Filter bank based wideband cognitive radio sensing using joint distribution of pseudo WIShart matrix Eigenvalues

Cognitive RAdio sensing Based on the joint distribution of pseudo WIShart matrix Eigenvalues (CRABWISE) is a narrowband multi-antenna sensing algorithm, which applies random matrix theory performing near to the optimum detector at the cost of extra complexity. In this paper, CRABWISE is modified such that it is applicable for wideband spectrum sensing in cognitive radio with a single antenna. To that end, a tree-structured filter bank is used for wideband analysis of the time-frequency components of a secondary user received signal. The sensing performance of the CRABWISE hypothesis test is characterized by using a synthesis filter bank for generating a test signal with a random occupancy over the considered subband. Optimizing the threshold of each subband is a fundamental task for achieving reliable sensing performance. The performance is assessed in terms of the average receiver operating characteristic and the probability of a wrong decision in each subband.