Optimal wideband spectrum sensing order based on decision-making tree in cognitive radio

In the cognitive radio networks, the cognitive user must scan the multi-channel to acquire an idle channel before its cognitive transmission. If the sensing capability is limited, the user must sense the multi-channel one by one. So the sensing order is very crucial to the realization of maximum throughput. In this paper, we aim to find the optimal wideband sensing order in polynomial time which is much faster than brute-force search and dynamic programming approach. First, we present two greed solutions, with the trait: fast but not optimal. Also, we analyze and improve the potential function in the optimality principle of sequencing. Then, we propose the decision-making tree method with the proper branching rule and the lopping rule, to search the optimal wideband spectrum sensing order. In the end, through the performance analysis, we validate the low computational complexity and the optimality.