Joint estimating based location and state of mobile primary user in spectrum sensing

Spectrum sensing, as one of the most important aspects, plays a crucial role on mitigating interference of secondary users (SU) in cognitive radio. However, moving primary user (PU) will sharply decrease the stability of observable information by considerably deteriorating the sensing performance. In this paper, a new joint estimating scheme is proposed for tracking PU proactively and detecting the occupation of primary band meanwhile. In view of both PU's state and its location, the united mathematical model based dynamic state-space model (DSM) is established in the new scheme. On this basis, a Bernoulli filter algorithm is suggested to jointly estimate the PU's state and its location recursively, which is further implemented by Particle filtering. Furthermore, an adaptive horizon expanding method is subtly designed to deal with loss of tracking resulting from the intermittent disappearance of PU's state. Experimental simulations demonstrate that, aiming at mobile PU, sensing performance of the new scheme is apparently better than other traditional methods, and the estimated PU's location may be further utilized by resource allocation of cognitive network.