Statistical characterization of transmitter locations based on signal strength measurements

In this paper we study the problem of characterizing transmitter locations based on signal strength measurements. Instead of attempting localization of individual transmitters, we adopt a model-based approach, attempting to infer either a model or selected key statistics describing how transmitters are distributed over the region of interest. We show that such estimates can be made with much smaller number of measurements and with higher degree of accuracy than would be required for solving the full localization problem. We study two scenarios distinguished by assumptions made on the capabilities of node(s) carrying out the measurements. In the first scenario only capability of measuring total received power is assumed, whereas in the second scenario we consider receivers capable of distinguishing between transmissions from different sources. We also discuss in some detail the potential applications of the developed techniques, especially focussing on radio resource management problems in cognitive wireless networks.