Question: How many 802.11 wireless transceivers would it take to cover a 50-mile stretch of road?
Answer: Only one, if its a railroad. Just put the transceiver in the train. Then fly a plane with broadband microwave on board just above the train.
No, I aint kidding.
Daunted by the capital cost of trying to get Wi-Fi to travellers, the worlds ISPs are starting with rail commuters. And to reach train riders, some of the tricks they are getting up to are starkly unbelievable.
Fitting a Wi-Fi access point to a train is relatively simple. Theres plenty of electricity with no anxieties about battery life. Signal propagation, inside a metal tube, is usually excellent. The train almost acts as a wave-guide. The problem is that train commuters arent really interested in networking with each other. They want access to the outside Internet world, and thats really quite a challenge.
The first, obvious trick is to stick an antenna on the train to track a satellite. In really flat countries around the equator this works flawlessly. But move out of the tropics and into hilly country or urban areas and the satellite is often lost below the horizon. Europeans, as a result, dont think much of geostationary satellites. You can never see one when you really need it.
The result is that theres an amazing amount of interest in low earth orbit wireless, or LEOs as they are called, and in other even wackier ideas. Tethered aerostats–big balloons–are being experimented with, but the Capanina project is going way beyond that, and looking for other ways to get an antenna into the sky without launching a satellite. Furthermore, in London last week, the surprise star of the conference on “Practical Strategies for Implementing Wi-Fi in Passenger Transport,” was North American railroad pioneer, Pointshot, which is already providing Canadian and Californian rail commuters with free Wi-Fi and moving into the area of wanting to start asking them for real money.
Go Fly a Blimp
Contd”> Behind the awareness that there are people planning to use the Internet on trains, the real research is going to be in two new areas. Capanina looks at permanent High Altitude Platforms, or HAPs, and the industry is looking at business models. Both are the subject of real skepticism.
On paper, the figures look exciting. While you cant seriously plan to feed automobile commuters satellite data, drivers arent really a prime market for browsing. Train commuters, however, really are. Consultants BWCS produced a report late last year, for example, predicting that worldwide laptop users “will spend more than 12.6 billion hours on trains during 2003 as compared to just under 2 billion hours spent at airports.” But the Wi-Fi optimism assumes that someone finds a way of feeding the data to the trains.
Currently, feeding the data to trains is a masterpiece of chewing gum, wet string, and tin cans. You have the satellite link, of course, but youve go to assume that youre going to lose the signal behind buildings, in tunnels, and under railway stations. So you have to set up a bonded link using multiple other sources, CDMA data, GPRS, and local Wi-Fi connections, coupled with special transceivers in the middle of long tunnels. It works, but its hardly ideal.
Instead, Capanina looks forward the development of fuel cells and photovoltaic aircraft like Nasas Helios. Capanina researcher Dr. David Grace at York University is eagerly predicting that unmanned HAPs will be flying overhead at 80,000 feet for six months between takeoffs and using steerable antennae to cover micro-cells of 4 miles diameter. The remote controlled planes dont even have to be overhead, he says. If safety anxieties become heightened, the planes can serve most of the UK from flight patterns well out to sea.
And users on the ground will get over 100 megabits per second of data. Thats certainly enough to satisfy a train full of commuting executives who would otherwise be restricted to whatever the mobile phone network could provide.