Cooperative Spectrum Sensing with a Progressive MAP Detection Algorithm

In this paper, a new cooperative spectrum sensing algorithm is proposed for a cognitive radio network with multiple secondary users (SUs) sharing spectrum with one or more primary users (PUs). Unlike most previous spectrum sensing algorithms that do not consider the time domain traffic statistics of the PU, the algorithm in this paper is developed by exploiting the statistical properties of the PU´s transmission pattern, which is modeled with a Markov chain with two states: busy (1) and idle (0). Each SU performs energy detection based on an observation of the Markov chain, and the detection results are forwarded to a fusion center (FC) through a noisy channel. The FC recovers the decisions of the SUs by using a new progressive maximum a posteriori (MAP) algorithm, where the a priori probability essential to the MAP detection is obtained by progressively estimating the transition probabilities of the Markov chain. Analytical expressions are derived for the probabilities of false alarm and missing detection, with both the majority data fusion rule and the OR data fusion rule. Both theoretical analysis and simulation results indicate that the proposed algorithm can provide reliable and efficient spectrum sensing over a large range of system configurations.