Date of Award
Electrical and Computer Engineering
Clustering approach is considered a management technology that arranged the distributed cognitive radio users into logical groups to improve the sensing performance of the network. A lot of works in this area showed that cluster-based spectrum sensing (CBSS) technique efficiently tackled the trade-off between performance and overhead issue. By employing the tree structure of the cluster, a multilevel hierarchical cluster-based spectrum sensing (MH-CBSS) algorithm was proposed to compromise between the gained performance and incurred overhead. However, the MH-CBSS iterative algorithm incurs high computational requirements. In this thesis, an energy-efficient low computational hierarchical cluster-based algorithm is proposed which reduces the incurred computational burden. This is achieved by predetermining the number of cognitive radios (CRs) in the cluster, which provides an advantage of reducing the number of iterations of the MH-CBSS algorithm. Furthermore, for a comprehensive study, the modified algorithm is investigated over both Rayleigh and Nakagami fading channels. Simulation results show that the detection performance of the modified algorithm outperforms the MH-CBSS algorithm over Rayleigh and Nakagami fading channels. In addition, a conventional energy detection algorithm is a fixed threshold based algorithm. Therefore, the threshold should be selected properly since it significantly affects the sensing performance of energy detector. For this reason, an energy-efficient hierarchical cluster-based cooperative spectrum sensing algorithm with an adaptive threshold is proposed which enables the CR dynamically adapts its threshold to achieve the minimum total cluster error. Besides, the optimal threshold level for minimizing the overall cluster detection error rate is numerically determined. The detection performance of the proposed algorithm is presented and evaluated through simulation results.
Alqawasmeh, Abeer Fuad, "Low Complexity Energy-Efficient Collaborative Spectrum Sensing for Cognitive Radio Networks" (2017). Electronic Theses and Dissertations. 7232.