The future looks bright for "broadband wireless," a term I've always taken as an oxymoron. I've been negative on wireless since deciding against radio for Ethernet in 1973.
However, US Federal Communications Commission Chairman Bill Kennard and TeraBeam CEO Dan Hesse are shedding new light.
Kennard's term ends in June 2001, and like other prominent Bills he's working on his legacy. Kennard is proposing that the 21st century FCC stimulate competition in speeding communications to all Americans and focus on managing the electromagnetic spectrum. See http://www.fcc.gov.
Kennard has long promoted wireless, but recently has been making it more important than wireline. And he's not just talking about mobile wireless, but also fixed wireless.
Knowing the difference between careless and carefree, I've always disliked the term wireless. Why not wirefree?
Fixed is another term that needs work. Does it mean that mobile wireless is broken? Was it recently neutered? No, wireless can be immobile, and thereby broadband.
Although we're clearing things up, wireless is confusingly specified in Hertz. AM Radio stations, for example, each have a carrier frequency stated in kHz (kilohertz), around which they use a band of frequencies, the width of which is also measured in kHz. Then there's the speed of data through a frequency band, also often called bandwidth, which has a baud rate (transitions per second), often expressed in Hertz. Data rates through frequency bands are best expressed in bits per second.
With that, Kennard's FCC worries about shortages of electromagnetic spectrum were settled. They've been shaking things up by gathering underutilised frequency bands and offering them up for auction. For example, frequencies long reserved for television channels 60 to 69 are now being recycled to compete with cable television modems and DSLs.
As if this were not causing enough excitement, the FCC recently filed a proposal for sharing existing frequencies using new spread-spectrum or code-division technologies. They're calling this UWB (Ultra Wide Band), and if interference tests go as expected, they will allow unlicensed UWB below 3GHz across old radio bands.
UWB doesn't refer to the speed of data, which can be high or low, but to how very many frequencies are used. UWB applications include radar for objects below ground, secure communications for police, fire, and rescue, and the big one: broadband wireless for internet access.
Now, with the FCC scrounging underutilised frequencies up to 300GHz in the spectrum it manages, and with them now proposing to share used frequencies, what more exciting could happen with broadband wireless?
Enter TeraBeam Networks, a 1997 startup that just raised $555 million and recruited AT&T wireless star Dan Hesse as CEO. TeraBeam and Lucent are rolling out Ethernet-compatible 160Gbps systems offering 1Gbps Internet access to office buildings in major metropolitan areas.
TeraBeam's service is wireless and way broadband. What's more, it's being rolled out without the FCC. TeraBeam's Gbps service operates at frequencies in the electromagnetic spectrum that are about 1000 times higher than the high end of the FCC's range.
TeraBeam's 1Gbps point-multipoint broadband wireless uses infrared light at 190THz. So actually their service is not wireless, but, as they trademark their system, Fiberless Optical. I would have called it Fiberfree.
Critics say pulling fibre is still your best bet. They worry that TeraBeam's 3-kilometre window-to-window transmissions will have a hard time penetrating fog. Hesse says TeraBeam has been running trials in Seattle for 18 months to be sure broadband fibreless works in all but catastrophic weather conditions.
And anyway, fibers have a hard time penetrating concrete, which is why, says Hesse, less than 2 per cent of office buildings in the United States have fibre access. See http://www.terabeam.com for the remaining 98 per cent.