WiFi and wireless

Wireless Internet access does not have to be slow, but it often is. I'm at the beach this week, and the access point is about thirty feet away, right at the edge of the property, but the speed of the paid service ($28/week, much higher than in hotels and other venues) is abysmal.

Wireless is often oversold by providers, who don't provision adequate backhaul and/or try to cram too many users on each access point.

The Public Knowledge folks have published their recommendations for how the FCC should handle the impending auction of 700 Mhz radio spectrum for broadband use. This frequency range, down around the broadcast TV spectrum, can carry a broadband data signal many miles and could be a boon for underserved areas waiting for community fiber efforts to build out fiber.

There is an AP article circulating this morning about failing muni WiFi projects (not yet on the Web). This is something I have been predicting for a long time, based on the past performance of early WiFi efforts.

Here is a short list of problems with municipal WiFi-only efforts:

If you like to believe everything vendors tell you, WiMax will solve all our broadband problems, give us younger looking skin, and get rid of grey hair.

WiMax has been lurking for ages, a technology that has been "just a few months away" for at least three years. But it takes a long time to bring an entirely new wireless technology to market, with extensive testing required to make sure the systems don't interfere with other wireless systems, among other issues.

India has announced an ambitious plan to provide free wireless broadband throughout the country.

It is not at all clear that "free broadband" is sustainable. The longstanding problems with free services (in any market, not just broadband) include market distortion and low quality service.

This article [link no longer available] speculates on whether or not Google has a mobile phone in the works. It would make sense for Google to do that, since Google now has a wide array of Web-enabled applications and services that would work nicely on a large screen mobile phone. The phone and its associated service might even be free or very low fee; if it was, Google would recover its costs by restricting what users can do on the phone and/or by interspersing ads with service access (you might have to view an ad to make a phone call or do a search).

Another study of study of tumors caused by cellphones has produced a conclusion that suggests an increased incidence of a rare kind of tumor called a glioma on the side of the head where the cellphone is used most. The study was done in Finland with people that had the rare kind of tumor, and their cellphone usage was compared to those who did not have tumors. Long term users of cellphones (more than ten years) had a higher risk of developing the tumor.

The Apple iPhone is being widely criticized for having relatively slow data service (about 256 kilobits/second) compared to Verizon's much speedier EVDO data service, which can run two or three times faster. The wireless wars have whipsawed back and forth over the past several years. Four or five years ago, many of us, including me, were enthusiastic about the potential of WiFi winning the wireless connectivity wars. But as the shortcomings of WiFi became clearer, Verizon began deployment of their EVDO wireless technology, which is piggybacked on top of their existing cellular network.

In a widely reported story, a Danish study indicates a no more than normal incidence of cancer among cell phone users. Unlike many other studies, this one involved hundreds of thousands of people--a number large enough to make it statistically reliable. Researchers caution that the we still don't have enough long term data to know for sure that cellphones are completely safe. There is widespread agreement that cellphone and cordless phone frequencies in the gigahertz range do cause measurable changes in cell activity. What we don't know is whether or not this causes health problems.

This article from mid-summer just came to my attention. It discusses some of the current municipal WiFi projects and the problems they are having. WiFi vendors tout the low cost of wireless and the "easy" installation--stick up a few towers and you are done. What they tend to leave out of the sales pitch is that current WiFi systems often have trouble penetrating trees with leaves on them, don't penetrate walls well, and the signal does not go around corners. Here is a portion of the article:

The Boston airport administration has tried for two years to force out every WiFi provider in the airport except the one with whom they signed a contract. This meant that travelers did not have a choice of providers, and that free WiFi in airline frequent traveler lounges had to be removed by the airlines. Logan Airport officials claimed the WiFi providers were causing radio interference with airport operations, thereby endangering safety. Right. Unlike the thousands of cellphones crowding the very same airwaves.

Network Computing has a short article with a headline that touts, "City governments are offering metro wireless services with speeds and latency that can't be beat." Sounds interesting, right? But if you read all the way to the end, where the article discusses the fabled WiMax, which will supposedly solve all the world's broadband problems, you find out that WiMax's multimegabit speeds drop to "1 to 2 megabits only at the outer edges."

Swedish-Finnish telecom company TeliaSonera has started selling hybrid phones that will automatically make phone calls via the Internet when in range of a WiFi hotspot, and use the normal cellphone network when not in a hotspot. Some other dual mode phones have been available, but this is the first phone (manufactured by Samsung) that will switch automatically between the two. The firm is targeting in home use first, which is clever, because we make a lot of calls from home.

This article discusses Intel's belief that wireless networks in the home are inadequate for high definition television. The highly touted 802.11g, with a theoretical maximum bandwidth will only deliver about 22 megabits under the best of conditions in an in-home network, and performance could be much worse. In other words, it will barely handle a single channel of HD TV--as long as you or any one in the family is not doing anything else on the network.

I have been traveling a lot lately, so I have had the opportunity to try out a lot of WiFi hotspots. The good news is that most hotels now have some kind of WiFi available. Business travelers can stay almost anywhere and count on having some kind of Internet access. I have not had to use my Earthlink dial up account in many months. Many "budget" hotels have some kind of free WiFi, with service about what you pay for it (nothing). Service at the far end of the hall in the hotel may be poor, and speeds may be pokey.

A coffee shop in Washington state had a guy arrested for using their "free" WiFi signal for three months without buying anything. The alleged WiFi poacher apparently just sat in his car in the parking lot nearly every day and used the Internet access without buying anything.

Citywide WiFi projects in Sacramento, California and St. Cloud, Florida are both having problems, supporting my long-standing contention that these efforts are risky. MobilePro, the company that got a city government contract to blanket the city with WiFi, is pulling out of the project entirely after the company and the city could not agree on how to finance the project. What's mind-boggling is how the company and the city agreed to move forward without a clear understanding of how the system would be paid for.

Anthony Townsend, an expert on the social impacts of technology, has written an important article about community WiFi projects. Townsend is concerned that community leaders, in the rush to show some progress in broadband, are inking deals that give away too much.

In a perfect world, we would throw our cellular phones away and move as fast as possible to an all Internet wireless system, using VoIP to make phone calls and the same packet-based IP transport for all other kinds of data--one kind of transport system for everything--voice, video, Web, you name it.

I am not a big fan of me-too municipal wireless projects. Wireless technology remains in flux, with new equipment and systems coming online constantly. Interference and bandwidth issues have to be considered very carefully when designing these systems. And you have to know how you are going to pay for the network management and maintenance.

In other words, a community should not be planning a big wireless initiative just because "that's what they are doing in Philadelphia."