Wireless touted as source of bandwidth in the local loop
An expert in wireless technology offers a way to bypass the local telephone loop`s bottlenecks, which prevent large amounts of bandwidth from reaching homes and small businesses.
StePHen n. brown, mev inc.
For years, fiber`s proponents have urged replacing the copper loop with a fiber one--a proposal that has been resisted because of the expense. A well-known proponent of wireless communications is offering his own solution: Use spread-spectrum technology to bring large amounts of bandwidth to the public. This proposal emerges from a so-called open letter by David R. Hughes of Old Colorado City Communications to Reed Hundt, chairman of the Federal Communications Commission (fcc). Hughes is the principal investigator in a National Science Foundation (nsf) project that assesses the performance of spread-spectrum devices that are providing schools with local-area and wide-area networks and Internet access. The nsf considers the project a sterling example of what wireless technology can do for infrastructure and has provided more than $500,000 of funding. Armed with knowledge gained from his work, Hughes berates Hundt for not giving supporters of spread-spectrum technology an opportunity to speak at the fcc`s bandwidth forum (see Lightwave, April, page 32): "I am...upset both with what you said, and...[with] what was not said because you did not give spread-spectrum manufacturers a seat at the table....Why?"
Converging economic and political events may answer Hughes`s question. Spread-spectrum technology is a spin-off of electronic cryptography, which codes a message, subdivides it into several parts, and transmits it in a camouflage of electronic noise by "spreading" the message over a bandwidth much larger than the message itself. This is an ironic and perhaps a wasteful use bandwidth. For example, fiber-to-the-home networks are criticized because they have too much bandwidth, so much in fact that no one can think of appropriate ways to use it all. Therefore, fiber-to-the-home is portrayed as bandwidth overkill. If this were a meaningful principle in public policy towards communications infrastructure, then it should be adhered to for all technologies. This is not the case because bandwidth overkill is a virtue in wireless networks. In spread-spectrum transmission, most of the bandwidth is deliberately kept empty to camouflage the message itself. Thus, the information content of a spread-spectrum message will always be far less than what could be achieved in a fiber network. Also, if the demand for bandwidth is expected to grow exponentially in the coming years, then spread-spectrum`s information-carrying capacity will be quickly exhausted. Despite this risk, Hughes and his fellow proponents suggest that the technology can benefit the public.
"I am sitting here connected to you at 3-Mbit/sec spread-spectrum...bypassing the local loop to our office...hav[ing] higher bandwidth than [what] US West can sell me...[and] using none of their resources," asserted Hughes in his letter to Hundt. "Multiply that by 100 million and you have a solution to [providing consumers with a cheap source of bandwidth]....If you encouraged and supported that solution...the regional Bell operating companies would both find the money and technical solutions to their local-loop crunch without trying to impose new fees [on Internet providers]."
However, the weakest aspect of radio communications is interference, where one signal corrupts another as they both compete for the same frequency. This explains why most two-way communications networks in the United States were developed through wire technology instead of radios.
The term "interference" also suggests that the spectrum is a limited resource that could be used up like a highway overcrowded at rush hour. This was one justification for the creation of the fcc in the 1930s and the development of radio-licensing procedures to conserve spectrum and protect the public interest. If interference is still a fundamental problem in radio communications, then providing bandwidth through 100 million spread-spectrum radios would seem to create a monumental traffic jam. Hughes disputes this in his letter, saying that a "recent doctoral thesis from mit [Massachusetts Institute of Technology in Cambridge, MA]...conclusively determines that millions of spread-spectrum radios can now, because of processing gain now possible, operate in metropolitan areas, without practicable levels of interference."
Last barriers to progress
Thus, Hughes suggests, the last barriers to progress are the fcc`s inappropriate technical specifications that prevent the technology from reaching the marketplace. He says: "The fcc, by restricting the [spread-spectrum device`s] power [output, in order] to minimize interference...rather than holding the manufacturers to a high standard of process gain, which [also] minimizes interference, and by keeping the frequencies in the bands that require ever-more rigorous `line-of-sight` placement...[has] limited the ability of the radios to fully serve the markets."
His criticism of the fcc`s specifications is apparently based on expected improvements in spread-spectrum radios. The fcc allows the radios to be used only in certain frequency ranges that do not require the operators to get a license from the agency. The most commonly used range is the 902- to 928-MHz band. Unlicensed devices can also operate in the 1910- to 1930-MHz and 2390- to 2400-MHz frequency bands. In addition, unlicensed devices can share the 2390- to 2400- and 2402- to 2417-MHz bands with the amateur radio services. The manufacturer has the fcc`s authorization for use and passes that authorization on to the radio`s purchaser. The radios have an effective line-of-sight range of 5 to 7 miles and transmit at a rate of 115 kbits/sec. Line-of-sight means that the signal can be blocked by a concrete building or a hill and that the signal cannot bend over the horizon. When obstacles block the path, the typical solution is to add a repeater and another radio, which together improve the range--but at additional expense. Hughes says he is working with radios that have a 25-mile range, with a speed of 10 Mbits/sec operating around the 2500-MHz band. The implication of his criticism is that the new devices should be allowed higher power output and be given an opportunity to operate in bands other than those already assigned.
Hughes`s claim that a doctoral thesis from mit shows how to eliminate "practicable levels of interference" is disputed by Marvin Sirbu, professor of engineering and public policy at Carnegie-Mellon University in Pittsburgh, PA. Regarding the thesis, Sirbu says that "the analysis makes a very large number of simplifying assumptions....[There is] no accounting for the hills and valleys or even buildings that characterize real terrain....There will still be continuous changes in the behavior of the radio path...due to weather or movement of reflecting objects such as cars and trucks....I wouldn`t get too excited until there is some experimental evidence that...the real world is accurately reflected in the model [used in the thesis]." Sirbu does not dispute the viability of spread-spectrum but just the particular method suggested by Hughes: "There are other approaches to the use of spread-spectrum in licensed as opposed to unlicensed bands...which have substantial promise....I`m willing to bet we will see a million direct sequence...personal communications systems handsets operating in licensed spectrum long before we see a million unlicensed [spread-spectrum] radios communicating using the model [in the mit thesis]."
Economic and political issues
Behind the technical arguments are larger economic and political issues. If Hughes were correct in asserting that interference is no problem for unlicensed spread-spectrum radios, then there is no justification for the fcc to confine unlicensed spread-spectrum to the currently assigned MHz ranges. Also, there would be no reason to require a license for any spread-spectrum radio. The purpose of a license is to limit the number of users in the spectrum and prevent the overcrowding that causes interference. If there is no interference, spectrum is not like a highway that can be overcrowded at rush hour. Instead, spectrum becomes a resource that is too plentiful to be controlled by any company and thus free to everyone: Its supply can never be exhausted. Consider this in light of the fcc`s spectrum auctions, which treat spectrum as if it were private property and as if a company could corner the market on spectrum and then rent it to radio users. In the past two years, the fcc`s spectrum auctions have raised more than $24 billion from companies that apparently believe the spectrum is a limited resource that can be controlled and rented to radio users. In his letter, Hughes tells Hundt: "You know as well as I do that a huge part of the motivation of companies that have bid those obscene amounts...is to get a government-granted monopoly on that spectrum."
In the spectrum auctions, at&t purchased 10 MHz for use by a new product that the company unveiled to a group of regulators. In a speech to the National Association of Regulatory Utility Commissioners, at&t president John Walter said, "We are announcing the creation of what we believe will be the communications medium for the twenty-first century." at&t`s product gives consumers the ability to use their wireless phones not only on the road but also at home. In residential neighborhoods, a single antenna mounted on a utility pole can link 2000 homes to a digital switch. Nick Kauser, chief technology officer at at&t Wireless Services and vice president of at&t Labs, was quoted: "We are [offering the]...quality that people have come to expect from fiber optics, but we`re doing it over radio waves....We`ve created a wireless fiber-optic system." Neither executive mentioned spread-spectrum but that is the basis for the new system.
Imagine what would happen to at&t`s "wireless fiber" market and to the government`s revenues from the fcc`s spectrum auctions if Hughes`s claims about unlicensed spread-spectrum gained credibility. Perhaps this answers Hughes`s question. q