Post sensing optimal channel selection in cognitive radio networks

In cognitive radio networks parallel sensing the spectrum is not possible due to hardware limitations. This poses a challenge for good sequential sensing algorithms. We present a simple channel selection method for selecting the best channel amongst the sequentially scanned channels for the secondary user transmission taking the cost of scan per channel into consideration. We model the problem in the framework of the optimal stopping problem. Under practical assumptions, we show that this problem is a monotone problem and that the optimal policy is completely characterized by the one stage look ahead or 1-sla set. We impose conditions on the states and analytically characterize the 1-sla set. We show that the policy from the finite horizon MDP and the analysis of the 1-sla set yield the same result for different distributions on the channel state.