Spectrum sensing in cognitive radio using recurrence diagrams

Abstract

In this article, a new technique for the detection of primary users in cognitive radio (CR) is presented. The proposed technique is based on statistical quantifiers that take into account causal information from the received signal. Results show that a quick decision about the transmission of the data is possible by sensing a very short time. Furthermore, heavily noise-contaminated signals are detectable, even without prior information about the level of noise in the channel. We show the performance of the method and conclude that is comparable to those used classically. Finally, we present the hardware implementation and the resources employed.