Abstract
Cognitive radio is an effective technique to enhance the spectrum utilization. In cognitive radio systems, spectrum sensing is the most important part. However, a single secondary user cannot exactly detect the existence of the primary user due to the effects of hidden nodes, shadowing, and fading channels. In order to combat these effects, cooperative spectrum sensing has been proposed. For the conventional cooperative spectrum sensing method, the control channel bandwidth required for reporting the sensing results of secondary users would be significant as the number of secondary users is large. It is also not a good approach to use equal weighting for all secondary users’ decisions in cooperative spectrum sensing when they are under different sensing environments. In this thesis, we propose a weighting scheme based on the water-filling principle and apply it to cooperative spectrum sensing with double threshold energy detection for cognitive radio systems. For the proposed cooperative spectrum sensing scheme, each secondary user is allocated a weighting factor based on the instantaneous signal-to-noise ratio (SNR) of the corresponding channel, where the weighting factor is set to be zero as the instantaneous SNR is low enough. Since the zero-weighting secondary user is ineffective in the sensing process, it does not need to report the local decision to the secondary-user base station. To assess the detection performance of the proposed scheme as well as to determine the threshold for energy combination, we also provide analyses of the false alarm probability and the normalized average number of sensing bits under Rayleigh fading channels. Simulation results show that, under the same false alarm probability, the proposed weighting scheme achieves better detection performance with a smaller normalized average number of sensing bits (i.e., a smaller control channel bandwidth) than the equal weighting scheme, and they agree well with the analysis results.