Reliability-based decision fusion scheme for cooperative spectrum sensing

In this study, the authors propose a reliability-based cooperative decision fusion scheme which considers the reliability of the secondary users (SUs') local decisions when making a final decision at the fusion centre in cognitive radios. The authors use past information about the local and global decisions to estimate the reliability of the sensing decision obtained from each SU and then reflect this difference in reliability in the weighting of each SU's decision. The authors formulate the problem of minimising the probability of sensing error at the fusion centre, subject to a limit on the network probability of detection, as a constrained non-linear integer programming problem. To solve this problem, the authors implement an iterative solution based on the generalised non-linear Lagrangian relaxation. Simulation results show that our proposed solution can achieve optimal results with zero duality gap using only a few number of iterations. Results also demonstrate that the proposed reliability-based fusion scheme provides performance improvement, in terms of the minimum probability of sensing error, when compared to the OR and AND fusion schemes. This improvement is more pronounced as the number of users increases since by assigning weights differently to users, the multiuser diversity gain is better exploited.