After studying this section you should be able to do the following:
- Understand the last-mile problem and be able to discuss the pros and cons of various broadband technologies, including DSL, cable, fiber, and various wireless offerings.
- Describe 3G and 4G systems, listing major technologies and their backers.
- Understand the issue of Net neutrality and put forth arguments supporting or criticizing the concept.
The Internet backbone is made of fiber-optic lines that carry data traffic over long distances. Those lines are pretty speedy. In fact, several backbone providers, including AT&T and Verizon, are rolling out infrastructure with 100 Gbps transmission speeds (that’s enough to transmit a two-hour high-definition [HD] movie in about eight seconds)1 (Spangler, 2010). But when considering overall network speed, remember Amdahl’s Law: a system’s speed is determined by its slowest component (Gilder, 2000). More often than not, the bottleneck isn’t the backbone but the so-called last mile, or the connections that customers use to get online.
High-speed last-mile technologies are often referred to as broadband Internet access (or just broadband). What qualifies as broadband varies. In 2009, the Federal Communications Commission (FCC) redefined broadband as having a minimum speed of 768 Kbps (roughly fourteen times the speed of those old 56 Kbps modems). Other agencies worldwide may have different definitions. But one thing is clear: a new generation of bandwidth-demanding services requires more capacity. As we increasingly consume Internet services like HD streaming, real-time gaming, video conferencing, and music downloads, we are in fact becoming a bunch of voracious, bit-craving gluttons.
With the pivotal role the United States has played in the creation of the Internet, and in pioneering software, hardware, and telecommunications industries, you might expect the United States to lead the world in last-mile broadband access. Not even close. A recent study ranked the United States twenty-sixth in download speeds, (Lawson, 2010) while others have ranked the United States far behind in speed, availability, and price (Hansell, 2009).
Sounds grim, but help is on the way. A range of technologies and firms are upgrading infrastructure and developing new systems that will increase capacity not just in the United States but also worldwide. Here’s an overview of some of the major technologies that can be used to speed the Internet’s last mile.
When folks talk about bandwidth, they’re referring to data transmission speeds. Bandwidth is often expressed in bits per second, or bps. Prefix letters associated with multiples of bps are the same as the prefixes we mentioned in Chapter 5 “Moore’s Law: Fast, Cheap Computing and What It Means for the Manager” when discussing storage capacity in bytes: Kbps = thousand bits (or kilobits) per second, Mbps = million bits (or megabits) per second, Gbps = billion bits (or gigabits) per second (or terabit), and Tbps = trillion bits (or terabits) per second.
Remember, there are eight bits in a byte, and one byte is a single character. One megabyte is roughly equivalent to one digital book, forty-five seconds of music, or twenty seconds of medium-quality video (Farzad, 2010). But you can’t just divide the amount of bytes by eight to estimate how many bits you’ll need to transfer. When a file or other transmission is sliced into packets (usually of no more than about 1,500 bytes), there’s some overhead added. Those packets “wrap” data chunks in an envelope surrounded by source and destination addressing and other important information.
Here are some rough demand requirements for streaming media. For streaming audio like Pandora, you’d need at least 150 Kbps for acceptable regular quality, and at least 300 Kbps for high quality2. For streaming video (via Netflix), at a minimum you’d need 1.5 Mbps, but 3.0 Mbps will ensure decent video and audio. For what Netflix calls HD streaming, you’ll need a minimum of 5 Mbps, but would likely want 8 Mbps or more to ensure the highest quality video and audio3.
Roughly 90 percent of U.S. homes are serviced by a cable provider, each capable of using a thick copper wire to offer broadband access. That wire (called a coaxial cable or coax) has shielding that reduces electrical interference, allowing cable signals to travel longer distances without degrading and with less chance of interference than conventional telephone equipment.
One potential weakness of cable technology lies in the fact that most residential providers use a system that requires customers to share bandwidth with neighbors. If the guy next door is a BitTorrent-using bandwidth hog, your traffic could suffer (Thompson, 2010).
Cable is fast and it’s getting faster. Many cable firms are rolling out a new technology called DOCSIS 3.0 that offers speeds up to and exceeding 50 Mbps (previous high-end speeds were about 16 Mbps and often much less than that). Cable firms are also creating so-called fiber-copper hybrids that run higher-speed fiber-optic lines into neighborhoods, then use lower-cost, but still relatively high-speed, copper infrastructure over short distances to homes (Hansell, 2009). Those are fast networks, but they are also very expensive to build, since cable firms are laying entirely new lines into neighborhoods instead of leveraging the infrastructure that they’ve already got in place.
DSL: Phone Company Copper
Digital subscriber line (DSL) technology uses the copper wire the phone company has already run into most homes. Even as customers worldwide are dropping their landline phone numbers, the wires used to provide this infrastructure can still be used for broadband.
DSL speeds vary depending on the technology deployed. Worldwide speeds may range from 7 Mbps to as much as 100 Mbps (albeit over very short distances) (Hansell, 2009). The Achilles heel of the technology lies in the fact that DSL uses standard copper telephone wiring. These lines lack the shielding used by cable, so signals begin to degrade the further you are from the connecting equipment in telephone company offices. Speeds drop off significantly at less than two miles from a central office or DSL hub. If you go four miles out, the technology becomes unusable. Some DSL providers are also using a hybrid fiber-copper system, but as with cable’s copper hybrids, this is expensive to build.
The superspeedy DSL implementations that are popular in Europe and Asia work because foreign cities are densely populated and so many high-value customers can be accessed over short distances. In South Korea, for example, half the population lives in apartments, and most of those customers live in and around Seoul. This density also impacts costs—since so many people live in apartments, foreign carriers run fewer lines to reach customers, digging up less ground or stringing wires across fewer telephone poles. Their U.S. counterparts by contrast need to reach a customer base sprawled across the suburbs, so U.S. firms have much higher infrastructure costs (Hansell, 2009).
There’s another company with copper, electricity-carrying cables coming into your home—the electrical utility. BPL, or broadband over power line, technology has been available for years. However, there are few deployments because it is considered to be pricier and less practical than alternatives (King, 2009).
Fiber: A Light-Filled Glass Pipe to Your Doorstep
Fiber to the home (FTTH) is the fastest last-mile technology around. It also works over long distances. Verizon’s FiOS technology boasts 50 Mbps download speeds but has tested network upgrades that increase speeds by over six times that (Higginbotham, 2009). The problem with fiber is that unlike cable or DSL copper, fiber to the home networks weren’t already in place. That means firms had to build their own fiber networks from scratch.
The cost of this build out can be enormous. Verizon, for example, has spent over $23 billion on its FTTH infrastructure. However, most experts think the upgrade was critical. Verizon has copper into millions of homes, but U.S. DSL is uncompetitive. Verizon’s residential landline business was dying as users switch to mobile phone numbers, and while mobile is growing, Verizon Wireless is a joint venture with the United Kingdom’s Vodaphone, not a wholly owned firm. This means it shares wireless unit profits with its partner. With FiOS, Verizon now offers pay television, competing with cable’s core product. It also offers some of the fastest home broadband services anywhere, and it gets to keep everything it earns.
In 2010, Google also announced plans to bring fiber to the home. Google deems its effort an experiment—it’s more interested in learning how developers and users take advantage of ultrahigh-speed fiber to the home (e.g., what kinds of apps are created and used, how usage and time spent online change), rather than becoming a nationwide ISP itself. Google says it will investigate ways to build and operate networks less expensively and plans to share findings with others. The Google network will be “open,” allowing other service providers to use Google’s infrastructure to resell services to consumers. The firm has pledged to bring speeds of 1 Gbps at competitive prices to at least 50,000 and potentially as many as 500,000 homes. Over 1,100 U.S. communities applied to be part of the Google experimental fiber network (Ingersoll & Kelly, 2010; Rao, 2010).
Mobile wireless service from cell phone access providers is delivered via cell towers. While these providers don’t need to build a residential wired infrastructure, they still need to secure space for cell towers, build the towers, connect the towers to a backbone network, and license the wireless spectrum (or airwave frequency space) for transmission.
We need more bandwidth for mobile devices, too. AT&T now finds that the top 3 percent of its mobile network users gulp up 40 percent of the network’s capacity (thanks, iPhone users), and network strain will only increase as more people adopt smartphones. These users are streaming Major League Baseball games, exploring the planet with Google Earth, watching YouTube and Netflix, streaming music through Pandora, and more. Get a bunch of iPhone users in a crowded space, like in a college football stadium on game day, and the result is a network-choking data traffic jam. AT&T estimates that it’s not uncommon for 80 percent of game-day iPhone users to take out their phones and surf the Web for stats, snap and upload photos, and more. But cell towers often can’t handle the load (Farzad, 2010). If you’ve ever lost coverage in a crowd, you’ve witnessed mobile network congestion firsthand. Trying to have enough capacity to avoid congestion traffic jams will cost some serious coin. In the midst of customer complaints, AT&T committed to spending $18 billion on network upgrades to address its wireless capacity problem (Edwards & Kharif, 2010).
We’re in the midst of transitioning from third generation (3G) to fourth generation (4G) wireless networks. 3G systems offer access speeds usually less than 2 Mbps (often a lot less) (German, 2010). While variants of 3G wireless might employ an alphabet soup of technologies—EV-DO (evolution data optimized), UMTS (universal mobile telecommunications systems), and HSDPA (high-speed downlink packet link access) among them—3G standards can be narrowed down to two camps: those based on the dominant worldwide standard called GSM (global system for mobile communications) and the runner-up standards based on CDMA (code division multiplex access). Most of Europe and a good chunk of the rest of the world use GSM. In the United States, AT&T and T-Mobile use GSM-based 3G. Verizon Wireless and Sprint use the CDMA 3G standard. Typically, handsets designed for one network can’t be used on networks supporting the other standard. CDMA has an additional limitation in not being able to use voice and data at the same time.
But 3G is being replaced by high-bandwidth 4G (fourth-generation) mobile networks. 4G technologies also fall into two standards camps: LTE (Long Term Evolution) and WiMAX (Worldwide Interoperability for Microwave Access).
LTE looks like the global winner. In the United States, every major wireless firm, except for Sprint, is betting on LTE victory. Bandwidth for the service rivals what we’d consider fast cable a few years back. Average speeds range from 5 to 12 Mbps for downloads and 2 to 5 Mbps for upload, although Verizon tests in Boston and Seattle showed download speeds as high as 50 Mbps and upload speeds reaching 25 Mbps (German, 2010).
Competing with LTE is WiMAX; don’t confuse it with Wi-Fi. As with other 3G and 4G technologies, WiMAX needs cell towers and operators need to have licensed spectrum from their respective governments (often paying multibillion-dollar fees to do so). Average download and upload speeds should start out at 3–6 Mbps and 1 Mbps, respectively, although this may go much higher (Lee, 2010).
WiMAX looks like a particularly attractive option for cable firms, offering them an opportunity to get into the mobile phone business and offer a “quadruple play” of services: pay television, broadband Internet, home phone, and mobile. Comcast and Time Warner have both partnered with Clearwire (a firm majority-owned by Sprint), to gain access to WiMAX-based 4G mobile.
4G could also rewrite the landscape for home broadband competition. If speeds increase, it may be possible for PCs, laptops, and set-top boxes (STB) to connect to the Internet wirelessly via 4G, cutting into DSL, cable, and fiber markets.
Wireless systems provided by earth-bound base stations like cell phone towers are referred to as terrestrial wireless, but it is possible to provide telecommunications services via satellite. Early services struggled due to a number of problems. For example, the first residential satellite services were only used for downloads, which still needed a modem or some other connection to send any messages from the computer to the Internet. Many early systems also required large antennas and were quite expensive. Finally, some services were based on satellites in geosynchronous earth orbit (GEO). GEO satellites circle the earth in a fixed, or stationary, orbit above a given spot on the globe, but to do so they must be positioned at a distance that is roughly equivalent to the planet’s circumference. That means signals travel the equivalent of an around-the-world trip to reach the satellite and then the same distance to get to the user. The “last mile” became the last 44,000 miles at best. And if you used a service that also provided satellite upload as well as download, double that to about 88,000 miles. All that distance means higher latency (more delay) (Ou, 2008).
A firm named O3b Networks thinks it might have solved the challenges that plagued early pioneers. O3b has an impressive list of big-name backers that include HSBC bank, cable magnate John Malone, European aerospace firm SES, and Google.
The name O3b stands for the “Other 3 Billion,” of the world’s population who lack broadband Internet access, and the firm hopes to provide “fiber-quality” wireless service to more than 150 countries, specifically targeting underserved portions of the developing world. These “middle earth orbit” satellites will circle closer to the earth to reduce latency (only about 5,000 miles up, less than one-fourth the distance of GEO systems). To maintain the lower orbit, O3b’s satellites orbit faster than the planet spins, but with plans to launch as many as twenty satellites, the system will constantly blanket regions served. If one satellite circles to the other side of the globe, another one will circle around to take its place, ensuring there’s always an O3b “bird” overhead.
Only about 3 percent of the sub-Saharan African population uses the Internet, compared to about 70 percent in the United States. But data rates in the few places served can cost as much as one hundred times the rates of comparable systems in the industrialized world (Lamb, 2008). O3b hopes to change that equation and significantly lower access rates. O3b customers will be local telecommunication firms, not end users. The plan is for local firms to buy O3b’s services wholesale and then resell it to customers alongside rivals who can do the same thing, collectively providing more consumer access, higher quality, and lower prices through competition. O3b is a big, bold, and admittedly risky plan, but if it works, its impact could be tremendous.
Wi-Fi and Other Hotspots
Many users access the Internet via Wi-Fi (which stands for wireless fidelity). Computer and mobile devices have Wi-Fi antennas built into their chipsets, but to connect to the Internet, a device needs to be within range of a base station or hotspot. The base station range is usually around three hundred feet (you might get a longer range outdoors and with special equipment; and less range indoors when signals need to pass through solid objects like walls, ceilings, and floors). Wi-Fi base stations used in the home are usually bought by end users, then connected to a cable, DSL, or fiber provider.
And now a sort of mobile phone hotspot is being used to overcome limitations in those services, as well. Mobile providers can also be susceptible to poor coverage indoors. That’s because the spectrum used by most mobile phone firms doesn’t travel well through solid objects. Cell coverage is also often limited in the United States because of a lack of towers, which is a result of the NIMBY problem (not in my backyard). People don’t want an eighty-foot to four-hundred-foot unsightly tower clouding their local landscape, even if it will give their neighborhood better cell phone coverage (Dechter & Kharif, 2010). To overcome reception and availability problems, mobile telecom services firms have begun offering fentocells. These devices are usually smaller than a box of cereal and can sell for $150 or less (some are free with specific service contracts). Plug a fentocell into a high-speed Internet connection like an in-home cable or fiber service and you can get “five-bar” coverage in a roughly 5,000-square-foot footprint (Mims, 2010). That can be a great solution for someone who has an in-home, high-speed Internet connection, but wants to get phone and mobile data service indoors, too.
Net Neutrality: What’s Fair?
Across the world, battle lines are being drawn regarding the topic of Net neutrality. Net neutrality is the principle that all Internet traffic should be treated equally (Honan, 2008). Sometimes access providers have wanted to offer varying (some say “discriminatory”) coverage, depending on the service used and bandwidth consumed. But where regulation stands is currently in flux. In a pivotal U.S. case, the FCC ordered Comcast to stop throttling (blocking or slowing down) subscriber access to the peer-to-peer file sharing service BitTorrent. BitTorrent users can consume a huge amount of bandwidth—the service is often used to transfer large files, both legitimate (like version of the Linux operating system) and pirated (HD movies). Then in spring 2010, a federal appeals court moved against the FCC’s position, unanimously ruling that the agency did not have the legal authority to dictate terms to Comcast4.
On one side of the debate are Internet service firms, with Google being one of the strongest Net neutrality supporters. In an advocacy paper, Google states, “Just as telephone companies are not permitted to tell consumers who they can call or what they can say, broadband carriers should not be allowed to use their market power to control activity online5.” Many Internet firms also worry that if network providers move away from flat-rate pricing toward usage-based (or metered) schemes, this may limit innovation. Says Google’s Vint Cerf (who is considered one of the “fathers of the Internet” for his work on the original Internet protocol suite) “You are less likely to try things out. No one wants a surprise bill at the end of the month” (Jesdanun, 2009). Metered billing may limit the use of everything from iTunes to Netflix; after all, if you have to pay for per-bit bandwidth consumption as well as for the download service, then it’s as if you’re paying twice.
The counterargument is that if firms are restricted from charging more for their investment in infrastructure and services, then they’ll have little incentive to continue to make the kinds of multibillion-dollar investments that innovations like 4G and fiber networks require. Telecom industry executives have railed against Google, Microsoft, Yahoo! and others, calling them free riders who earn huge profits by piggybacking off ISP networks, all while funneling no profits back to the firms that provide the infrastructure. One Verizon vice president said, “The network builders are spending a fortune constructing and maintaining the networks that Google intends to ride on with nothing but cheap servers.…It is enjoying a free lunch that should, by any rational account, be the lunch of the facilities providers” (Mohammed, 2006). AT&T’s previous CEO has suggested that Google, Yahoo! and other services firms should pay for “preferred access” to the firm’s customers. The CEO of Spain’s Telefonica has also said the firm is considering charging Google and other Internet service firms for network use (Lunden, 2010).
ISPs also lament the relentlessly increasingly bandwidth demands placed on their networks. Back in 2007, YouTube streamed as much data in three months as the world’s radio, cable, and broadcast television channels combined stream in one year (Swanson, 2007), and YouTube has only continued to grow since then. Should ISPs be required to support the strain of this kind of bandwidth hog? And what if this one application clogs network use for other traffic, such as e-mail or Web surfing? Similarly, shouldn’t firms have the right to prioritize some services to better serve customers? Some network providers argue that services like video chat and streaming audio should get priority over, say, e-mail which can afford slight delay without major impact. In that case, there’s a pretty good argument that providers should be able to discriminate against services. But improving efficiency and throttling usage are two different things.
Internet service firms say they create demand for broadband business, broadband firms say Google and allies are ungrateful parasites that aren’t sharing the wealth. The battle lines on the Net neutrality frontier continue to be drawn, and the eventual outcome will impact consumers, investors, and will likely influence the continued expansion and innovation of the Internet.
Hopefully, this chapter helped reveal the mysteries of the Internet. It’s interesting to know how “the cloud” works but it can also be vital. As we’ve seen, the executive office in financial services firms considers mastery of the Internet infrastructure to be critically important to their competitive advantage. Media firms find the Internet both threatening and empowering. The advancement of last-mile technologies and issues of Net neutrality will expose threats and create opportunity. And a manager who knows how the Internet works will be in a better position to make decisions about how to keep the firm and its customers safe and secure, and be better prepared to brainstorm ideas for winning in a world where access is faster and cheaper, and firms, rivals, partners, and customers are more connected.
- The slowest part of the Internet is typically the last mile, not the backbone. While several technologies can offer broadband service over the last mile, the United States continues to rank below many other nations in terms of access speed, availability, and price.
- Cable firms and phone companies can leverage existing wiring for cable broadband and DSL service, respectively. Cable services are often criticized for shared bandwidth. DSL’s primary limitation is that it only works within a short distance of telephone office equipment.
- Fiber to the home can be very fast but very expensive to build.
- An explosion of high-bandwidth mobile applications is straining 3G networks. 4G systems may alleviate congestion by increasing capacities to near-cable speeds. Fentocells are another technology that can improve service by providing a personal mobile phone hotspot that can plug into in-home broadband access.
- The two major 3G standards (popularly referred to as GSM and CDMA) will be replaced by two unrelated 4G standards (LTE and WiMAX). GSM has been the dominant 3G technology worldwide. LTE looks like it will be the leading 4G technology.
- Satellite systems show promise in providing high-speed access to underserved parts of the world, but few satellite broadband providers have been successful so far.
- Net neutrality is the principle that all Internet traffic should be treated equally. Google and other firms say it is vital to maintain the openness of the Internet. Telecommunications firms say they should be able to limit access to services that overtax their networks, and some have suggested charging Google and other Internet firms for providing access to their customers.
Questions and Exercises
- Research online for the latest country rankings for broadband service. Where does the United States currently rank? Why?
- Which broadband providers can service your home? Which would you choose? Why?
- Research the status of Google’s experimental fiber network. Report updated findings to your class. Why do you suppose Google would run this “experiment”? What other Internet access experiments has the firm been involved in?
- Show your understanding of the economics and competitive forces of the telecom industry. Discuss why Verizon chose to go with fiber. Do you think this was a wise decision or not? Why? Feel free to do additional research to back up your argument.
- Why have other nations enjoyed faster broadband speeds, greater availability, and lower prices?
- The iPhone has been called both a blessing and a curse for AT&T. Why do you suppose this is so?
- Investigate the status of mobile wireless offerings (3G and 4G). Which firm would you choose? Why? Which factors are most important in your decision?
- Name the two dominant 3G standards. What are the differences between the two? Which firms in your nation support each standard?
- Name the two dominant 4G standards. Which firms in your nation will support the respective standards?
- Have you ever lost communication access—wirelessly or via wired connection? What caused the loss or outage?
- What factors shape the profitability of the mobile wireless provider industry? How do these economics compare with the cable and wire line industry? Who are the major players and which would you invest in? Why?
- Last-mile providers often advertise very fast speeds, but users rarely see speeds as high as advertised rates. Search online to find a network speed test and try it from your home, office, mobile device, or dorm. How fast is the network? If you’re able to test from home, what bandwidth rates does your ISP advertise? Does this differ from what you experienced? What could account for this discrepancy?
- How can 4G technology help cable firms? Why might it hurt them?
- What’s the difference between LEO satellite systems and the type of system used by O3b? What are the pros and cons of these efforts? Conduct some additional research. What is the status of O3b and other satellite broadband efforts?
- What advantages could broadband offer to underserved areas of the world? Is Internet access important for economic development? Why or why not?
- Does your carrier offer a fentocell? Would you use one? Why or why not?
- Be prepared to debate the issue of Net neutrality in class. Prepare positions both supporting and opposing Net neutrality. Which do you support and why?
- Investigate the status of Net neutrality laws in your nation and report your findings to your instructor. Do you agree with the stance currently taken by your government? Why or why not?
1Zacks.com, “AT&T Tests 100 Gb Ethernet in Move toward Faster Internet,” SeekingAlpha, March 10, 2010.
2Pandora, “Frequently Asked Questions,” http://blog.pandora.com/faq.
3LG Knowledge Base, “Bandwidth Needed for Instant Streaming,” http://lgknowledgebase.com/kb/index.php?View=entry&EntryID=6241.
4“What Is Net Neutrality?” The Week, April 7, 2010.
5Google, “A Guide to Net Neutrality for Google Users,” 2008, http://www.docstoc.com/docs/1064274/A-Guide-to-Net-Neutrality-for-Google-Users.
Dechter G. and O. Kharif, “How Craig McCaw Built a 4G Network on the Cheap,” BusinessWeek, May 24, 2010.
Edwards C. and O. Kharif, “Sprint’s Bold Play on a 4G Network,” BusinessWeek, March 30, 2010.
Farzad, R., “AT&T’s iPhone Mess,” BusinessWeek, February 3, 2010.
Farzad, R., “The Truth about Bandwidth,” BusinessWeek, February 3, 2010.
German, K., “On Call: Welcome to 4G,” CNET, March 9, 2010.
Gilder, G., Telecosm: How Infinite Bandwidth Will Revolutionize Our World (New York: Free Press, 2000).
Hansell, S., “The Broadband Gap: Why Is Theirs Faster?” New York Times, March 10, 2009.
Higginbotham, S., “Verizon Tests 10 Gbps to the Home. Yeah, You’ll Have to Share,” GigaOM, December 17, 2009.
Honan, M., “Inside Net Neutrality,” MacWorld, February 12, 2008.
Ingersoll M. and J. Kelly, “Think Big with a Gig: Our Experimental Fiber Network,” The Google Blog, February 2, 2010.
Jesdanun, M., “As the Internet Turns 40, Barriers Threaten Growth,” Technology Review, August 31, 2009.
King, R., “Telecom Companies Scramble for Funding,” BusinessWeek, August 3, 2009.
Lamb, G., “O3b Networks: A Far-Out Plan to Deliver the Web,” Christian Science Monitor, September 24, 2008.
Lawson, S., “US Ranks 26th in New Broadband Index,” Computerworld, May 25, 2010.
Lee, N., “Sprint’s 4G Plans Explained,” CNET, May 19, 2010.
Lunden, I., “Broadband Content Bits: Web Drama Investment, PPL Video Store, Telefonica to Charge?” paidContent:UK, February 11, 2010.
Mims, C., “A Personal Cell Phone Tower,” Technology Review, April 7, 2010.
Mohammed, A., “Verizon Executive Calls for End to Google’s ‘Free Lunch,’” Washington Post, February 7, 2006.
Ou, G., “Why Satellite Service Is So Slow,” ZDNet, February 23, 2008.
Rao, L., “The Final Tally: More Than 1100 Cities Apply for Google’s Fiber Network,” TechCrunch, March 27, 2010.
Spangler, T., “Cisco Clarifies 100-Gig AT&T Backbone Claim,” Multichannel News, March 9, 2010.
Swanson, B., “The Coming Exaflood,” Wall Street Journal, January 20, 2007.
Thompson, R., “DSL Internet vs. Cable Internet,” High Speed Internet Access Guide, March 23, 2010.