Distributed Spectrum Detection Algorithms for Cognitive Radio

Cognitive radio technology may offer the capability for suitably enabled terminals to sense spectrum activity and reuse free channels, thus allowing an increase in efficiency by permitting dynamic sharing of spectrum between users. Accurate spectrum sensing is a vital requirement. This work presents simulation of an algorithm that improves the accuracy of spectrum sensing by combining results from teams of co-operating sensing terminals, thus overcoming problems caused by localised shadowing. Results show that the detection algorithm performance is greatly improved by allowing neighbouring nodes to influence the channel sensing process at a particular cognitive radio location, and that a detection accuracy approaching 100% can be achieved in certain cases. Various other trade-offs exist in the weighting factors used to tune the algorithm performance.