In this paper is carried out an investigation that the performance of the cooperative spectrum sensing with relay may be interfered by redundant relays, the number of busy cognitive users (CUs) and the secondary relay(SR), detection threshold and transmission error of the channel, so that a global adaptive optimization algorithm is proposed. In this algorithm, based on the maximum interference-free power relay selection (MIFPRS), the set of secondary relay is defined adaptively; and in order to reach a maximum of the probability of detection, the single cognitive user minimizes the transmission error of the channel and selects the relay from the set of secondary relays. To achieve the maximum of the total channel throughput under the given probability of detection, ale ues the entire adaptive optimization algorithm. Numerical simulations demonstrate that the proposed algorithm has a high degree of channel transmission precision, and a large amount of channel throughput, so as to spend less frequency band.