Potential collapse of whitespaces and the prospect for a universal power rule

Abstract

The TV whitespaces have recently been opened up for semi-unlicensed use by frequency-agile radios. However, there is a potentially significant flaw in the adopted rules: they try to treat the whitespaces in a manner similar to the ISM bands — with per-device transmit-power constraints. Unfortunately, wireless interference aggregates and the population density across the United States of America varies by orders of magnitude. This means that the aggregate interference that TV receivers might face could increase as whitespace devices are deployed, and could collectively cause a loss of reception within the supposedly protected contours. However, it is not too late. The adopted geolocation plus databases approach lets us avoid this problem by changing database behavior — instead of just controlling where white-space devices operate, we should also hold their aggregate emissions to within a certain power density (i.e. by area). With the looming problem resolved, we can also try to address one of the main tensions within the entire TV whitespace approach: any set of allowed power/height/distance rules is implicitly prioritizing rural vs urban needs and picking favorites among different applications. Alas, the reality of aggregate interference prevents us from making everyone simultaneously perfectly happy. To enable higher transmit powers further from TV stations, we must necessarily reduce the allowed powers closer in. But amazingly, the properties of wireless propagation and information-theory combine to suggest that universally approximately-optimal approaches might be possible that could compromise between these competing interests in a principled way. We explore a pair of such rules and show that indeed, most people can get a data-rate close to what they would have gotten if the rules had been written especially for them.