Towards Sensing Performance and Energy Efficiency Trade-off of Spectrum Sensing in Cognitive Radio Networks

Different priorities of users’ access are defined in cognitive radio networks (CRNs) in order to differentiate the spectrum usage: primary user (PU) preferentially occupies the band, whereas secondary users (SUs) can only have access to the band if PU is available. Firstly, PU activities are optimized to improve the energy efficiency (EE). Too much presence of PU could leave less opportunities for SU to access and transmit data; too much absence of PU could influence the received signal to noise ratio of SUs and therefore increase the false alarm probabilities. Hence, in this thesis, the tradeoff problem of PU’s activities is considered in the coordinated PU-SUs formulation of EE. Using bi-section search method, PU’s active ratio is optimized with sensing time ratio of a SUs’ frame to always achieve the maximal EE of the whole network. Secondly, a mixed utility function is formulated as “throughput-energy cost” to determine whether it is worthy of transferring energy into additional throughput, in order to dynamically control SUs’ access attempts using an evolutionary game in cooperative sensing model. A re-activation scheme is especially presented to conquer the great fallings of throughput and sensing performance at low sensing time and to make SUs more active on the transmissions. Simulation results show that our methods can better manage the resources in CRNs, EE and sensing performance are improved between PU and SUs. The observation of evolutionary equilibrium also shows that a balance can be achieved by adapting SUs’ access behaviors in a cooperative way.