Optimal Consensus Based Decentralized Cooperative Spectrum Sensing in Cognitive Radio Networks

We study the problem of decentralized cooperative spectrum sensing in cognitive radio ad hoc networks without a fusion center, using a consensus based solution. First, the received logarithmic likelihood ratio (LLR) is employed as the local detection Outcome; then, it is exchanged between neighboring CR users and is updated based on the LLR received from the neighbors. It is proved that with the proposed decentralized algorithm, the optimal sensing performance is achieved via just local information exchange between neighbors. In addition, it is shown that the proposed algorithm outperforms previous work in which the received energy was employed as the local detection outcome.