Parallel cooperative sensing in multichannel cognitive radio networks with partial CSI

In this paper, to maximize the total throughput of secondary users (SUs), we investigate the problem that how to allocate SUs cooperatively parallel sensing multiple primary users (PUs) channels, how to select the best SUs making full use of the available channels, and how to determine the sensing time and detection threshold for each PU channel. To efficient protect PUs from harmful interference, the problem is optimized under the average interference power constraint. Thus, we firstly present the exhaustive search algorithm for the problem with high complexity and partial channel state information (CSI). Subsequently, to decrease the complexity and solve the optimization problem in practice, we propose a detection threshold maximization algorithm (DTMA) to maximize the total SUs´ throughput by maintaining the detection threshold as large as possible. The algorithm determines the sensing group, sensing time, and detection threshold jointly. Simulations are conducted to verify our analysis, and show the DTMA can get the similar performance and variation as the exhaustive algorithm with low complexity.