Radio is cool and exciting because it allows us to send voice and broadband data through the air. Advances in radio technology have resulted in more throughput, greater reliability and better range. And we've made advances across the board, in basic technologies, radio chip architecture and fabrication, packaging, power consumption, antennas, protocols and software.

There's just one little problem, and that's the radio spectrum itself.

You can think of the electromagnetic spectrum as one big wire in the sky, although thinking of it as multiple chunks, each with different properties, is a bit more correct. Most important, radio-wave propagation characteristics vary with frequency, with some bands better for certain applications. Shortwave radio, for example, from 3 MHz to 30 MHz, propagates globally under the right conditions. The 700-MHz bands, for which auction rules were set Tuesday, are going to be very popular for mobile communications applications because such waves represent a great compromise between high information-carrying capacity and reasonable in-building penetration.

But, no matter what, no one is making any more spectrum. For this reason, access to the spectrum is highly regulated and the spectrum itself is very valuable, hence all of the interest in investing billions in 700 MHz.

All of the choice radio spectrum is now fairly well subscribed, and there's plenty of discussion about how to "re-farm" the spectrum for new applications. That's actually what's happening at 700 MHz. That spectrum used to belong to television broadcasters and was used for UHF TV. But the broadcasters clearly weren't making the best use of that chunk of real estate, so the Federal Communications Commission took it back, made any stations in those bands move (conveniently, as part of the digital television cutover process), and can now sell the spectrum for big bucks.

But not all spectrum is radio, and not all of it is subject to regulation. In the U.S., spectrum above 300 GHz is unregulated -- not unlicensed, mind you, which implies regulation, but totally unregulated. The problem, of course, is that building radios that operate at 300 GHz is expensive, and the highly directional nature of the waves there make products operating at those frequencies problematic. But if we go much higher in frequency, say, to the vicinity of 300 terahertz (that's trillions of cycles per second), something wonderful happens.

This is the land of infrared radiation (IR), so-called because it exists just below the red part of the visible-light spectrum. And it's wonderful because we have the technology to build many forms of low-cost emitters and photodetectors that operate in this range. There's a huge amount of spectrum available, literally many terahertz, so we have a huge amount of spectrum (meaning potentially very high throughput) and unlicensed operation on a worldwide basis. Cheap? Fast? What's the catch?

Well, like light, infrared radiation is blocked by solid objects such as walls. While it's possible to bounce IR, such action doesn't work well, so IR tends to be a line-of-sight technology. It's used, of course, in TV remote controls, some specialized wireless LANs, and in a variety of outdoor point-to-point links, often using IR lasers, which provide fairly good range (a kilometer or so) in clear air and in outer space.

But one of the major applications for IR was the Infrared Data Association (IrDA) standard, which resulted in IR technology being installed in many notebook computers and similar devices. IrDA was never a huge success in terms of number of users (although it certainly was in terms of number of units -- check the back of your notebook for a little red window), and it was a bit redundant once most notebooks got on the increasingly wireless LAN. But IrDA protocols formed the basis for much of what we have in Bluetooth today, even though Bluetooth is based on radio. I thought IrDA was terrific, and I worked on a lot of IR projects in the 1990s.

Interestingly, the IrDA folks remain quite active and are still cooking up innovations. The low cost/high throughput/unlicensed nature of IR means that it will never be completely replaced. Perhaps all IR really needs is better marketing.

I'm reminded of the great job the pork producers did by calling their product "the other white meat." Maybe a similar campaign about "the other wireless technology" might do some good. Regardless, we've not heard the last about IR -- stay tuned.