Wireless Access Substitution?

At a recent carrier event hosted by Denver-based carrier neutral co-location provider Comfluent, an independent provider of Ethernet access in the Rocky Mountain region was musing about how he would be able to compete with the likes of Qwest in the enterprise space, not to mention Time Warner Telecom. One of his particular worries was wireless substitution in the Ethernet access domain.

So Brough Turner notes that DoCoMo, in an experiement, has squeezed about 50 bits into one Hertz of free space bandwidth, albeit at the cost of 100 MHz of spectrum in an unused and very high frequency. The best modulation techniques out there commercialy for mass applications run at about 50 bits per Hertz (the symbolic information conveyed by the oscillation time of a single radio wave in one second). Impressive.

But there are issues. That much continuous spectrum is going to be hard to find in North America or Europe. And when found, it will be at very high frequencies. And the problem with very high frequencies is signal attenuation. Signals won't go very far. Nor will they penetrate tree leaves or walls very well at all. So while useful for mobile, outdoor or short-range indoor distribution, really high modulation techniques won't rival optical Ethernet.

A single optical wavelength replicates the bandwidth of the entire radio frequency, plus all the optical frequencies, and every wavelength can replicate the same frequency again. Every fiber can replicate all of the bandwidth provided by all the multiplexed waves. There's just no way free space is going to keep pace with that.