Compressive spectrum sensing in the cognitive radio networks by exploiting the sparsity of active radios

Spectrum sensing is a key technology to detect spectrum holes in cognitive network. It has been demonstrated that collaboration among cognitive users can improve the probability of detecting the primary users, but the fusion center is the bottleneck when a lot of collaborative information is transmitted. In this paper, we consider the cognitive radio users only transmit part of sensing information to relieve the transmission load. Besides, the sensing information will be inevitably influenced by various noise in the process of transmission. Therefore, the challenge is how we can detect spectrum holes successfully from these incomplete and inexact measurements. Most recently, there are some research results on this but the detection performance is not satisfactory. In this paper, we firstly formulate the collaborative spectrum sensing as an optimization model and then present a novel adaptive orthogonal matching pursuit algorithm by exploiting the sparsity of active primary users. Statistical property of the sensing data plays a crucial role in spectrum sensing. Theoretical analysis shows the presented scheme can detect active primary users rapidly and efficiently. Simulation results verify that the proposed method can obtain better detection performance with stronger noise background, which is more attractive in real applications.