Cooperative Spectrum Sensing and Sharing in PD-NOMA Cognitive Radio Networks

Spectrum efficiency, spectrum sensing (SS) and sharing are core requirements of a reliable communication network to maximize the utilization of the scarce radio frequency spectrum. This research advances the development of efficient spectrum sharing techniques in CRN by integrating cooperative spectrum sensing and power domain non-orthogonal multiple access (PD-NOMA). As wireless systems are expected to utilize the spectrum resources efficiently, they must be able to mitigate interference with primary users and operate within the required throughput levels. Existing works consider either spectrum sensing or spectrum sharing separately, and often ignore the effect of cooperation and hybrid spectrum sharing on the overall system performance. This work presents and evaluates a novel framework that combines cooperative spectrum sensing and a hybrid interweave-underlay spectrum sharing mode for application in IoT-CRN. The network features multiple cognitive users (CUs) switching between interweave and underlay modes depending on availability of spectrum holes. This hybrid mode combines the advantages of opportunistically accessing the channel and transmitting with controlled power and interference. The model features a Gaussian mixture model (GMM)-based channel estimator sensing mechanism that enables individual CUs to determine the availability of holes. The formulated channel occupancy problem is evaluated using adaptive particle swarm optimization (PSO). The occupancy is cooperatively shared to the fusion center (FC) for the final decision on the availability of PUs with penalties applied to misbehaving CUs. The results of the proposed model show a 28% and 35% better throughput performance than the existing models and a 15% higher probability of detection while accommodating different CU traffic.