The Polite Radio
By Mark Frauenfelder, Wed May 04 08:00:00 GMT 2005

Why are radios becoming more polite? Because mobile users are becoming more greedy. We want our mobile phone to make calls, connect to Wi-Fi and communicate with Bluetooth devices -- all at once.

The radio started out as a boor. The earliest transmitters sent signals over a wide range of frequencies, hogging huge slices of spectrum for themselves. It wasn't a problem in the beginning, because there weren't enough broadcasters around for there to be interference problems.

When more broadcasters arrived, interference did become a problem. The solution was modulation -- superimposing the original sound or data signal on a constant, high-frequency carrier signal. That way, signals could be restricted to a particular frequency.

As the years went on, radio became even more polite: it learned to divide channels into sequential time slots, and assign different devices to coded channels. Then, in 1991, a computer scientist named Joseph Mitola coined "software radio" to describe wireless devices that were capable of transceiving over different frequencies and protocols simply by changing the software that ran the devices. It was a good idea -- instead of having to uses multiple chipsets to be able operate in different networks, a software radio could emulate whatever chipset the system required to work with it.

In November 2004 the FCC gave its first blessing to a such a system -- Vanu's Anywave software radio GSM base station, which support multiple wireless services and standards entirely in software.

The Cognitive Radio

And in March of this year, the FCC released a Report and Order (R&O) on "cognitive radio" systems, another intriguing buzzword term coined by Mitola, who in 2003 described a cognitive radio to me as a software radio that additionally has "the etiquette to listen to what's going on to determine the nature of the noise or interference of the channels and to predict what kind of effect its transmission at a given power level is going to have on other users and to correct its behavior if it turns out that somebody else needs to use the channel."

The FCC's R&O was favorable for cognitive radio, concluding that cognitive radio architectures can make efficient and intelligent use of the radio spectrum. It stated the report is intended to "facilitate continued growth in the deployment of radio equipment employing cognitive radio technologies and make possible a full realization of their potential benefits."

The Cognizant Radio

One person who would agree with the FCC is Ali Niknejad, a professor of electrical engineering at the University of California Berkeley, and a member of the Berkeley Wireless Research Center. Niknejad is developing a system called the Cognizant Universal Radio (COGUR) that could bring cognitive radio closer to reality.

The overall goal of COGUR is reducing the chip count. While some handset manufacturers claim to make single-chip phones, in reality, they contain dozens or even hundreds of components in addition to the radio circuitry. Multi-band phones, for instance, require separate filters for each band. And they require switching circuitry to change bands. These additional components increase power requirements and complexity. COGUR is meant to address chip proliferation inherent in conventional multi-band design.

A major part of COGUR has already been designed and built by one of Niknejad's graduate students, Axel Berny. It's a highly tunable wideband voltage-controlled oscillator in the form of an integrated circuit. An oscillator is the part of a radio transmitter that generates the carrier signal upon which the voice or data signal rides. Conventional voltage-controlled oscillators (VCOs) have a tuning range in the neighborhood of 20 percent, but the COGUR VCO can double or halve its midpoint frequency. If the midpoint frequency is, say, 5 gigahertz, it can shift up to 10 gigahertz or down to 2.5 gigahertz. Better yet, Niknejad and Berny are working on a universal frequency synthesizer chip (of which the VCO will be a component) that will greatly increase the tuning range.

Power Play

COGUR will be able to vary power output and receiver sensitivity on the fly, as well. There's no need to pump out as many watts for a Bluetooth connection as there is for a voice call that requires the phone to send a signal to a cell tower. The cognitive radio will assess the given conditions, and, paying proper respect to both battery life and other devices in the general vicinity, make the necessary signal strength adjustments.

Another exciting thing about cognitive radios is that, at least in theory, they can happily coexist in legacy wireless environments. If the intelligence is kept at the edge of the network, then legacy cell tower systems don't need to be overhauled. The cognitive mobile device will do all the work. And when industry standards change, as they invariably do, chances are the flexibility of the software-driven COGUR will be able to make the transition to the new protocols with ease.

Niknejad and the Berkeley Wireless Research Center hope to be able to integrate all of COGUR's components onto a single chip within the next few years, after which time, it could be ready for commercial implementation.