A Practical Multibit Data Combining Strategy for Cooperative Spectrum Sensing

In this paper, we study a cooperative spectrum sensing scheme for cognitive radio systems where each sensor transmits multibit quantized information to a fusion center where the decision about the availability or occupancy of the channel is made. In particular, we introduce a linear-quantization scheme that is based on a single parameter Δ and derive the optimal detector at the fusion center for this scheme. We also derive the performance of this detector as a function of Δ and use it as a cost function in an optimization problem to find the Δ that provides minimum error. Furthermore, we propose a suboptimal detector with much lower complexity and compare its performance with the optimal detector. Finally, we compare the performance of our multibit combining scheme with the hard and soft combining schemes and show that, with transmission of a few bits of information from each sensor, the system can achieve an error rate very close to the optimal soft combining scheme.