Detection of a Spectrum at the Receiver in a CR Network

Abstract

Cognitive radio is being touted as the next Big Bang in wireless communication. A large part of the usability of cognitive radio rests upon its ability to detect white spaces in the spectrum, i.e. to detect the presence of the primary or licensed user. As such, spectrum sensing occupies most of the research potential as far as cognitive radio is concerned, and in this paper, a spectrum sensing technique will be looked upon. A optimal detector for spectrum sensing can be reduced to other well-known detectors, three signal processing techniques, viz. matched-filter, energy detection and cyclostationarity detection, is considered. Of special interest is the topic of cooperative sensing, where a number of cognitive nodes can interact with each other or a master node to efficiently carry out the spectrum sensing, often more effectively. We look at a technique, varying from narrowband to wideband detection, with special emphasis on optimal data and decision fusion. Finally, we consider one detection technique, which can be applied when the frequency information about the channel or the primary signal is available.