Can We Find (and Use) "Spectrum Holes"? Spectrum Sensing and Spatial Reuse Opportunities in "Cognitive" Radio Systems

A main concern in spectrum overlays ("Cognitive Radio") has been the reliability of sensing techniques to predict the performance in the primary communication link if the secondary user decides to use the spectrum. The fundamental difficulty is that sensors detect the transmissions (i.e. the transmitters), whereas the communication performance is determined by the interference environment around the (possibly silent) receivers. In previous work, specific scenarios with rather simple propagation models, mostly neglecting the correlation properties between signal paths, have been used. In this paper, more realistic propagation models are used to compare the performance of various sensing and resource management strategies. Results show that simple transmitter based sensing works well only in very flat terrains and at low frequencies. In these scenarios unfortunately, the opportunities for spectrum reuse are limited. In more rough terrains, where the potential for secondary use of the spectrum is large, simple sensing schemes on the other hand perform poorly. Even advanced sensing schemes and/or responsive primary system may not be sufficient to achieve acceptable performance.