Subspace based multi-user spectrum sensing in frequency selective cognitive radio systems

This paper proposes an optimal subspace based detector for multiple primary user spectrum sensing in frequency selective Rayleigh fading wireless channels. The proposed sensing scheme employs the optimal zero-forcing algorithm to null the effect of inter block interference (IBI) occurring due to the multipath nature of the wireless channel and also the multi user interference (MUI) arising due to the presence of other interfering primary users. This is followed by deriving the optimal Neyman-Pearson (NP) criterion based subspace detector to efficiently sense the presence/absence of the desired user. To characterize the efficacy of the proposed detection framework, expressions for the probabilities of false alarm and detection are also derived. Finally, performance comparison with existing detectors is presented.