Performance analysis of coalition formation algorithms based on matching theory for cognitive radio networks

We consider the problem of increasing the throughput in cognitive radio networks by forming coalitions among cognitive radio user in additive white Gaussian noise (AWGN) channel. For coalition formation using matching theory, we analyze two algorithms, namely Gale-Shapely algorithm and one-sided stable matching algorithm. For the first algorithm for coalition formation, well-known gale shapely algorithm is used to achieve cooperation among the cognitive radios for spectrum detection and sharing. Each cognitive radio prepares a preference list of other radios in the vicinity for cooperation and hence to form a coalition formation. The second algorithm is based one-sided matching theory which is a variant of the Gale-Shapely algorithm, however, to achieve a stable cooperation, certain criteria must be satisfied. The procedure is similar to the first algorithm (.i.e. formation of preference list and then making offers to other cognitive radio for cooperation) however the difference is in how the coalition formation takes place among the cognitive radios. Finally, using simulations we investigate various aspects of the algorithms and analyse their performance. The proposed algorithms result in improved spectrum detection as well as increasing the spectrum efficiency.