Smart Mobs

by Howard Rheingold

  His experience in Mongolia gave Hendricks the idea that he could experiment with new technologies that could provide even higher bandwidth and greater distances by finding sympathetic partners outside the FCC’s jurisdiction. In 1998, Hendricks was introduced to the Crown Prince of Tonga. Because the policy regarding use of the electromagnetic spectrum in Tonga is whatever the Crown Prince says it is, Hendricks’s Tonga company, licensed in Tonga as a common carrier, will be free to experiment with technologies and power levels prohibited by FCC regulation. Then, as Hendricks told a Wired magazine writer, “I learned about an FCC initiative to improve communications services on Indian reservations, and it dawned on me that 551 sovereign nations were close at hand.”57

  While Hendricks was unwiring North American Indian reservations, Hughes paid a visit to the land of “nine generations of rebellious Welsh minister” forebears. At a New Year’s Eve celebration in a pub, Hughes recalled, “I realized that most of the Welsh population lived within wireless range of a pub. A couple of hundred dollars per pub, easy install, and you have instant national broadband infrastructure for a country 50 miles wide by 150 miles long.”58 An activist from Wales came to Old Colorado City to videotape Hughes proclaiming his vision for the future of Welsh telecommunications. The video is online—a remarkable combination of tour, demo, how-to, and polemic.59 In the activist’s video, Hughes emphasizes the same benefit to local small businesses he pushes in rural Colorado: Wireless broadband isn’t just bringing more things to consume and buy but offers a channel to create, sell, and promote their local point of view to the rest of the world. Hughes returned to Wales in February 2002, as a guest of the Welsh Digital College, to evangelize Welsh movers and shakers. While the United States and other nations tie up the development of their communication systems because of the investment by telephone companies in 3G licenses, watch places like Wales to see the future media sphere emerge first. Or visit Okinawa, which some policy analysts in Japan are pushing to become a “radio haven” for the development and deployment of advanced wireless technologies.60

  Under National Science Foundation sponsorship, Hendricks is attempting to bring wireless broadband access to the Turtle Mountain, Fort Berthold, Fort Peck, and Sitting Bull community colleges.61 In these communities, basic voice communication infrastructure is nonexistent. Wireless broadband can bring telephony to the reservation along with high-speed Internet access to its educational institutions at a fraction of the cost of “wireline” telephones alone. Hendricks is collaborating with researchers at the University of California, San Diego, who have deployed solar-powered relays to bring wireless broadband to the La Jolla and Pala Native American reservations.62 Investigators reported that the high-speed wireless network using 802.11b access points “proved to be very cost effective and relatively easy to install when compared to wireline options. For one thing, installation took months instead of years and a few hundred thousand dollars instead of a few million dollars.”63

  Because Hendricks intends to use new technologies that are currently prohibited at higher power levels and in wide bands of the spectrum, his partners on the reservations are prepared for possible conflict with the FCC. One of these technologies, “wideband,” doesn’t focus on a single frequency but transmits encoded pulses, each a billionth of a second long, across the entire spectrum, at extremely low power levels—literally in the background noise of radio technologies that concentrate on specific frequency bands. If each transmitting device can share the common resource of spectrum by transmitting during a billionth of a second when no other device is using it, and the receivers are smart about what to listen for, a gargantuan amount of spectrum room becomes available for sharing bandwidth. Some researchers believe ultrawideband data transmission could achieve gigabit speeds—a billion bits per second.64 Gigabit wireless would turn the present technological, regulatory, and economic regime of networked communications upside down. Today, a T1 line is 1.5 megabits per second—1/666 of 1 gigabit per second—and costs $1,000 a month or more.

  The FCC, the military, and emergency services fear that a major potential problem with the new wireless technologies is “interference”—the danger that multiple broadcasters on the same frequency could make it difficult for receivers to distinguish one competing signal from another. As I’ll discuss, this fear is based in part on the limitations of radio technology of the early twentieth century. However, caution is clearly in order when it comes to fire department, ambulance, military, and police communications and GPS devices. The FCC is properly worried that the noise from too many spread-spectrum users could interfere with emergency communications and threaten public safety. Technologists counter that the answer is not in regulation of spectrum but in improving the devices used to broadcast and receive signals. National security interests are also concerned that wideband technologies are difficult to intercept and easy to encrypt.

  Hendricks, Hughes, David Reed, and Lawrence Lessig are part of a growing movement of advocates who would like to see much more of the spectrum opened, far beyond today’s tiny experimental loopholes. Another technology that David Reed and Hendricks both mentioned to me, “dense-packet radio networks,” has the valuable property of automatically growing carrying capacity as use expands, accommodating, potentially, billions of simultaneous transmissions in the same region. Indeed, an MIT student’s doctoral thesis proved that efficiency of spectrum use can increase as the number of devices increases, provided they are smart devices that cooperate electronically to use the spectrum efficiently.65 If you could buy a dense-packet radio receiver and turn it on, not only would you be turning on access to the Internet for you and your local network, but your receiver would serve as a relay for other nearby, similarly equipped transceivers. Your radio would act as a “router,” picking those bits that are meant for it out of the stream and passing along all the others to the nearest other radio. The network’s carrying capacity grows more plentiful and each radio needs to use less power as more people join the network, provided that each user serves as a relay for nearby routers—one of those “sheep shit grass” situations that economists describe more decorously as “the law of increasing returns.”66

  Within ten years, Intel intends to integrate “software-configurable radios” in every chip it manufactures; the way a software-configurable radio uses spectrum can be changed on the fly by the computer, so a single PC add-on could switch from being an FM radio tuner to a cell phone to an 802.11b card.67

  Spread spectrum, wideband, software-configurable radios, and dense-packet radio networks don’t exhaust the list of technologies competing to make traditional spectrum allocation obsolete. We’re accustomed to thinking of our telephones, for example, as devices that plug into the telephone or cable company’s network. What if the devices themselves could become the network? Dense-packet radio networks and other state-of-the-art but ahead-of-the-regulatory-structure technologies make possible fully ad-hoc, self-organizing, multi-hop mesh networks. Imagine telephones that directly communicate with each other, relaying signals from device to device the way Usenet nodes do, without using any other communication network other than the telephones that happen to be nearby.

  Yet another new technology is known as “ad-hoc peer-to-peer networking.” If any one node of what is also called a “mesh” network has a fat enough pipeline to the Internet, then the network of devices can cooperatively distribute the bandwidth. In a mesh network, each node also can serve other users simultaneously as infrastructure, like a cellular system without fixed cells that relays messages among telephones.

  A company called MeshNetworks was funded by DARPA (today’s successor to ARPA) to make it possible to “parachute two or three thousand soldiers in the middle of nowhere and have them instantaneously form an ad-hoc peer-to-peer network and communicate.”68 The company is planning a $35 chip that could serve as a wireless access point; telephones equipped with such chips could serve as relay stations for other nearby telephones. Imagine those 1,500 people
in Shibuya Crossing, and the people within range of them, rippling out through Tokyo, using their telephones as the routers in an ad hoc communication network. Mesh technologies can transmit data at 6 million bits per second, enough for data, voice, Internet, audio, and video. In February 2002, the FCC granted MeshNetworks an experimental license to test their technology in limited frequency bands.69 Nokia markets “wireless routers” based on mesh network technology in the unlicensed bands.70 Strange as the notion seems, population densities and mobile telephone penetration in major metropolitan areas make self-sufficient peer-to-peer networks entirely possible.71

  I wasn’t surprised to discover teenagers growing mesh networks by putting their toys together. The “Cybiko handheld” is an inexpensive device for the youth market, created by a Russian company. Priced at around $100, Cybiko (Japanese for “cyber girl”) devices combine a walkie-talkie, texting terminal, FM radio player, voice recorder, game and music player, email, and organizer.72 Cybiko base stations can provide up to 200 Cybiko users with wireless Internet access, email, and instant messaging. America Online is one of the investors. As of this writing, over half a million first-generation devices are in circulation. The next generation incorporates protocols for ad hoc peer-to-peer networks.

  Hidekazu Umeda created mobile peer-to-peer software protocols that DoCoMo and other traditional wireless service providers can’t be happy about, since his mobile p2p software could turn all the mobile telephones, personal digital assistants, Cybikos, and other wireless devices in Tokyo into one giant ad-hoc network that moves voice and data around without the intermediation of the traditional telecommunication networks.73 I met with Umeda and his colleague, Yuichi Kawasaki, after normal office hours in the Web design firm that employed them during the day. There was a slight irony to convening a discussion with hard-core digital revolutionaries in the rosewood-paneled presentation room on the seventeenth floor of Tokyo’s posh Cerulean Tower. Kawasaki dreams of a mobile gift economy: “I’d like to see people able to use their mobile devices to swap data/games/music completely outside of centralized control. With technologies like Bluetooth, virtual communities could be formed purely by the exchange of data between mobile devices.” In other words, they want to enable smart mob formation.

  “Bluetooth” is the name of a standard for short-range radio communication chips that can link the broadband Internet to the ad hoc networks of devices in a computation-pervaded environment. It won’t be a technology most people will use, but it will be inside devices most people will use. In 1994, Ericsson Mobile Communications started studying low-power, low-cost radio devices as a way to eliminate cables connecting mobile telephones, headsets, PCs, and printers.74 The Bluetooth Special Interest Group, founded in 1998 by Ericsson, Nokia, Intel, IBM, and Toshiba, was joined by Microsoft and Motorola—almost everybody in telecommunications.75 Whenever two devices equipped with Bluetooth chips are within ten meters of each other, they automatically open communication; each Bluetooth chip periodically broadcasts a query for other devices in the area. Just as boxy cathode ray tube computer displays are being replaced by flat-panel liquid crystal screens, cables will give way to short-range wireless chips over the next ten years. As a side-effect of replacing cables, Blue-tooth chips enable the creation of local ad-hoc networks—the kind that pervasive computing devices will use.

  The industry heavy-hitters who back Bluetooth anticipate widespread adoption of Bluetooth when the price per chip drops to $5. Like WiFi, Bluetooth technology must solve existing security and interference problems. Nevertheless, UPS plans to deploy Bluetooth, along with 802.11b LANs, in its worldwide distribution hubs. 76 Forrester Research analyst Lars Godell forecasts “235 million Bluetooth-enabled mobile phones, PDAs and laptops, versus 22 million WLAN-enabled devices” by 2006. 77 Bluetooth is the immediate-vicinity link connecting mobile devices and roaming people in a computation-pervaded environment, from your telephone to your PC to your printer to your MP3 player, to a vending machine. WiFi is the broadband zone at home, work, or the café where you can plug into the worldwide Net.

  When it comes to wireless communications, however, politics is as important as technology.

  Open Spectrum versus the Good Old Boys

  Law professor and activist Lawrence Lessig has prepared a pro bono legal defense team for Dewayne Hendricks. They intend to challenge the legal basis of current spectrum regulation. Tribal colleges were informed by the project administrators that “Wireless technology raises questions about who controls the spectrum on reservations. Campuses that are interested in participating in the project need to demonstrate awareness that there could be problems with the Federal Communications Commission, local telephone companies, and others.”78 The early skirmishes in the battle for the electromagnetic spectrum have been joined.

  One new radical idea is backed up by solid science: Abandon the idea of a regulated spectrum and dumb devices, say the advocates of “open spectrum,” and turn the spectrum into a commons anyone can use as long as they use a broadcasting or receiving device that is smart enough to cooperate with all the other devices. Why not? That’s exactly how, and why, the Internet worked so well. There is no central regulation of how the Internet’s communication bandwidth is used, just a standard protocol for connecting.79

  Regulate the devices, not the spectrum, say open-spectrum advocates, and create conditions for entrepreneurial innovation and broad economic benefits that extend far beyond the existing large corporations that now are the sole beneficiaries of spectrum regulation. To some, this position sounds vaguely communistic. “Are you a communist when you use your cordless phone? Because the unlicensed band your cordless phone uses is a little commons,” retorts Dave Hughes, the last guy you’d want to call a communist to his face.80 To complicate the issue, advocates of “open-source software- defined radio” are designing radios that can transmit and receive on any frequency and modulate with any scheme (i.e., AM, FM, spread spectrum), making it far more difficult to regulate devices, since modifications take place only in the software.81

  During my journeys into cybersociology, I had discovered Elinor Os-trom’s studies of commons that were not tragically mismanaged and encountered the notion of “public goods” in the experimental economics games probing cooperation. And Lawrence Lessig had referred to an “innovation commons” built into the Internet’s end-to-end architecture. When the same notion showed up in the hot center of policy debate concerning wireless Internet regulation, another conceptual Schelling point in the smart mobs literature revealed itself. The commons is where smart mobs could gather; commons are what smart mobs have the potential to create and what they have to be careful not to overconsume.

  I asked Lessig to explain what he meant when he said that the Internet was a public resource “held in common,” rather than divided up among private owners. Lessig pointed to the difference between railroad and highway regulation. In a railroad, the individual cars have no intelligence, and only one train can be on a specific stretch of track at a time, so railroads must be carefully centrally coordinated. Automobiles, however, presumably have intelligent drivers who can figure out how to get where they need to go without colliding with other vehicles. Central coordination is no longer required. “The highway is a commons,” Lessig explained. Everybody has access to the highway, nobody needs permission to use the highway system, anyone can start a trucking company and use the system. The devices that you can use on the highway commons are regulated—you can’t drive a tank, and if you have no lights, you’ll be pulled over. Lessig noted, in light of the railroad/highway comparison, “Regulation of spectrum could move from the world of railroads, where central coordinators have to figure out who uses the track when, to the world of highways, where smart devices figure out how to use their common resource as they actually want.”82

  New York University law professor Yochai Benkler proposed in a 1998 article that current technology puts the present rationale for licensing spectrum into question.

&n
bsp; The central institutional choice regarding wireless communications is whether to rely on centralized control by identifiable organizations, or on multilateral coordination among numerous users. On the one hand, it is possible to treat spectrum as a resource whose use must be centrally determined by someone with the power to decide how wireless communications equipment will be used in a given spectrum unit. That entity can either be “the owner” of the defined spectrum unit, if privatization is chosen, or the licensee operating within parameters set by the regulator, if licensing continues to be the rule. On the other hand, it is now technically possible to rely on standards and protocols to enable multilateral coordination of transmissions among equipment owners, without identifying any person whose choices trump those of all other potential users. The central question then is no longer how to allocate spectrum channels—how to decide who makes unilateral decisions about who may communicate using a frequency band and for what types of communications—but whether to coordinate by defining channel allocations. While the answer may be that we should permit a commons to develop alongside proprietary allocations, we will fail to permit that development if we continue to misperceive the choice at hand as one between licensing and exhaustive privatization.83

  Benkler used the term “open spectrum” in the summer of 2001, and analyst Kevin Werbach, former counsel for New Technology Policy at the Federal Communications Commission, publicized it in Esther Dyson’s influential Release 1.0, describing the coalition of technologists, academics, and legal activists emerging around the idea of deregulating spectrum.84 The idea is not to do away with auctions but to mix several ways of allocating spectrum and then see which works best. Big players will be able to buy pieces of spectrum at auction, and other large amounts of spectrum will be held as a commons.