN Coexistence
663Wireless LANs and Bluetooth operate at the same ISM band (2.45 GHz). Therefore, there
664is a high possibility that the signals from the two technologies may interfere with each
665other. Since Bluetooth is the more recent of the two technologies, it is only natural that
666the designers of Bluetooth had the responsibility of ensuring that Bluetooth minimized
667interference with the existing wireless LAN technology. Therefore, the Bluetooth standard
668defines two mechanisms to minimize interference between 802.11 and 802.15.
669The first of these two mechanisms is collaborative and occurs where Bluetooth and
670802.11 communicate with each other. This is possible when both 802.11 and 802.15
671are present on one device, such as a laptop with both 802.11 and 802.15 capability. In
672the collaborative mechanism, Bluetooth avoids transmission during an ongoing 802.11
673transmission. Alternately, Bluetooth transmits signals on a different channel than the
674channel on which the ongoing 802.11 communication is taking place.
675The second mechanism is non-collaborative. The non-collaborative method is used when
676communication between 802.11 and 802.15 systems is not possible. For example, Bluetooth
677keyboards do not have 802.11 capability and the Bluetooth system on the keyboard has no
678way to collaborate with 802.11. In the non-collaborative method, the 802.15 system senses
67913 https://en.wikipedia.org/wiki/Malaysia_Airlines_Flight_370.
680NFC
681Another technology for communication among proximate devices is called near-field
682communication (NFC). NFC is intended for communication among devices closer
683than 10 cm from each other. Example applications include smartphone payments and
684smartphone check-in.
685476 • Supplementary Chapter 15 / Wireless
686the medium before transmitting. It tries to find a channel in the 2.45 GHz ISM band that
687is not very busy and transmits signals on that channel.
688Bluetooth Categories
689The early Bluetooth specification supported data rates of up to 1 Mbps. However, as
690Bluetooth grew in popularity, newer applications for the technology were identified, each
691with slightly different requirements. As a result, additional categories of Bluetooth have
692been defined as subcategories of Bluetooth. The traditional Bluetooth specification is now
693called 802.15.1.
694The first additional Bluetooth category supports higher data rates. It is useful to be
695able to transfer pictures from digital cameras to computers without the need to take out
696the picture card or connect the camera to the computer using a wire. Since digital images
697can get very large (a compressed picture from a 4-megapixel camera is about 1.5 MB), high
698data rates are very useful for image transfers. Accordingly, the high-data-rate Bluetooth
699specification, 802.15.3, supports data-transfer rates of up to 25 Mbps. This is accomplished
700by improving the efficiency at which the physical layer encodes data into signals, so that
701more data can be sent using the same bandwidth. Currently, higher data rates are becoming
702possible by integrating Wi-Fi with Bluetooth.
703The second additional category of Bluetooth is for remote-control devices such as the
704remote controls for TVs, door openers, fans, lights, etc. These devices need very low data
705rates because, after all, the only data these remotes send is “ON” or “OFF” or “CHANNEL
706= 2.” However, we expect the batteries in remote controls to work for at least a couple of
707years. A unique feature of remote controls is that they are idle most of the time, used only for
708a few seconds in a day to operate remote devices. In almost every case, it is also true that the
709remote control does not need to be a connectable device. Only the controlled device, which
710usually is connected to a power outlet, needs to be connectable. To meet the requirements
711of remote controls, the 802.15.4 standard supports very low data rates, up to 250 Kbps. But
712to achieve long battery life, 802.15.4 devices are not in a connectable state (are switched off)
713when they are not being used. As a result, they do not lose power by scanning the medium,
714listening for other devices that may be interested in connecting to them.
715802.16—Wireless Metropolitan-area Networks
716The final category of wireless networks operating in the ISM band is wireless metropolitanarea
717networks (MANs). A wireless metropolitan-area network (MAN) is a moderately high-speed
718computer network that usually spans a city or large enterprise campus. It is commonly used to
719interconnect LANs within its coverage area to each other and to the Internet.
720Though these networks aren’t well known yet, wireless MANs are being pushed by leading
721computer manufacturers. The professional organization of the wireless MAN industry
722is called the Worldwide Interoperability for Microwave Access (WiMAX). Accordingly,
723wireless MANs are often known as WiMAX networks.
724Wireless MANs have been standardized by the IEEE 802.16 group and, accordingly,
725these networks are also called IEEE 802.16 networks. The 802.16 network was initially
726designed to serve as an alternative to data connections by ISPs over cable and DSL. 
727802.16—Wireless Metropolitan-area Networks • 477
728Since ISPs connect to homes from a central office to a fixed, wired router in the home,
729802.16 was designed to provide connectivity between a fixed-base station (similar to cell
730phone towers) and stationary subscriber stations (similar to TV antennas at some homes).
731The technology was designed to support data rates exceeding 20 Mbps to a range of up to
73210 miles. With the evolution of technology, in 2005 support for mobile stations was added
733to WiMAX as 802.16e. This is a very promising development because 802.16 networks can
734now be used to create wireless “metro zones” to cover large cities and provide metro-wide
735mobile Internet access. Broadband wireless access could be made available to mobile users
736anywhere within the “metro zone,” just as 802.11 enables wireless access in a BSS.
737Most efforts to create city-wide LANs using 802.11 technologies have failed due to the
738high costs of maintaining the large number of 802.11 access points required to provide
739city-wide coverage. 802.16 will enable wireless coverage over the same area using just one
740base station. Perhaps 802.16 will enable the vision of city-wide LANs after all.
741In some developing countries, WiMAX has been used successfully to provide phone
742connectivity to far-flung rural areas. Pulling a wire from a phone company central office
743to a remote rural switch is extremely expensive. Instead, WiMAX can be used to wirelessly
744connect a small rural phone switch to the central office. Homes can be connected to the
745switch using the traditional twisted pair. These links are affordable because they are short.
746In a typical data-communication application, 802.16 provides data rates up to 40 Mbps
747per channel to stations as far as six miles away from a base station. A very popular data rate
748for small businesses is 1.5 Mbps. 802.16 data rates are sufficient to support hundreds of
749such small businesses from a single base station. Alternately, each 802.16 base station can
750be used to support thousands of homes with high-speed Internet access.
751802.16 Data Rates
752Figure 15.8 shows the theoretical data rates supported by 802.16 as a function of distance.
753It also shows the modulation techniques used to achieve the data rates. As expected, stations
754close to the base station can receive the highest data rates (up to 134 Mbps). As the station
755gets farther away, the data rates fall to 90 Mbps. The farthest stations can expect data rates
756of up to 45 Mbps. These are theoretical data rates; actual data rates may be lower. Stations
757closer to the base station use quadrature amplitude modulation (QAM) and farther stations
758use quadrature phase shift keying. You may recall that amplitude modulation generally
759supports high data rates but is very susceptible to noise. It is therefore suitable near the
760base station where signals are strong. Phase modulation offers high immunity to noise but at
761the expense of slower data rates using the same bandwidth. It is useful at longer distances.
762802.16 vs. 802.11
763802.16 networks are likely to be used in a similar manner as 802.11 networks—to obtain
764Internet connectivity on mobile laptops. The two technologies also share other similarities.
765Both are standardized by the IEEE. Both use ISM frequency bands. 802.11 uses the
7662.4 GHz and 5.5 GHz bands; 802.16 networks can use either of these bands.
767In spite of these similarities, there are some significant differences between 802.11 and
768802.16. The primary difference is that, whereas 802.11 access points cover only a radius of 
769478 • Supplementary Chapter 15 / Wireless
770about 300 feet, a single 802.16 base station can cover large parts of a metro area. Another
771difference is that all communication in 802.16 networks is required to pass through the base
772station. Though most 802.11 networks also send all communication through the access point,
773802.11 allows ad hoc networking where stations can directly communicate with each other.
774A very significant difference between 802.16 and both 802.11 and 802.15 is that 802.16
775allows operators to use licensed frequencies if desired. This is to facilitate commercial
776adoption and to improve signal quality. We saw earlier in this chapter that licensed frequencies
777generally have superior signal-transmission properties compared to ISM frequencies. Their
778use is also strictly controlled by the licensee. As a result, 802.16 devices operating on
779licensed bands are likely to see better network performance. The trade-off is that the use
780of licensed bands costs money to compensate the carriers for the costs incurred in acquiring
781the spectrum. Current 802.16 hardware is designed to support signals at 2.5 GHz–3.5 GHz.
782802.16 access over licensed frequencies is likely to cost a subscription fee.
783Satellites—the next frontier in Internet access
784Approximately two-thirds of the world’s population has no access to the Internet. In
7852014, Google announced that it was planning to invest $1 billion to set up a network
786of 180 satellites to bring Internet access to these parts of the world. Other experiments
787include using hot air balloons to transmit signals over large areas. These experiments
788in global connectivity, including Teledesic and Iridium, have failed in the past. Perhaps
789advertising, social media, and the newer revenue models will make this proposed system
790viable.14
79114 Alistair Barr and Andy Pasztor, “Google invests in satellites to spread Internet access,” Wall Street Journal,
792June 1, 2014.
793Figure 15.8: WiMAX data rates
794Example Case—The Oil Industry • 479
795EXAMPLE CASE—The Oil Industry
796When gas prices rose rapidly in recent years, oil companies earned record profits during
7972007–2008. When the economy slowed down in late 2008, oil companies experienced
798slumping demand for the first time in years. Improved supply-chain management using
799computer networks is helping oil companies deal with these boom-and-bust cycles and also
800to lower oil prices.
801Integrated petroleum firms are some of the largest companies in the world. The industry
802had sales of $1.99 trillion in 2008, comparable to US government tax revenues of $2.54
803trillion in 2008. The industry is very volatile, however, with estimated net sales in 2009
804of $1.28 trillion—a drop of more than 35% compared to 2008, due to falling prices and
805reduced demand resulting from weaknesses in the economy. The industry also has very
806low profit margins, with net margins of only 8.1% in 2008. With political sensitivities and
807customer behavior limiting price increases, the way to improve profitability in the industry
808is to lower costs.
809We saw in the Wal-Mart case that utilizing point-of-sale data to optimize distribution
810and manufacturing can eliminate inventory accumulation, reduce waste, and lower costs. In
811most industries, this requires information sharing between many companies that complete
812the supply chain. But the petroleum industry has a unique advantage. It is probably the only
813industry still dominated by vertically integrated firms. Vertical integration refers to a single
814firm controlling all aspects of a product’s manufacture, from raw materials to distribution.
815The major oil companies such as Exxon-Mobil, British Petroleum, Shell, and Chevron own
816or control all factors of production starting from the oil fields where oil is drilled from the
817ground to the gas stations where drivers fill their cars.
818Whereas retailers such as Wal-Mart have to work out legal and technical barriers to
819protect their intellectual property from being stolen by business partners with whom they
820share information, vertically integrated oil companies have a unique advantage. With the
821right systems, they can share point-of-sale data from company-owned gas stations all the
822way up the supply chain to company-owned or company-controlled oil rigs and refineries,
823thereby optimizing drilling, distribution, and storage to lower costs. Publicly available
824information indicates that Chevron is a leader in its industry.
825A large distribution company such as Chevron has a number of places where unnecessary
826costs can add up. For safety reasons, containers such as oil tankers and trucks do not deliver
827unless they can be emptied fully. Ships waiting for storage space to accept their crude can
828pay port charges as high as $30,000 per day. Delivery charges for a truckload of gas can be
829as high as $150. If a truck returns because the gas station is not empty enough to receive
830the entire load of fuel (retain), the company incurs an unnecessary expense. On the other
831hand, if the gas station is out of fuel when a customer arrives (run-out), Chevron could
832lose the customer and earn a bad reputation.
833Chevron uses wireless and satellite networks in many parts of the company to manage
834its supply chain. Many of its gas stations are linked by a satellite network to a central
835dispatch center. These stations have electronic level monitors in the underground gas tanks
836to monitor fuel levels in real time. A wired network in the gas station transports the data to
837the satellite dish on top of the station from where the data is sent to the dispatch station.
838Using this real-time data, the dispatch station is able to optimally schedule fuel deliveries 
839480 • Supplementary Chapter 15 / Wireless
840from terminals located on the outskirts of major metros, minimizing unsuccessful deliveries
841(retains) and stock-outs (run-outs). Chevron can also use the computer network to monitor
842fuel levels in the terminal tanks to schedule oil tanker deliveries so that tankers do not have
843to wait at ports to make deliveries.
844The popular satellite communication technology for retail data-communication
845applications is called VSAT, or very small aperture terminals. This technology uses
846small-sized dish transmitter-receivers communicating with a central hub with a large dish
847transmitter-receiver through a geostationary satellite. The high-performance antenna
848at the hub improves the data-transmission capabilities of the overall network. The lowperformance
849requirements from antennas at each retail outlet such as gas stations reduce
850the costs of the overall system. The data-transfer mechanism in a VSAT system is shown
851in Figure 15.9. All data exchange occurs through the hub. The geostationary satellite acts as
852a broadcast medium for the network. If a station wants to send data to another station in
853a VSAT system, the transmission has to be routed through the hub to get adequate signal
854strength, making two hops through the satellite.
855The satellite network has helped Chevron move toward being a demand-driven company
856where upstream activities are performed only in response to observed demand. The company
857has been moving in this direction since 1997. In 2000, demand-driven operations were
858improving profits by almost 15% in parts of the company where they were being applied.
859Chevron and other oil companies also use computer networks in other parts of their
860businesses. By linking equipment to Internet-based networks, these companies are able to
861centralize all operational information to a central monitoring location. Improved monitoring
862helps companies respond quickly to problems, preventing fires and other hazards, and
863Figure 15.9: VSAT system operation
864References • 481
865improving the uptime of pipelines and storage tanks. Industry experts believe that the
866benefit from these efforts is equivalent to adding 2% to 5% refining capacity.
867Wireless networks are particularly useful to Chevron in its drilling operations. As
868new oilfields are becoming difficult to find, oil companies are focused on improving the
869productivity of existing oil fields. Installing wireless sensors on pumps and other equipment
870allows operators to access maintenance data on all equipment in a location directly from
871their trucks, significantly improving their productivity. Equipment defects that may have
872gone undetected for months can now be attended to in days.
873At remote oil fields, Chevron has used a type of wireless network called a “mesh”
874network to monitor its oil wells. Unlike traditional 802.11 networks where dedicated devices
875act as base stations and routers, in mesh networks each field device acts as both a sensor
876and a wireless router. By placing devices suitably close to each other, each device in a mesh
877network requires very little power because the radio signal from the device has to travel only
878a short distance to the nearby node. The devices can also be designed to transmit only when
879needed, further reducing power requirements and allowing sensor batteries to last for up to
880seven years. For many sensor-deployment projects, wiring costs can be up to 75% of the
881cost of the project. Wireless technologies can eliminate this huge cost.
882References
8831. Davis, A. “Job Losses Cut into U.S. Driving.” Wall Street Journal, Jan. 2, 2010, A3.
8842. Malik, N.S. “Refiners Keep Tab with Real-Time Monitoring.” Wall Street Journal,
885Dec. 23, 2009, B2.
8863. Mir, R.M. “Satellite Data Networks.” http://www.cse.wustl.edu/~jain/cis788-97/ftp/
887satellite_data.pdf (accessed Apr. 9, 2016).
8884. Pister, K., and G. LaFramboise. “Wired Warriors.” www.isa.org.
8895. Value Line, Industry Report. “Integrated Petroleum Firms.”
8906. Worthen, B. “Drilling for Every Drop of Value.” CIO, June 1, 2002.
891Summary
892Wireless networks enable mobility and have become extremely popular in homes and
893businesses. Wireless networks enable inexpensive Internet access in many homes and
894offices. Most wireless networking technologies use ISM frequency bands. Frequencies in
895these bands can be used without cost or licensing restrictions. To meet the requirements of
896different applications that benefit from wireless access, three different categories of wireless
897technologies have been defined.
898The best-known wireless technology is the 802.11 wireless LAN technology, sometimes
899also called Wi-Fi. Wireless LANs use access points to provide high-speed Internet access
900within a range of about 300 feet from the access point. Multiple access points can be
901connected using a distribution system to provide wireless LAN coverage over an arbitrarily
902large area. There are three wireless LAN standards popular today: 802.11a, 802.11b, and
903802.11g. 802.11 technologies can provide network connection speeds of up to 54 Mbps. 
904482 • Supplementary Chapter 15 / Wireless
905The newest wireless LAN technology, 802.11n, is expected to provide data rates of up to
906600 Mbps.
907The second category of wireless networks is personal-area networks, better known as
908Bluetooth. These are standardized by the IEEE as 802.15 networks. Bluetooth provides
909data rates of up to 1 Mbps within a radius of about 30 feet. Bluetooth helps eliminate wire
910clutter created by peripheral devices such as keyboards and mice. The primary design goal
911of Bluetooth is to provide adequate data connectivity while maximizing battery life and
912minimizing costs.
913The last category of wireless networks is IEEE 802.16 wireless metropolitan-area
914networks, also known as WiMAX. These networks can substitute for cable and DSL
915connections and provide high-speed connectivity to fixed receivers at large distances. Mobile
916functionality has recently been added to WiMAX. WiMAX is likely to be offered as a paid
917service in many metro areas in the coming years.
918About the Colophon
919The line in the colophon was uttered by the fox to the little prince in French aviator Antoine
920de Saint-Exupéry’s most famous novella, The Little Prince. The novella is believed to have
921been inspired by the aviator’s real-life experiences in the Sahara desert. It is one of the
922best-selling books ever—80 million copies—and has been translated into more than 180
923languages. Though written and illustrated for children, the book makes many thoughtful
924observations about life. One of the best known of these is “On ne voit bien qu’avec le
925cœur. L’essentiel est invisible pour les yeux,” which translates as, “It is only with the heart
926that one can see rightly. What is essential is invisible to the eye.”
927Computers and networks have no heart. At the heart of their operations though, are
928properties of the universe that are invisible to the eye. Computer networks do not need a
929visible medium to transport data. The properties required from nature to support signal
930transmission are invisible to the naked eye. The invisible outer space can transport data just
931as effectively as visible wired networks. In an earlier age, this ability of the universe to carry
932electronic signals was given a name—ether.
933REVIEW QUESTIONS
9341. What are wireless networks? Why are they useful?
9352. Some cities took up projects to set up wireless LANs all over the city. Read about the
936project taken up by one such city. Was the project a success? Why or why not?
9373. What are some of the concerns with using wireless networks?
9384. What are ISM frequency bands? Why are they useful?
9395. What are some differences between wired and wireless LANs? How do they impact
940the design of the wireless LAN header?
9416. What is a basic service set? A basic service area?
9427. What is an access point? What are some reasons why you would prefer access points to
943wireless routers when creating a wireless network in your organization?
9448. What is a distribution system in wireless LANs?
9459. What is an extended service set?
946Example Case Questions • 483
94710. What is a portal in a wireless LAN?
94811. What are some differences between the physical layers in wireless and wired LANs?
94912. What are the common wireless LAN categories? What are the important differences
950between them?
95113. What is 802.11n? What are some likely advantages of 802.11n over the traditional
952wireless LANs? How does 802.11n obtain these advantages?
95314. What are personal area networks? How are they different from LANs?
95415. What are some important characteristics of Bluetooth?
95516. What is a piconet? What are some differences between a piconet and a basic service set?
95617. What are master and slave devices in a piconet?
95718. What is a scatternet?
95819. What are some advantages of having distinct physical channels in Bluetooth?
95920. Why is device discovery useful in Bluetooth? How is device discovery accomplished?
960Why is device discovery not needed in wireless LANs?
96121. Describe the mechanisms that have been defined for WLANs and WPANs to coexist
962at the same frequency bands without interfering with each other.
96322. What are the different categories of Bluetooth? What are they used for?
96423. What are wireless MANs? What are their primary uses?
96524. What data rates and ranges are likely to be available on wireless MANs?
96625. What are the differences between wireless LANs and wireless MANs?
967EXAMPLE CASE QUESTIONS
9681. What are the different kinds of wireless data communication technologies used in the
969case?
9702. What is a retain in the context of supply chains?
9713. What is a run-out in the context of supply chains?
9724. What is a mesh network in the context of wireless sensor networks? What are its
973advantages and disadvantages? (Wikipedia is a good resource.)
9745. Why do companies with a nationwide footprint often use satellite-based data networks
975for data transmission instead of wired networks such as DSL?
9766. A leading provider of satellite-based data communication services is DirecPC. Visit the
977company’s website and write a short (one-paragraph) report on the services offered by
978the company based on information provided at the website. Include information such
979as data rates, plan prices, and other information relevant to new subscribers.
980HANDS-ON EXERCISE—AirPCap Wireshark Captures
981You have used Wireshark to capture packets on your local computer. In this exercise, you
982will use two Wireshark captures on the wireless interface to visualize the operation of IEEE
983802.11 networks. Both captures show the download of a web page over a wireless LAN.
984The network topology of the setup for the capture is shown in Figure 15.10.
985The two captures are included in the readings for this chapter on the companion
986website—Wireshark HTML capture—802.11-header only (called “802.11 only” in this 
987484 • Supplementary Chapter 15 / Wireless
988exercise) and Wireshark HTML capture—802.11 header and radio header (called “radio
989header” in this exercise). These captures have been made using AirPCap, which includes
990the hardware and software required to capture information specific to the IEEE 802.11
991protocol. You may need to look at online resources to answer some of these questions from
992the 802.11-only capture:
9931. Look at the first frame. The information field for this packet states that this is a beacon
994frame. What is the role of a beacon frame in IEEE 802.11? Why is this frame not
995necessary in IEEE 802.3 Ethernet?
9962. Which device on the wireless LAN sends out the beacon frame? Based on this
997information, what is the MAC address of the wireless router (which is also used as an
998access point for this capture)?
9993. Based on the above question and the information in the beacon frame header fields,
1000what information serves as the basic service set (BSS) identifier?
10014. The second packet in the capture is a probe request. What is the role of a probe-request
1002frame in IEEE 802.11? Why is this frame not necessary in IEEE 802.3 Ethernet?
10035. Which device(s) send(s) out probe requests?
10046. What is the BSS ID of the destination in the probe request? What does this number
1005signify?
10067. How are frames identified as beacon frames or probe-request frames or data frames?
1007(Hint: look at the type/subtype field.)
10088. Examine the MAC address fields in a few frames. What are the three MAC addresses
1009included in all frames?
10109. Recalling the wired Wireshark captures, there were only two MAC addresses in the
1011Ethernet header: source and destination. Why is it necessary to include a third MAC
1012address, the BSS ID, in 802.11 frames?
1013Now, answer the following questions from the radio-header capture:
10141. Select any frame in the capture and expand all the subheaders of the radiotap header
1015(e.g. present flags and flags). What is the channel frequency at which the frame was
1016transmitted?
1017Figure 15.10: AirPCap capture topology
1018Wireless router
1019MAC address:00-21-91-E6-35-DB
1020Laptop running
1021web browser and
1022Wireshark with AirPCap
1023MAC address: 00-19-7E-30-0A-2B
1024Laptop running
1025Apache web server
1026MAC address: 00-0E-9B-C3-E0-95
1027Critical-Thinking Exercise—Ubiquitous Wi-Fi • 485
10282. Briefly describe the channels used by 802.11b/g.
10293. Why is channel 6 one of the recommended channels for transmitting 802.11 wireless
1030LAN data?
10314. Was the frame transmitted using FHSS (Frequency Hopping Spread Spectrum) or
1032OFDM (Orthogonal Frequency Division Multiplexing)?
1033CRITICAL-THINKING EXERCISE—Ubiquitous Wi-Fi
10341. Technologies such as 802.11ac are bringing wireline speeds to Wi-Fi. Do you think
1035Wi-Fi will eliminate the need for wired LANs within the next decade? Why or why not?
1036IT INFRASTRUCTURE DESIGN EXERCISE—Add Wi-Fi
1037The employees at the Amsterdam service center use laptops for work and need wireless
1038coverage throughout the two floors of the building. The company has therefore decided to
1039install a wireless LAN at this location. (In general, the company will also create an additional
1040subnet for this wireless network, but you can ignore this detail for this class.)
1041Answer the following questions:
10421. What wireless technology would you recommend to create the wireless LAN—
1043IEEE802.11a, IEEE802.11b, IEEE803.11g, or IEEE802.11n? Justify your choice.
10442. Assume that both floors of the building have the same dimensions. Making typical
1045assumptions about the needed work space for each employee, what is the total area in
1046the building that needs wireless coverage?
10473. Given your technology choice and the area calculated above, how many access points
1048would you need to provide satisfactory coverage, where needed, on both floors?
10494. Update your network diagram to include the portal for the wireless network in
1050Amsterdam.
1051_____________________________________________________________________
1052Chapters 15 and 16 are online supplements to Business Data Communications and IT
1053Infrastructures, 2nd Edition, by Manish Agrawal & Rekha Sharma. © 2017, Prospect Press
1054
1055487
1056S U P P L E M E N T A R Y
1057CHAPTER 16
1058Phone Networks
1059Mr. Watson—come here—I want to see you.
1060Alexander Graham Bell
1061Overview
1062The earlier chapters have focused on the technologies that enable the Internet. However,
1063a large volume of information exchange occurs over a more humble technology—the
1064telephone. Toward the end of the 20th century, phone networks also served as the precursors
1065of modern WAN networks. In recent years, the telephone has evolved with the introduction
1066of cell phones, which add mobility to phone technology. Given the importance of the
1067technology for modern businesses, it is useful to know about the technologies that underlie
1068land and cellular phone networks. At the end of this chapter you should know about:
1069• the architecture and components of landline phone networks
1070• the signals used in phone networks
1071• DSL—an early technology developed by phone companies to offer high-speed
1072Internet access
1073• the architecture and components of cellular networks
1074• the evolution of cell phone networks
1075• code division multiple access—the signaling scheme used in modern cell phone
1076networks
1077Introduction
1078Phone networks served as the early access mechanism for the Internet. In the 1990s,
1079as Internet service providers such as AOL were perfecting their business models and
1080technologies (recall the reference to AOL’s growth), the phone network served as the
1081medium over which most residents obtained their Internet connections. Only in the early
1082years of the 21st century have carriers installed dedicated networks for Internet service. It is
1083therefore useful to have a high-level understanding of the architecture of phone networks.
1084The phone network is best seen as an analog information-transmission system. Signals
1085are not digitized or packetized by landline phones. Instead, signals are simply allowed to flow 
1086488 • Supplementary Chapter 16 / Phone Networks
1087over the wires in the same form in which they are generated at the source. In this chapter, we
1088will look at the components of the phone system, and some significant events that helped
1089the phone system evolve to its current form.
1090The data communication timeline shows that for more than 100 years (1840–1969),
1091the only technologies available for long-distance information exchange were the telegraph
1092and the telephone. Of the two, the telephone continues to be widely used globally. The
1093basic technology underlying the phone system is quite simple. When a user dials a phone
1094number, the switches in the phone network establish a dedicated circuit from the sender to
1095the receiver so that signals can flow uninterrupted between the sender and the receiver. As
1096long as the call is in progress, the network resources allocated to the call are unavailable to
1097other callers. This mechanism of connecting users is called circuit switching. Circuit switching
1098is a process that, on demand, connects two or more communicating devices and permits the exclusive use of
1099a data circuit between them until the connection is released.
1100Once the circuit is established, a microphone in the handset converts the talker’s voice
1101to electronic signals. These signals are carried over the phone network from the sender to
1102the receiver. A speaker in the receiver handset converts these signals to voice that is heard
1103by the receiver.
1104Since the phone technology is so simple, particularly compared to the more complex
1105packet-switched technologies used on the Internet, the phone network is sometimes also
1106called plain old telephone service (POTS), with the term “plain old” implying that the
1107phone is a plain vanilla kind of service. However, such descriptions should not lead us to
1108underestimate the importance of phone service. Though the phone technology is quite
1109simple, it is one of the most important technologies for businesses. Figure 16.1 shows
1110the levels of adoption of landline phones in different parts of the world.1
1111 As seen in the
1112figure, over the last decade, landline adoption has been relatively steady around the world.
1113Assuming three inhabitants per family, there would be approximately 35 families for every
1114100 inhabitants (100/3) and with about 50 phones for every 35 families, the figure shows
1115that on average, there is more than one phone line per family in the developed world.2
1116Phone Network Components
1117The phone network is called the public-switched telephone network (PSTN). For basic
1118understanding, the components of the network are shown in Figure 16.2. The PSTN
1119terminates on the walls of end-user homes and offices. Customers maintain the phone
11201 www.itu.int/ITU-D/ict/statistics/ict/
11212 Since the statistics also include business phones, it is fair to say that most homes in the developed world have
1122a landline phone connection.
1123Why didn’t the telephone pass eighth grade?
1124Because it wasn’t a “Smart Phone.”
1125Source: Boys’ Life magazine, July 2012
1126Phone Network Components • 489
1127networks within their own premises. Between the customer premises and the central office
1128is a very visible and critical component of the phone network—the local loop. The local loop
1129is a circuit from the customer premises to the last switch of the phone company’s network. The local loop
1130ends at the end office or central office of the phone company. The end office, also called the
1131central office, is the location where the phone company operates equipment that is responsible for providing
1132the customer’s dial tone. The last switch of the PSTN is in the end office.
1133If a call is made to a user connected to the same end office, it is connected to the user by
1134the end office. The user then hears the phone ring. If the call is being made to a user who
1135is connected to a different end office, the call is forwarded to the appropriate end office
1136for completion. Figure 16.2 shows this as a link to the interexchange carrier (IXC). IXCs
1137are networks that carry traffic between end offices. When the remote end office connects the call,
1138there is a completed circuit between the sender and receiver over which signals can flow.
1139The IXC link in Figure 16.2 has some special properties. This link is where phone
1140networks get integrated with WAN technologies. It may be noted that whereas the local loop
1141is dedicated to a single user, the IXC link is shared among all users of the end office. Since
1142the end user can only make one phone call at a time, there can only be one signal at any
1143given time on a local loop. However, at any given time, multiple customers of an end office
1144may need connections to the IXC to get connected to users in other end offices. Hence,
1145the phone network needs a multiplexing technology in the IXC link to enable multiple
1146users to simultaneously share the common IXC link. This can be accomplished using WAN
1147Figure 16.1: Landline adoption
1148490 • Supplementary Chapter 16 / Phone Networks
1149technologies. Hence, the IXC link shown in Figure 16.2 is where WAN technologies such
1150as ATM were first deployed for widespread use.3
1151Phone Signals
1152Voice is carried over the phone wires as signals. In this section, we will take a high-level
1153look at the properties of phone signals. We have avoided getting into the details of signals
1154in this book because the information can get very technical. However, some aspects of
1155phone signals are useful to know because they define the capabilities of phone modems.
1156The phone system tries to reproduce the speaker’s voice at the receiving end. This can
1157be accomplished if all the tones in the speaker’s voice can be captured and transmitted to
1158the receiver. Tones are determined by signal frequencies. Frequencies are measured in hertz.
1159Hertz is a unit of frequency that is equivalent to one cycle per second. Any operation that repeats once
1160per second is said to operate at 1 Hertz. An operation that repeats twice per second is said
1161to operate at 2 Hertz, and so on.
1162Observations suggest that the human voice lies in the frequency range 80 Hz to about
11631,100 Hz. Male voices typically lie at the lower end of this range and female voices are at the
1164higher end of this range. A system that can capture, transmit, and reproduce all the signals
1165in the range 80 Hz–1,100 Hz should therefore be able to reproduce voice with high fidelity.
11663 A pretty infographic that shows the evolution of the phone network gained widespread coverage along with
1167the introduction of Google Voice, “A modern history of human communication,” http://visual.ly/modernhistory-human-communication
1168(accessed Dec. 2015).
1169Figure 16.2: Phone network components
1170Legal Developments • 491
1171The phone system has been designed to transmit signals in the 300 Hz–3,400 Hz range.
1172Since 3,400 is greater than 1,100, the high end of the range captures all human voice at
1173high frequencies. However, the cutoff at the low end, at 300 Hz, causes some loss of
1174information. So, why does the phone system not capture the signals in the frequency range
117580 Hz–300 Hz? Why are we willing to lose the information in these frequencies? The
1176reason to eliminate lower-frequency signals is that the power line transmits power at 60 Hz.
1177Filtering away signals below 300 Hz eliminates the strong hum that is likely to be created
1178in the phone receiver by the power line and its harmonics. Transmitting signals within the
1179frequency band of 300 Hz–3,400 Hz has been found to be adequate to convey the human
1180voice over the telephone. While this does not create a high-fidelity (hi-fi) reproduction of
1181the sender’s voice, the sender’s voice is clearly recognizable over the phone network.
1182The phone system only needs to transmit a signal within a relatively narrow bandwidth
1183of about 3 kHz (3,400 Hz - 300 Hz = 3,100 Hz). The transmission of phone signals does
1184not require a very high quality of wire in the local loop. As a result, the phone network
1185was built using relatively inexpensive copper cables (inferior in quality even to Cat3 cables)
1186in the local loop. This kept costs low and served the phone network well throughout the
118720th century. By comparison, Ethernet signals require thousands of times the bandwidth
1188for signal transmission. When customers began to demand broadband connections in the
1189last decade of the 20th century, the poor signal-carrying properties of Cat3 phone cables
1190in the local loop became a stumbling block for phone companies. Serving high data rates
1191requires transmission of high-bandwidth signals, something the local loop could not do. To
1192overcome this limitation, phone companies could either invest in a new local loop or figure
1193out a way to do the best they could with the existing local loop. Phone companies therefore
1194developed DSL, a technology that allowed high-data-rate signals to be carried over phone
1195lines for short distances.
1196Legal Developments
1197The evolution of the phone network in the United States has been influenced not just by
1198competitive forces, but also by legal action taken by the government.4
1199 Alexander Graham
1200Bell had created AT&T to commercialize his invention of the telephone. Over the years,
1201AT&T invested in creating a nationwide telephone network. Many residents were located
1202in remote areas. Even though it was very expensive to connect these users to the telephone
1203network, AT&T provided these customers with phone service at the same rates as urban
1204customers. To facilitate these investments, in the early years of the 20th century, the US
1205government allowed AT&T to operate as a regulated monopoly. The monopoly status
1206guaranteed the company suitable returns on its investments in developing a national network
1207that provided phone service to all residents in the United States. This was a lot like the way
1208most states allow gas and electric utilities to operate as regulated monopolies. The utilities
1209agree to invest in providing services to all residents, however inaccessible they might be.
12104 For a very informative graphic that shows the evolution of the phone network in the United States, from
1211Graham Bell to the contemporary cell phone providers, please see the graphic, “A tangled family tree,” Wall
1212Street Journal, Mar. 29, 2011, http://www.wsj.com/articles/SB10001424052748704471904576229250860034510
1213(accessed Feb. 2016).
1214492 • Supplementary Chapter 16 / Phone Networks
1215In return, the state protects them from competition (monopoly) and allows them to earn
1216reasonable (regulated) profits.
12171984—Competition in Long-distance Phone Service and the
1218Modified Final Judgment
1219By the mid 1970s, technological innovations had led to the emergence of a number of
1220competitors providing long-distance phone service. There was no longer any need to
1221provide monopoly protection to AT&T in long-distance phone service. To level the playing
1222field for competitors in the long-distance phone business, the US government filed an antitrust
1223lawsuit against AT&T in 1974. The lawsuit ended in 1982 when AT&T and the US
1224government agreed on the terms of a solution to enable competition in long-distance phone
1225service. As part of the agreement, seven regional phone companies that offered local phone
1226service were divested (separated) out of the old AT&T. AT&T continued in business as a
1227long-distance phone company, competing with other firms such as MCI. The seven local
1228phone companies were called Regional Bell Operating Companies (RBOCs). These RBOCs
1229were granted regulated monopoly status in providing local phone service, as the government
1230believed that this was necessary to ensure continued investments in operating the local loop.
1231The RBOCs were prohibited from providing long-distance service. This new structure was
1232put in place effective January 1, 1984. Figure 16.3 shows a map of the operating areas of
1233the seven RBOCs in the United States after the breakup of AT&T.
1234Early mis-evaluations of telephony technology
1235In 1876, Western Union, the largest American telegraph company, refused to buy
1236Graham Bell’s patent for $100,000, arguing that people are not savvy enough to handle
1237a phone: “Bell expects that the public will use his instrument without the aid of trained
1238operators. Any telegraph engineer will at once see the fallacy of this plan. The public
1239simply cannot be trusted to handle technical communications equipment.”
1240A group of British experts thought somewhat differently: “The telephone may be
1241appropriate for our American cousins, but not here, because we have an adequate supply
1242of messenger boys.”5
1243Note: These statements were written so beautifully in the original, they have not
1244been paraphrased.
1245The judgment that led to the divestiture of the seven RBOCs in 1984 is sometimes called
1246the modified final judgment. The name relates to a judgment in 1956 from an earlier antitrust
1247lawsuit filed against AT&T. As a result of the 1956 judgment, AT&T agreed to restrict
1248its activities to running the national telephone system and performing other government
1249work. The 1956 judgment is referred to in the industry as the final judgment. The decision
1250in 1984 modified the final judgment from 1956, and is therefore called the “modified final
1251judgment” in the telecom industry.
12525 Gerd Gigerenzer, Risk Savvy: How to Make Good Decisions (Penguin, 2015).
1253Legal Developments • 493
12541996—Competition in Local and Long-distance Phone Service:
1255Telecommunications Act
1256There were rapid developments in the telecom industry following the modified final
1257judgment. Competitors entered into the long-distance telephony market. New technologies
1258were beginning to be deployed for Internet service. Competitive markets reduced the need
1259for government to impose price controls, but prior regulatory barriers had to be removed
1260to allow free entry of competitors for providing current and emerging telecommunications
1261services. To respond to the changes in the industry from these developments, in 1996 the US
1262Congress passed the Telecommunications Act. The website of the Federal Communications
1263Commission states that “the Telecommunications Act of 1996 was the first major overhaul
1264of telecommunications law in almost 62 years. The goal of this new law was to let anyone
1265enter any communications business—to let any communications business compete in any
1266market against any other.” Specifically, for our purposes, the major provision of the act
1267was that it introduced competition in both local and long-distance phone service. Thus,
1268whereas a judicial decision introduced competition in long-distance phone service in 1984,
1269the legislative process introduced competition in both local and long-distance phone service
1270in 1996 through a new law, the Telecommunications Act of 1996.
1271The Telecommunications Act of 1996 had other features that affected the phone
1272industry. In a very novel provision, in order to create competition in local phone service, the
1273act required RBOCs or other local phone companies to provide access to their networks at
1274reasonable rates to competitors who wanted to provide local phone service. This provision
1275created two classes of local phone companies—the incumbents and the competitors.
1276Figure 16.3: Map showing operating areas of the seven RBOCs in 1984
1277494 • Supplementary Chapter 16 / Phone Networks
1278The incumbents (the RBOCs) were called the incumbent local exchange carriers (ILECs).
1279The competitors were called the competitive local exchange carriers (CLECs). You may
1280remember regularly receiving solicitations from startup phone companies offering phone
1281services at very competitive rates in the late 1990s and early 2000s. These startups were the
1282CLECs. The law also allowed the RBOCs to offer long-distance phone service.
1283Digital Subscriber Line (DSL)
1284We saw earlier that when customers began to demand broadband Internet access, the
1285limitations of the phone network immediately became apparent. The local loop in the
1286phone network was designed to carry narrow band signals in the range 300 Hz–3,400 Hz.
1287Broadband signals need cables capable of carrying much higher bandwidths than what the
1288local loop was designed to carry.
1289So, when users began to demand broadband connections, the phone companies were at a
1290disadvantage compared to cable companies. The coaxial cable used by cable TV companies
1291is capable of carrying signals over a very high bandwidth. Cable companies were therefore
1292capable of offering broadband Internet access to customers using their existing network
1293infrastructure. Cable companies began to offer service packages that bundled cable TV,
1294phone, and broadband Internet access. If replacing the cables in the local loop is extremely
1295expensive, how were phone companies going to compete with cable companies once
1296broadband Internet access became important to customers?
1297To respond to this business need, phone companies developed the digital subscriber
1298line (DSL) technology. DSL is technology that provides full-duplex service on a single, twisted, metallic
1299pair of phone wires at a rate sufficient to support basic high-speed data service. Phone companies
1300realized that though Cat3 cables could not carry high-bandwidth signals over long distances,
1301they could carry these signals over short distances, say up to three to five miles. From a
1302business perspective, this was very useful. In densely populated areas (e.g. places with many
1303apartment complexes), phone companies could create end offices in a central location and
1304offer DSL services to as many nearby customers as possible. By suitably locating end offices,
1305DSL could allow phone companies to offer broadband Internet access to a large number
1306of customers.
1307When the phone line carries DSL signals, the signals are transmitted in the frequency
1308ranges shown in Figure 16.4. The phone signals are carried as before in the 300 Hz–3,400
1309Hz range. The upstream (upload) DSL signals are carried in the frequency band 25.875
1310kHz–138 kHz. The downstream (download) signals are carried in the frequency band 138
1311kHz–1,104 kHz. For reference, Figure 16.4 also shows the frequencies used by the phone
1312channel (the narrow column at the extreme left, identified as “phone channel” in the figure).
1313Why was the cell phone wearing glasses?
1314Because it lost its contacts.
1315Source: Boys’ Life magazine, March 2012
1316Digital Subscriber Line (DSL) • 495
1317We see from Figure 16.4 that the upstream bandwidth is about 112 kHz (138 - 25.875),
1318while the downstream bandwidth is about 966 kHz (1,104 - 138). Thus, the downstream
1319bandwidth is more than eight times the upstream bandwidth. Why this asymmetry? Is there
1320something special about the downstream signal that it requires such a high bandwidth
1321compared to the upstream bandwidth?
1322The large downstream bandwidth has to do with typical customer behavior. Recall the
1323relationship between bandwidth and data rates. For the same signal and noise levels, the
1324data rate of the channel increases in direct proportion to the bandwidth of the channel.
1325Most end users download huge quantities of data, but upload very little data. Downloads
1326include video, web pages, and other Internet content. The most common data uploaded by
1327end users is e-mail. The total bandwidth of the upstream and downstream channels is fixed
1328(1,104 - 25 = 1079 kHz). Since most end users care most for high-speed downloads and not
1329so much for high-speed uploads, most Internet service providers engineer their systems to
1330provide acceptable upload speeds and the highest download speeds possible.
1331You may note a slight gap in Figure 16.4 between the phone channel and upstream DSL
1332channel. This gap is deliberate. End users of DSL service install a DSL splitter that separates
1333out the phone and DSL signals on the cable. The gap between phone and DSL channels
1334helps the DSL splitter separate out the phone and DSL signals.
1335As an analogy for how Cat3 cables can carry high bandwidth signals over a short distance,
1336but not long distances, imagine driving a Corvette over a dirt road. You will be able to
1337reach high speeds over short distances. But if you try to maintain the high speed over long
1338distances, you will end up with a sprained back, a damaged vehicle, or both. A Cat3 cable is
1339like a rough road for signals. You can carry low-speed signals for long distances, but highspeed
1340signals can only be carried for short distances.
1341Figure 16.4: Phone and DSL signal frequencies
1342496 • Supplementary Chapter 16 / Phone Networks
1343Cell Phones
1344Cellular telephony is a mobile communications system. It uses a combination of radio transmission and
1345conventional telephone switching to permit mobile users within a specified area to access full-duplex telephone
1346service.
1347The rapid adoption of cellular telephony is one of the most important developments in
1348telecommunication technology in the first decade of the 21st century. Figure 16.5 shows
1349how cell phones have rapidly become popular in most regions of the world. There is almost
1350one cell phone for every individual in the developed world. In developing countries, cell
1351phone adoption seems to be rising at an even faster rate than in developed countries, and
1352there is now approximately one cell phone per family in the developing world (approximately
135345 cell phones for every 35 families).
1354To enable cellular telephony, large geographical areas are segmented into many smaller
1355areas. Each small area is called a cell. Each cell has its own radio transmitters and receivers
1356and a single controller interconnected with the public-switched telephone network.
1357Since one cell phone tower can provide phone coverage to a wide area, and laying landlines
1358to each home can get very expensive, building a cell phone infrastructure can sometimes
1359actually be cheaper than building a landline infrastructure. Thus cellular telephony can help
1360developing countries build nationwide phone networks at lower costs than is possible with
1361conventional landline telephony.
1362Figure 16.5: Cell phone adoption
1363Cell Phones • 497
1364Cell Phone Technology Evolution
1365Commercial cell phone networks have been available in different parts of the world from the
1366early 1980s. Since then, there has been steady evolutionary improvement in the technology.
1367There have also been two clearly identifiable revolutionary improvements in cell phone
1368technologies. Each revolutionary improvement in cell phone technology is labeled as a
1369generation in the telecommunications industry. Figure 16.6 shows the key features of each
1370generation of cell phones.6
1371The earliest cell phone networks were built starting around 1980 and are now called
1372first-generation cell phone networks. These phones used analog signals, using frequency
1373modulation to transmit the speaker’s voice over the allocated carrier frequency. This is a very
1374simple technology that had been used for more than 50 years for wireless communication
1375in applications such as car radios. Though the phones using this technology were heavy and
1376unwieldy, did not support data transmission, and used wireless frequencies very inefficiently,
1377the first generation set the stage for future development of the technologies and markets.
1378First-generation cellular phone networks were popular throughout the 1980s.
1379In the first revolutionary change, digital signals replaced analog signals to carry voice.
1380Networks and equipment using this technology are called second-generation cell phone
1381networks. The primary advantage of using digital signals is that using data-compression
1382techniques, it is possible to send multiple digital signals using the same bandwidth used
1383by one analog signal. Thus, second-generation networks made more efficient utilization
1384of bandwidth compared to the first-generation cell phone networks. Second-generation
1385networks began to be deployed just after 1990. Since the second generation used digital
1386signals, second-generation networks also added packet-data service, enabling cell phones to
1387be used as data modems at data rates of approximately 15 Kbps.
1388The current cell phone networks are called third-generation cell phone networks.
1389Specifications for the third generation of cell phone networks were defined under the
1390leadership of the International Telecommunications Union (ITU) in 2000. The revolutionary
1391feature of third-generation cell phone networks compared to second-generation networks
1392is much higher data rates. Both second- and third-generation cell phone networks use
1393digital signals and both support voice and data. Thus, the distinction between secondgeneration
1394and third-generation cell phone networks is not as distinct as the distinction
1395between the first and second generation of cell phones. However, as the ITU states, the
13966 M. Agrawal, K. Chari, and R. Sankar, “Demystifying Wireless Technologies: Navigating Through the Wireless
1397Technology Maze,” Communications of the AIS, 12 (2003): 166–182. (Excerpt used with permission from the
1398Association for Information Systems, Atlanta, GA. All rights reserved.)
1399Ethan: Two antennas met on a roof, fell in love, and got married.
1400Grace: How was the ceremony?
1401Ethan: I don’t know, but the reception was terrific!
1402Source: Boys’ Life magazine, July 2014
1403498 • Supplementary Chapter 16 / Phone Networks
1404third-generation cell phone networks “raised the bar.” Third-generation networks offer
1405performance levels significantly in excess of those obtainable from second-generation (2G)
1406cell phone networks. In particular, minimum data speeds for various environments are
1407defined for third-generation cell phone networks.7
14087 www.itu.int/ITU-D/imt-2000/DocumentsIMT2000/What_really_3G.pdf.
1409Figure 16.6: Cell phone technology evolution
1410Cell Phones • 499
1411In general, third-generation cell phone networks are designed to offer data rates that
1412are sufficient to simultaneously support voice and high-speed data communication. Thirdgeneration
1413(3G) networks are expected to offer minimum speeds of 2 Mbps for stationary
1414or walking users, and 348 Kbps in a moving vehicle. By comparison, second-generation
1415systems only provided data rates in the range of 9.6 Kbps–28.8 Kbps. Third-generation
1416networks help provide desktop-like network experience on smart phones such as the iPhone.
1417Figure 16.6 summarizes the evolution of cell phone networks. There were two early cell
1418phone services—TACS and AMPS. All current cell phone networks have evolved from these
1419services. The figure also shows the important evolutionary technologies that were introduced
1420between generations. For example, before 2G was deployed, some operators were already
1421experimenting with a technology called the intermediate system. This technology introduced
1422digital signals and TDM over FDM, which later became the accepted standard for 2G.
1423Similarly, before high-data-rate 3G networks were deployed, cell phone carriers
1424introduced higher data rates than were available from 2G, though less than proposed 3G
1425speeds. The first of these networks was called General Packet Radio Service (GPRS). These
1426networks offered data rates of about 128 Kbps (compared to about 14 Kbps in 2G). The
1427128-Kbps GPRS is popularly called 2.5G. GPRS evolved to EDGE (Enhanced Data rates
1428for Global Evolution), offering data rates of 384 Kbps. EDGE is popularly called 2.75G.
1429One interesting technological feature of 3G cell phone networks is that all 3G networks
1430are built using a form of multiplexing that we have not seen before. This is called code
1431division multiple access, or CDMA. To support a large number of wireless users, with
1432each user transmitting at very high data rates as required by the 3G standard while using
1433the limited wireless bandwidth that is available, requires multiplexing technology that uses
1434bandwidth very efficiently. Fortunately, CDMA is such a technology. CDMA is discussed
1435later in this chapter.
1436Cell Phone System Architecture
1437The architecture of the cell phone system is shown in Figure 16.7. The service area of the
1438mobile-phone network is divided into small areas called cells. Cells can be of different sizes
1439but it is quite common for cells to be approximately two to four miles in diameter. Each
1440cell is served by a base station that houses antennas and other electronic equipment to send
1441and receive signals from end-user devices (cell phones) within the cell. Base stations in an
1442area are connected to a mobile-telephone switching office (MTSO). The MTSO connects
1443the cell phone network in its area to the PSTN, or the landline phone system, through a
1444connection to a nearby phone exchange. This way, phone calls can be seamlessly connected
1445between landline and cellular phones.
14464G LTE
1447Most users now use 4G networks that are sometimes marketed as LTE (long-term
1448evolution) networks. The 4G standard specifies peak downlink data rates of 300 Mbps
1449and peak uplink data rates of 75Mbps.
1450500 • Supplementary Chapter 16 / Phone Networks
1451Frequency Reuse
1452Why do we divide the service area into cells? Why can’t we just have one tower to serve all
1453customers in an entire metro area? Such a system would cost far less than the current system
1454that uses multiple towers to cover a single metro area. At first glance, it might appear that
1455multiple cells are created to provide strong signals. We saw the case of 802.11 networks in
1456Chapter 15, where signals from a base station only reach about 100 meters away. So, is the
1457reason because, similarly, signals from a cell phone base station can only reach about one to
1458two miles in each direction from the base station, for a cell diameter of two to four miles?
1459Actually, this is not the case. The frequencies used for cell phone communication have
1460excellent propagation properties. Signals from the base station can reach as far as 40 miles
1461away. Therefore, a single cell tower can indeed serve an entire metro area. In fact, base
1462stations are designed to limit the signal range within the boundaries of the cell served by
1463the base station. Then why do we create cells and deliberately raise the costs of the cell
1464phone system?
1465Figure 16.7: Cell phone system architecture
1466Tower locations
1467Figure 16.7 shows towers in the middle of each cell, although in practice, towers are
1468generally at the cell intersections, with each tower carrying multiple antennas, and each
1469antenna serving one neighboring cell.
1470Cell Phones • 501
1471Tower accidents
1472The buildout and maintenance of the cell phone tower network may be one of the most
1473dangerous contemporary industries. There were 10 fatalities in 2014 and 14 in 2013.
1474This number fell to 4 in 2015.8, 9
1475There are also concerns about radiation exposure to workers who work near these
1476towers for extended periods, as well as residents directly in front of these antennas.
1477As of 2014, an estimated 10% of towers violated regulations regarding barricades and
1478signage near towers warning people of possible danger from exposure.10
1479Yet another hazard in working on cell towers is that ospreys like to use these towers
1480as nesting spots because of the clear view of hunting spots in the distance. The nests
1481can weigh as much as half a ton, and their presence slows down construction work.
1482Ospreys are protected birds, and there are strict regulations on how their nests can be
1483handled.
1484The cellular design of the cell phone system is primarily motivated by the need to
1485serve a large number of users using the limited amount of wireless bandwidth that is
1486allocated for cell phone service. We have seen that spectrum auctions are the mechanism
1487by which service providers obtain bandwidth, and frequencies are extremely expensive
1488to obtain at these auctions. To efficiently
1489use the available bandwidth, cell phone
1490networks reuse frequencies across cells.
1491We know that a single frequency can
1492serve one customer at a time in any single
1493area covered by one cell phone tower. By
1494reusing the same frequency in different
1495nonadjacent towers, the same frequency
1496can serve multiple customers within a
1497metro region. Frequencies are not reused in
1498adjacent cells to avoid mutual interference.
1499Thus, the network in Figure 16.7 may use a
1500frequency reuse pattern as shown in Figure
150116.8. In the figure, the network uses four
1502frequencies f1
1503, f2
1504, f3
1505, f4 and allocates them
1506in cells so that no two adjacent cells use
1507the same frequency. By creating appropriate
1508cell patterns, these frequencies can be
1509reused as often as necessary to cover the
1510entire service area.
15118 http://wirelessestimator.com/content/fatalities (accessed Feb. 2016).
15129 Ryan Knutson, “A New Spate of Deaths in the Wireless Industry,” Wall Street Journal, Aug. 21, 2013.
151310 Ianthe Dugan and Ryan Knuttson, “Cellphone Boom Spurs Antenna Safety Worries,” Wall Street Journal,
1514Oct. 2, 2014.
1515Figure 16.8: Cell phone frequency reuse
1516pattern example
1517502 • Supplementary Chapter 16 / Phone Networks
1518A simple example can help us understand the benefits of frequency reuse. Say you
1519operate a cell phone network and are allocated enough bandwidth to serve about 1,000
1520subscribers. If you have just one tower to serve your entire territory, only 1,000 of your
1521subscribers may call at any given time. Suppose, instead, that you use the frequency-reuse
1522pattern shown in Figure 16.8, and you divide your available bandwidth into four frequency
1523sub-bands. Each sub-band would be capable of serving 250 (1,000/4) subscribers. Thus,
1524each cell would be capable of supporting 250 simultaneous conversations.
1525But since the network now reuses frequencies, the same sub-band can support another
1526250 conversations in another cell. Thus, by dividing the service area into cells as necessary,
1527cell phone operators can support as many subscribers as they can sign up. For example, if
1528Figure 16.8 represents the layout of cells in your network, you will have 39 cells. With each
1529cell supporting 250 simultaneous conversations, you will be able to support 39 * 250 = 9,750
1530simultaneous cell phone calls. It is expensive and complicated to divide a service area into
1531cells, but by doing so there is no limit to the number of subscribers who may be supported
1532using a limited amount of bandwidth. If a particular cell
1533becomes too busy, you can simply subdivide it into more
1534cells. Figure 16.9 shows an example.
1535Though cells are most frequently shown as regular
1536hexagons, in practice, cell shapes and sizes are influenced by
1537population, terrain, buildings, hills, etc. Figure 16.10 shows
1538the actual locations of cell towers in Pomona, California.11
1539We see that the arrangement of cells is not very regular.
1540Areas with a high density of users have more towers, with
1541each tower covering a smaller cell. The figure shows that
1542the regular hexagonal pattern commonly used to represent
1543cell phone coverage areas of cell phone towers is only a
1544convenient representation of the actual pattern of cells.
154511 Retrieved from www.antennasearch.com.
1546Figure 16.9: Resizing cells
1547to accommodate subscribers
1548M&A activity driven by spectrum needs
1549In most industries, merger and acquisitions (M&A) activities are triggered by an
1550ambition to expand to a new market, eliminate a competitor, or acquire new customers.
1551While all these drivers are present in the cellular industry, too, one of the primary
1552drivers has been a need to acquire wireless spectrum from competitors. For example,
1553AT&T proposed to buy T-Mobile in 2012 for almost $40 billion in order to get access
1554to T-Mobile’s spectrum. Regulators blocked the deal for fear of its potential to reduce
1555competition and lead to higher consumer prices.
1556Since then, some spectrum has been freed from TV broadcasting and made available
1557to cellular providers in spectrum auctions.
1558Cell Phones • 503
1559Network investment expenses
1560Building out the voice and data network is incredibly expensive and one of the largest
1561capital-intensive projects in the country. In 2014, AT&T invested $21 billion in its
1562network.12 Verizon invested $80 billion on its network during 2009–2014.13
1563Roaming and the Role of the MTSO
1564The mobile telephone switching office (MTSO) is the nerve center of cell phone service, as
1565shown in Figure 16.7. The MTSO is the switching office that connects all of the individual cell towers
1566to the Central Office (CO). The MTSO is responsible for monitoring the relative signal strength of cell
156712 http://about.att.com/story/att_fourth_quarter_earnings_2014.html (accessed Feb. 2016).
156813 https://www.verizon.com/about/sites/default/files/2014_vz_annual_report.pdf (accessed Feb. 2016).
1569Figure 16.10: Cell phone towers in Pomona, CA
1570504 • Supplementary Chapter 16 / Phone Networks
1571phones as reported by each of the cell towers, and switching conversations to the cell towers with the best
1572possible reception.
1573Every cell phone has a home MTSO. When a cell phone is switched on, it periodically
1574broadcasts its presence. These broadcasts are received by all the towers in its coverage
1575area, and all the towers send the signal strength and other information about the phone to
1576the MTSO. The MTSO identifies the tower receiving the best signal from the phone and
1577instructs the tower to handle calls from the phone. The selected MTSO also informs the
1578home MTSO of the mobile phone user that the phone is located within its service area. For
1579example, the home MTSO for a cell phone number with area code 407 will be in Orlando,
1580Florida. If the cell phone is currently located in Kingston, Rhode Island, the local MTSO
1581in Kingston, Rhode Island, will inform the home MTSO of the cell phone in Orlando that
1582the cell phone is located in Kingston. When someone dials the 407 number, the call is first
1583connected to the user’s home MTSO in Orlando. The home MTSO will then direct the call
1584to the MTSO in Kingston, Rhode Island, where the user is located. This MTSO will forward
1585the call to the tower that is responsible for the phone.14
1586If the user moves away from a cell, its signals to its current cell tower weaken.
1587Simultaneously, the signals get stronger at a tower in a neighboring cell. The MTSO uses
1588these differences in signal strengths to transfer responsibility of handling the call to the
1589appropriate neighboring tower. This process is called handoff. Handoff is the process of
1590transferring a phone call in progress from one cell transmitter and frequency pair to another cell transmitter
1591and receiver, using a different frequency pair without interruption of the call.
1592Code Division Multiple Access (CDMA)
1593Third-generation cell phones use a kind of multiplexing method that we have not seen
1594before. It is called CDMA, which stands for code division multiple access. CDMA is a coding
1595scheme, used as a modulation technique, in which multiple channels are independently coded for transmission
1596over a single wideband channel. Several transmissions can occur simultaneously within the same bandwidth,
1597with the mutual interference reduced by the use of unique codes in each transmission.
1598We saw frequency-division multiplexing (FDM), where different signals are sent at
1599different frequencies. By tuning into one frequency, the receiver can obtain the signal at
1600the frequency, eliminating the signals being transmitted at all other frequencies. Another
1601common multiplexing technique is time division multiplexing (TDM). In this scheme,
1602each station is allowed to transmit in an allocated time slot. The receiver only listens to
1603transmissions at the specified time slots and ignores transmissions made at other times.
1604SONET is an example of TDM.
1605Whereas multiplexing schemes such as FDM and TDM are useful, they have a major
1606limitation. Each channel is only allowed to use a fraction of the transmitting capacity of
1607the medium. Thus, if an FDM scheme has 10 frequency slots, it can serve at most 10 users.
1608The 11th user will have to wait till one of the earlier users hangs up.
1609CDMA eliminates this limitation. CDMA allows an almost unlimited number of users to
1610transmit signals at any time using the entire bandwidth of the medium. Each communication
161114 For a good early read on optimizations being attempted in the industry, see Scott Woolley, “The $10 Phone
1612Bill,” Forbes, Nov. 2009.
1613Example Case—Cell Phones and Global Development • 505
1614is allocated a unique chipping code. Before transmission, signals are processed using
1615the assigned chipping code. Analogous to the FDM example, the receiver processes
1616the incoming signal using the same chipping code used by the sender. This extracts the
1617communication of interest and eliminates most of the information in all other signals. Any
1618number of chipping codes may be generated, and therefore any number of users may be
1619added to a cell. The only limitation is that, as used in the cell phone system, CDMA does
1620not eliminate all information from other signals. As a result, the background noise level in
1621CDMA increases when the number of users increases. Eventually this can make the sound
1622quality in the cell unacceptable. When this happens, the operator divides a large cell into
1623smaller cells and adds new towers to serve the new cells. This is why, as shown in Figure
162416.9, densely populated areas have many towers in close proximity. The hands-on exercise
1625uses a spreadsheet example to demonstrate how four users can use CDMA to send and
1626receive 5 bits each.
1627EXAMPLE CASE—Cell Phones and Global
1628Development
1629Mo Ibrahim founded MSI Cellular Investments (later Celtel) in Africa in 1998, when the
1630continent with a population of 950 million people had about 2 million phones. Recognizing
1631the opportunity for a business that connected people, Mo went about building his network
1632and company, investing more than $750 million in seven years. In 2005, when the company
1633was bought by MTC Kuwait for $3.4 billion, it had 24 million mobile subscribers in 15
1634different African countries. This made it one of Africa’s most successful businesses at the
1635time. Today, Africa has more than 350 million unique mobile subscribers. Mo has now
1636invested in a global satellite provider, O3B, which aims to bring cell phone service to remote
1637locations in developing countries. Mo believes Africa can offer well-run businesses growth
1638rates of more than 30% each year, a rate unattainable anywhere else in the world.
1639In spite of this growth, data usage continues to be expensive and unaffordable for most
1640users in Africa. Cell phones in Africa are primarily pre-paid, and pre-paid data costs about
1641$11/GB. Where daily incomes for large sections of the population are below $1.25/day,
1642$11/GB is a very heavy price to pay for Internet access.
1643This has created other business opportunities and the mobility-driven eco-system
1644continues to evolve. Entrepreneurs are outfitting mini-buses with free Wi-Fi access to the
1645Internet. These mini-buses are a popular mode of transportation in Africa, and buses that
1646offer this service have an advantage compared to their competitors on the same routes—
1647gathering more fare-paying passengers who use the time to browse the web and update their
1648social media status. Each bus pays about $25/month for data, but at every stop, passengers
1649choose Wi-Fi outfitted buses over the competition.
1650In another well-known underdeveloped part of the world, Afghanistan, there were only
165110,000 fixed-line phones in 2001. Services such as television were non-existent at the time.
1652Today, about 20 million of the 30 million people in the country use mobile phones, about
1653the same number of people who regularly watch television. Similarly, the road network has
1654grown from 32 paved miles to more than 7,500 paved miles. This connectivity will hopefully
1655bring political accountability and improve lives.
1656506 • Supplementary Chapter 16 / Phone Networks
1657In China, where development is more rapid and advanced, mobile telephony is bringing
1658other changes. In 2005, most consumer transactions in the country were conducted by
1659cash. In 2015, on the other hand, more than 350 million people used their cell phones for
1660payment, a growth of more than 60% compared to 2014. The volume of payments made
1661through mobile phones was estimated at $2.5 trillion in 2015. Alibaba and WeChat are the
1662leading players in the industry in China. Alibaba charges a maximum service fee of 0.6%
1663for each transaction. By comparison, credit card swipe fees in the US can exceed 1.5%
1664and are the second- or third-highest cost for retailers, after wages and healthcare benefits.
1665Alibaba and WeChat have established protections similar to those offered by US credit
1666card companies to establish reputations for trust among both buyers and sellers. Alibaba
1667even offers investment instruments such as money market funds that offer higher interest
1668rates than state-owned banks, so customers use these applications also as their wealthmanagement
1669solutions.
1670References
16711. Mohseni, Saad. “The Untold Story of Afghan Progress.” Wall Street Journal, Mar. 17,
16722013.
16732. Salvaterra, Neanda. “Mobile Pioneer Sees Rich Promise in Africa.” Wall Street Journal,
1674Apr. 18, 2011.
16753. Vogt, Heidi. “No Wi-Fi at home? Then Take a Bus.” Wall Street Journal, Apr. 15, 2014.
16764. Yuan, Li. “How Mobile Payments Reshape Lifestyles.” Wall Street Journal, Feb. 24, 2016.
1677Summary
1678This chapter provided a high-level overview of landline and cell phone networks. Though
1679many functions of the phone system are moving over to computerized communication
1680technologies, such as e-mail, IM, and VoIP, the phone continues to be an important medium
1681for business and personal communication. The phone system transmits signals in the
1682frequency range 300 Hz–3,400 Hz. Each end user is connected to the nearest end office
1683of the phone company using a dedicated pair of copper wires called the local loop. The
1684phone system is used in many parts of the world to offer high-speed Internet service using
1685a technology called Digital Subscriber Line (DSL).
1686The phone system in the United States was initially operated by one company, AT&T.
1687In 1984, the company agreed in a court settlement to focus exclusively on providing longdistance
1688phone service. AT&T’s local phone service networks were divested as seven local
1689phone companies. Later, in 1996, Congress passed the Telecommunications Act, which
1690opened up all sectors of telephony to competition. The Telecommunications Act also
1691required incumbent local phone companies with established phone networks to allow
1692competitors to use the incumbent phone company’s networks at reasonable prices to
1693compete with the incumbent.
1694Of late, cellular telephony is becoming increasingly popular, even in developing countries.
1695In many cases, developing countries actually find it cheaper to set up a cell phone network
1696than to set up a landline phone network. Cell phone networks divide the coverage area into 
1697Review Questions • 507
1698small cells. Users in each cell are served by a cell phone tower located in the cell. As users
1699move from cell to cell, their calls are handed off to the most appropriate cell phone tower in
1700the area. The division of the coverage area into cells permits frequency reuse, which allows
1701a small set of frequencies to be used to serve as many subscribers as necessary.
1702Cell phone technologies have evolved in three distinct phases. Each phase is called a
1703generation. We are currently using the third generation of cell phone technologies, creatively
1704called 3G. The third generation uses a multiplexing technology called CDMA, which is very
1705efficient in using bandwidth. CDMA enables cell phone networks to offer high-data-rate
1706network connections to a large number of end users using the very limited bandwidth
1707available for 3G networks.
1708In which of the following contexts were the words, “Watson, come in here please” used
1709in homage to Alexander Graham Bell:15
1710A. March 21, 2006, Biz Stone’s response, on seeing the first tweet by Jack Dorsey
1711B. April 8, 2008, Noah Glass to Evan Williams, on hearing about the Twitter IPO
1712C. November 6, 2008, Evan Williams to Noah Glass, on being offered a position at
1713 Twitter
1714D. January 23, 2009, Noah Glass to Biz Stone, on being fired from Twitter
1715About the Colophon
1716First impressions are lasting impressions. It is therefore not surprising that the first phrase
1717spoken on the telephone by the inventor of the technology is also one of the most
1718memorable phrases ever communicated using the technology.
1719Alexander Graham Bell maintained meticulous diaries of his experiments to create the
1720telephone. In his diary entry of March 10, 187616 (Figure 16.11), Alexander Graham Bell
1721described his first successful communication using the telephone. When he spoke through
1722the instrument, his assistant, Thomas A. Watson, was in the next room. In the diary, Graham
1723Bell wrote, “I then shouted into M the following sentence: ‘Mr. Watson—come here—I
1724want to see you.’ To my delight he came and declared that he had heard and understood
1725what I said.”
1726REVIEW QUESTIONS
17271. What are some of the ways in which the phone network has been important for data
1728communications over the years?
17292. What are some of the important landmarks in the development of phone service?
17303. What is circuit switching?
17314. How has the adoption of landline telephony evolved in the last decade?
173215 Answer: A. Source: Nick Bilton, Hatching Twitter (Penguin, 2013). All people mentioned are co-founders of
1733Twitter.
173416 www.loc.gov/exhibits/treasures/trr002.html.
1735Figure 16.11: Alexander Graham Bell’s diary, March 10, 1876
1736(Source: Alexander Graham Bell family papers, 1834-1974, Library of Congress, Manuscript Division)
1737Example Case Questions • 509
17385. What is the local loop in the context of phone service?
17396. What is the central office or the end office?
17407. What are inter-exchange carriers (IXC)? What are some similarities and differences between
1741the local loop and IXC links?
17428. What is hertz?
17439. Why is the phone system designed to carry signals in the frequency range 33 Hz–3,400
1744Hz?
174510. What is digital subscriber line (DSL)?
174611. What was the motivation for the development of DSL technology?
174712. What are the three kinds of signals on a cable providing DSL and phone service? What
1748are the frequency ranges used by the three signals?
174913. Why do most ISPs provide much higher downstream data rates than upstream data
1750rates?
175114. Why is the modified final judgment important to the development of phone service in
1752the United States?
175315. What was the outcome of the modified final judgment?
175416. What were the circumstances that led to the Telecommunications Act of 1996?
175517. What were some of the implications of the Telecommunications Act of 1996?
175618. What is cellular telephony?
175719. What are the three generations of cellular telephony service? Describe the important
1758features of each generation of cell phone service.
175920. What is frequency reuse in the context of cellular telephony? Why is frequency reuse
1760necessary for cellular telephony?
176121. Why are service areas divided into small cells for cellular telephony?
176222. What is the MTSO in cell phone service? What are the important roles of the MTSO?
176323. What is handoff? Why is handoff important? How does it work?
176424. What is CDMA?
176525. How is CDMA better suited than TDM or FDM for cellular telephony?
1766EXAMPLE CASE QUESTIONS
17671. What are some of the most promising investment opportunities in Africa and the
1768developing world today?
17692. If you were appointed the regional manager for a cell phone company in Africa, with
1770responsibility to increase sales, what are some specific steps you would take? Briefly
1771justify your recommendations.
17723. Have you used mobile payments? Briefly describe your experience with mobile
1773payments.
1774510 • Supplementary Chapter 16 / Phone Networks
1775HANDS-ON EXERCISE—CDMA
1776Table 16.1: Dot product of two codes
1777Element Code 1 Code 2 Product of elements
17781 1 1 1
17792 1 −1 −1
1780Dot product = sum of product of elements = 0
1781Creating a hands-on exercise for telephony is not simple because end users have no access to
1782phone carrier networks. Therefore, instead of trying to poke into phone company networks,
1783the hands-on exercise for this chapter will give you the opportunity to learn about one of
1784the most important recent developments in telephony—CDMA. You will create CDMA
1785codes and use these CDMA codes to multiplex data transmission. You will also decode the
1786data for reception at the receiver.17
1787In the discussion below, we will find it convenient to use -1 to represent binary 0. This
1788makes it easier to show the computations involved.
1789In CDMA, transmitters use mutually orthogonal codes, called chipping codes, to process
1790data. Orthogonality of chipping codes means that the dot product of any two chipping
1791codes is 0. The dot product, or inner product, of two codes is calculated by multiplying
1792the respective elements of the two codes, and taking the sum of the products. Table 16.1
1793shows an example. The two chipping codes in the example are [1 1] and [1 -1]. Writing the
1794codes in columns as in Table 16.1, we can evaluate the dot product of the two codes and
1795confirm that it is zero.
1796Techniques exist to create chipping codes of any length.18 The chip rate is higher than
1797the bit rate. Once a chipping code is selected, processing data for transmission involves
1798calculating the exclusive OR of each bit of the data with every element of the chipping
1799code. The result is transmitted into the medium.
1800The receivers multiply the resulting signal in the medium with the transmitter’s chipping
1801code and add the result for each bit period. This recovers the transmitted data. This is the
1802basic CDMA operation.
1803The spreadsheet CDMA.xlsx in the readings for this chapter at the companion website
1804has an example of four pairs of users transmitting 5 bits each. Figure 16.12 shows the data
1805and chipped signal for user D in the spreadsheet example.
1806The README worksheet walks you through the spreadsheet to see the data, chipping
1807codes, sender processing, and receiver processing. You will find it convenient to also
1808download the CDMA.xlw workspace and open the workspace. This will show you both the
1809README and data worksheets simultaneously.
181017 There is some level of math involved in this exercise, but every attempt has been made to keep it simple.
181118 For one such technique, please search Wikipedia or another resource for Walsh codes.
1812Hands-on Exercise—CDMA • 511
1813You may find it useful to use Excel’s Trace precedents feature to visualize how the
1814values are computed.19
1815Use the CDMA.xlsx worksheet to answer the following:
18161. Use Wikipedia or another resource to write a brief summary of CDMA.
18172. Use Wikipedia or another information resource to write a Walsh matrix of size 4 * 4.
18183. Pick any two different codes in the 8 * 8 Walsh matrix used in the data.xlsx worksheet
1819(cells A[28]–H[36]). Using a procedure similar to Table 16.1, confirm that the dot
1820product of the two codes is zero.
18214. Repeat Question 3, but use the chipping code for B for both codes. Confirm that the
1822dot product of a code with itself is not zero.
18235. The example in the spreadsheet uses the chipping codes from the first four columns
1824in the Walsh codes of size 8. Replace the chipping code for D with one of the unused
1825chipping codes (any code in columns E[29–36] … H[29–36]). Confirm that the data is
1826recovered correctly with the new code (contents in cells A22–D26 should not change).
18276. You saw in Question 4 above that the dot product of a code with itself is not zero. To
1828see the impact of this, repeat Question 5 above, but this time, reuse the chipping code
1829for user B for user D. Confirm that the data is not recovered correctly.
183019 William Rogers, a student in the University of South Florida’s fall 2013 class, suggested this.
1831Figure 16.12: Data and chipped-signal example
1832Chipped
1833data
1834Data
1835512 • Supplementary Chapter 16 / Phone Networks
1836CRITICAL-THINKING EXERCISE—Other Three
1837Billionaires
1838The company we saw in the example case, O3B, selected its name to represent the “other 3
1839billion” people who did not have Internet access. However, given the range of its customers,
1840the company is sometimes mocked as representing the “other 3 billionaires” for bringing
1841high-speed network connectivity to billionaires’ yachts, or the “other 3 battle-groups” for
1842bringing high-speed network connectivity to military ships.
18431. Should O3B protect the purity of its business by discontinuing service to profitable
1844but wealthy customers? Justify your response.
1845Critical-Thinking References
18461. http://spacenews.com/2014-top-fixed-satellite-service-operators-once-mocked-03binvestment-now-force-multiplier-for-ses/.
1847IT INFRASTRUCTURE DESIGN EXERCISE—Switch
1848to VoIP
1849While the Mumbai location uses VoIP for all its voice traffic, the Singapore location is trying
1850to decide whether to switch its users from traditional phones to VoIP phones. Answer the
1851following questions:
18521. Using the Internet or another information resource, compare the advantages and
1853disadvantages of VoIP compared to traditional phone service (PSTN) along dimensions
1854such as cost, performance, and reliability.
18552. Based on the above, would you recommend that TrendyWidgets switch from PSTN to
1856VoIP?
1857_____________________________________________________________________
1858Chapters 15 and 16 are online supplements to Business Data Communications and IT
1859Infrastructures, 2nd Edition, by Manish Agrawal & Rekha Sharma. © 2017, Prospect Press