Joint estimation based spectrum sensing for cognitive radios in time-variant fading channels

The traditional spectrum sensing schemes can only utilize the statistical probability of fading channels, which may fail to deal with the time-varying fading gain. Thus, the performance of such sensing techniques will degrade dramatically and may even become inapplicable to distributed cognitive radio networks. In this investigation, we develop a promising spectrum sensing algorithm for time-variant flat-fading (TVFF) channels. Firstly, a promising dynamic state-space model (DSM) is established to thoroughly characterize the evolution behaviors of both primary user (PU) state and fading channels, by utilizing a two-state Markov process and the finite-states Markov chain (FSMC), respectively. Relying on an optimal Bayesian inference framework, the sequential importance sampling based particle filtering is then suggested to recursively estimate PUs state and fading gain jointly. Experimental simulations demonstrated that the new scheme can significantly improve the sensing performance in TVFF channels, which provides particular promise to realistic applications.