Securing Radio Spectrum for Wireless Internet Access

Steve Cisler <sac@apple.com>
Apple Computer, Inc.

Writers have been talking about the convergence of different media and communications since 1915, when they predicted that telephone, telegraph, radio, and movies would merge by 1935![1] But the Net is causing different industries to plan alliances, make huge business deals, and also misunderstand what other industries are trying to do. Because of the enormous commercial opportunities, many new companies are interested in the Internet. This causes tensions among industry, education/research, and government officials. There are some tensions in the world of wireless spectrum allocation.

It is important to remember the big differences between the computer industry and the telecommunications industry: Personal computers developed without government regulation (there was no Federal Computer Commission), while the telephone, radio, and television industries developed in a very regulated environment.

A major conflict takes place when you mix computers and telecommunications. Many people want deregulation; this is part of an effort to force government out of trying to control the phone companies, the radio spectrum, and television. It is happening in the United States, and it is happening in other countries.

Others want continued government involvement in these areas. It depends on which part of society you trust. In the United States, many people, especially on the Internet, distrust all forms of government. They want less involvement of government everywhere. They usually have a strong faith in the market economy, in less control by government officials. They believe that the unregulated market economy is the best way to progress and expand the network and its services.

The Federal Communications Commission

On the other hand, you have the telephone companies, cable companies, and broadcasting industries, which have depended on a regulated environment since the early days of their formation. The most important agencies are the public utilities commissions in each state and the Federal Communications Commission (FCC). The broadcasters and telephone companies have many employees working on regulatory issues, but the computer companies did not--and still do not. Even now, most computer companies do not pay as much attention to what is happening with government regulations as the telecomms industries do. A group called Computer Systems Policy Project includes the heads of Sun, IBM, ATT, Apple, Silicon Graphics, and seven other companies. They have been working on government issues for about four years.

The FCC is an independent government agency with commissioners appointed by the President. They continue to serve even after the President leaves office. The head of the FCC is Reed Hundt, who is a close friend of Vice President Al Gore. He has taken a strong stand that the FCC should help achieve the goals of connecting the schools to the growing telecommunications infrastructure (the Internet). He has helped move the FCC into the computer age by being the first chairman to use electronic mail and by speaking out on various social issues that may be less important to commercial firms. During the NetDay 96 wiring extravaganza on March 8, 1996, he took part (as did other Clinton administration officials) in helping to wire a school.

This agency that controls communications only began using fax machines in 1992; it got e-mail in 1994, and this year it opened up a World Wide Web site, began placing its documents online, and even allowed the public to make comments via e-mail instead of just by paper documents. This shows they are willing to change.

Some parts of Congress want to eliminate the FCC, take away its many powers in controlling and enforcing spectrum regulations, and replace it with a small office with little power. The more radical group, as represented by the Progress and Freedom Foundation <www.pff.org>, wants to sell the radio spectrum and allow the owners to re-sell it or use it any way they can. They believe this will help spread advanced wireless services and make a lot of money for the people involved.

Most people believe, however, that if you eliminate this agency, you will have so much disagreement that you will spend more time in courts trying to win lawsuits, and that federal judges will make technical decisions that are becoming more complex with each new business deal and new technology. Most companies and organizations want the FCC to exist, but they want it to work more quickly.

Here are some basic issues facing the FCC:

Auctions and unlicensed wireless bands

When the FCC proposed to establish unlicensed wireless bands in the 1980s, the plan was for much of the spectrum to be designated for public use with no limit to the power. After the communications industry intervened, the so-called Part 15 rules applied only to a small amount of spectrum and at power levels of less than one watt. Still, a whole industry was born. There are dozens of companies now making Part 15 products for LAN and WAN applications. Generally, the speed is slow, with the major exception of WaveLAN.

In February 1995, an Apple petition to open up more spectrum was approved. The public now has 10 Mhz for license-free use. Apple does not own it. Everyone must use equipment designed for the band. Some of the bands are what the FCC privately calls "garbage bands"; the official term is ISM band, for industrial, scientific, and medical applications. These can cause interference to wireless radios, especially in an office environment. A microwave oven operates in the 2.4 GHz band, so radio equipment to handle this interference generally costs more than radio technology for use in a clear band. And sometimes one kind of radio can even interfere with another one.

How the government treats the country's resources--whether it's grazing land, mineral rights, or radio spectrum--has been a point of contention for many years. With the huge budget deficit, many lawmakers are happy that the recent auctions brought so much money into the federal treasury. The FCC just raised about $8 billion dollars by auctioning off spectrum for PCS (personal communicator services). Auctions for other parts of the spectrum have brought in billions in recent months, and many more auctions are planned.

The auctions only serve those who support licensed spectrum. Licensed spectrum is a good approach if you have the financial resources to make it through the FCC licensing process and to build the network in accordance with FCC rules. This generally means you have a strong business case and are using a proven technology. In addition, there has to be a license for the area you want to serve, and the service has to attract enough subscribers to recoup the capital you have invested in the license and the network.

Unlicensed networks mean no applications, no licenses, no service fees. It is a publicly available resource that entrepreneurs or anyone else can use, and we use the Internet as an inspiration and model when explaining this to the FCC and others in the commercial sector. Although the spectrum auctions have received much of the attention, the FCC's commitment to universal service, serving the underserved, and helping extend telecommunications services to schools makes them quite open to this approach.

Some parties have objected to the "giveaway" of public spectrum, in part because they believe it would devalue what they have already paid for newly licensed spectrum. What we are proposing is not a giveaway but more of a liberation or a set-aside for public use. Imagine a stretch of beach along a lake or ocean, and the government has sold or leased large portions of it to different hotel chains and resort owners who now oppose any public beach being set aside for use by the local inhabitants. The beach serves a public good, and other businesses will thrive there: parking lots, food vendors, and rental of boats and beach umbrellas. It makes sense to have public beaches and beaches leased for exclusive use by companies. So, too, does it make sense to have public spectrum alongside auctioned spectrum.

On May 24, 1995, Apple Computer proposed to the FCC that they set aside 300 Mhz of unlicensed spectrum for public use, what we called the NII Band. This consists of 5150 to 5300 MHz and the 5725-5875 MHz band. Aside from the technical challenges, we had to make assumptions about the political climate:

Following the publication of the filing in the Federal Register, a daily print journal of the U.S. Government, we posted it on Web sites at www.apple.com and www.warpspeed.com. Parties had 30 days to send comments to the FCC. Near the end of that period, the FCC notified people electronically that comments could be sent as e-mail, and this made it much easier for the average person or a school or small company to make its opinions known. Companies that dealt with the FCC used the formal paper process. It is important to remember that this is a public process, even though companies may request meetings with FCC employees to explain their position.

The growth of the Internet and the interest taken by Chairman Reed Hundt has resulted in the process being somewhat more democratic, yet it is clear that those organizations with the best connected lobbyists, technical staff, and political influence will usually prevail over those with fewer resources and less clout. With the passage of the Telecommunications Act of 1996, more people are aware of the importance and the value of the wired and wireless infrastructure. Consequently, the FCC is hearing from more sectors of American society. The ideal is if all parties can get together, agree on the terms of an agreement, and then present their solution the the Commission.

After a period of time, the FCC issues a Notice of Inquiry to gather more facts. If they have enough facts, they may issue a Notice of Proposed Rule Making (NPRM). After another 30 days, during which time there can be comments, they will issue a Report and Order, which is then included in the Code of Federal Regulations and becomes law.

This may take years, or it may take less than a year. We believe that the NII Band proposal will take more than a year, but it is on a reatively fast track. At the time this paper was submitted, the FCC was getting ready to release the Notice of Proposed Rule Making.

Technical aspects of the NII Band

Two needs will be met if the NII Band is made available for the public to use: very high (data) rates (VHR) and community networks. This does not mean that other uses are excluded, but these two are not now available for other frequencies.

VHR would provide the capacity of moving information at rates exceeding 20 Mbps. These would be primarily in-building networks, but the development of these systems is just in the formative stages. HIPERLAN, a VHR function that has been allocated spectrum in Europe and is optimized for packet data, is the most advanced. In the coming years we can expect Wireless ATM to be defined and implemented. We believe it is important for such systems to be refined for the NII Band.

The physical coverage radius of VHR devices could be from a few meters to tens of meters inside buildings. We advocate the establishment of two 50 MHz channels for VHR use, the lower one beginning at 5150 MHz and the upper one at 5725 MHz.

The most important benefit of the NII Band will be its use in what we call community networks. This does not refer to any formal network. It may be a combination of wired and wireless connections with various antennas over medium range and at varying speeds from perhaps 50 Kbps to Ethernet (10 Mbps) within a town, county, or local geographic area. A community network might use an omnidirectional antenna to link a library up to city hall, neighboring schools, and other buildings within line of sight of the library. Hospitals and outlying health clinics could use a combination of dish and omnidirectional antennas to join the different members. Individuals, small businesses, and government offices could also benefit. It could also be used by people with handheld devices in an exhibit hall, a park, or other large open area. Some rural areas are not served by either the cable or telephone companies, or the connections are so bad or so slow that this could significantly change the way a community, Indian reservation, or whole region communicates. Experience with the Telluride (Colorado) InfoZone, which uses Tetherless Access Ltd. 900 MHz devices, leads us to believe that wireless community networks will be very popular and relatively inexpensive, especially when compared with systems that use fixed microwave for the distance connections.

The users of these systems will not have the resources to pay for costly, industrial strength systems, but they will have a need for Internet access, and the NII Band will meet many of those. Most likely, parts of rural America and then developing countries (and counties) will avail themselves of the technology, if the letters of comment sent to the FCC in support of our petition are any indication (see www.warpspeed.com for examples). For an interesting scenario, see how the Taos, New Mexico, La Plaza Telecommunity Foundation is planning to use this, once it is allowed <http://laplaza.taos.nm.us>. For a field test of existing wireless devices for use in education, see the Web site for the NSF test in southern Colorado (Colorado Springs and the San Luis Valley around Alamosa) <http://wireless.oldcolo.com/>.

The cost of these NII Band devices and the appropriate antennas is only an estimate, but they could range from a few hundred dollars for a shrinkwrapped kit with installation video available from a local computer shop to thousands of dollars for a custom turnkey system provided by a consultant/value-added retailer. It depends on the development costs, competing technologies, economy of scale, and the general popularity of such a system.

The diagram below shows a number of configurations for handheld, roof-mounted, and parabolic antenna aimed at another dish antenna as far as 70 km away. These are best estimates, but it can vary by terrain, trees, and weather conditions. The transmitter power is one-tenth of a watt. Unlicensed community networks, however, will not support effective wide area mobile communications that would substitute for cellular, CDPD, PCS, mobile data, SMR, or paging networks. NII Band power levels in the United States will not be sufficient to create longer distance wide area coverage, but other countries may choose to set the power levels permitted at a higher level.

A resource that is too popular?

Some opponents of the concept of medium-reach point-to-point links where there are no formal frequency-sharing requirements or mutual-awareness channel-access etiquettes worry that a tragedy of the commons will occur. In this scenario the resource becomes unusable because of inefficient devices and excessive interference among systems. However, we are not proposing the telephony model; computer data can slow down, just as they do on the Internet, or people can delay transmission. We believe that it is critical that the fear of the tragedy of the commons--i.e., that the band will be so heavily used, so well-suited to meeting the communications needs of people across the country--not be overstated and generalized because of a few cases where congestion may occur. Nor should the band be burdened by too many requirements. Imagine if a Federal Road Commission required a traffic light at every intersection in the United States, no matter how heavy or light the traffic patterns might be. The NII Band should not be burdened by an excess of rules.

Other countries, other rules

Many members of the Internet Society come from an engineering background and may be impatient with the slow processes of government agencies, but it is very important to find out how spectrum is allocated in each country. That information is being collected by a number of companies, as they plan to extend their markets and technologies into new geographies. Individual ISOC members should discover just how spectrum is controlled in their own countries.

It is important to know about the FCC because it is going to determine what wireless technologies developed in the United States of America will be available in Latin America, Asia, and Africa. Also, the FCC provides a model for some countries as they decide how to shape their own telecommunications policy, but each country may handle this differently. In some countries, the policy is set only when the technology is introduced by an outside company. At that point, rules are tailored to help the project without looking at the big picture. Sometimes this will be a good decision; other times it may cause problems at a later date. And some policy is set on an ad hoc basis, depending on who is running the PTT or is Minister of Telecommunications. And in some places the regulatory agencies may exist only on paper, without resources to enforce a country's decisions about spectrum use; experimentation can be carried out without any fear of either causing interference or being prevented from operating by the government. In other places, the military will have more influence, and if forces opposing the central government have any power, use of wireless communications will be denied or severely curtailed. So, too, in countries with monopolies for the telephone company, wireless networks are seen as a threat and may be restricted.

A useful working group within the Internet Society, or in another organization, could pool these data for each country and make them available to members of the Internet community with an interest in developing and spreading these low-cost community networks outside the United States of America.

More information online

http://http.cs.Berkeley.edu/~randy/Daedalus/BARWAN/BARWAN_index.html
Bay Area Research Wireless Access Network (BARWAN)

http://www.ota.gov/publst.html
Wireless Technologies and the National Information Infrastructure

http://www.wirelessinc.ca/
Wireless Communications in Canada

http://www.apple.com/documents/fcc.html
http://www.warpspeed.com
Apple FCC NII Band petition

http://home.navisoft.com/nspi/bzk.htm
Bennett Kobb, wireless expert

Footnotes

[1] Pool, Ithiel de Sola; Forecasting the Telephone; Ablex, 1984.