Cognitive decision engine design for cognitive radio networks using gravitational search algorithm and flower pollination algorithm

In cognitive radio networks, the real-time setting of the transmission parameters required by the cognitive engine according to the quality of service requested by the cognitive users has become an essential task. This adjustment is becoming increasingly difficult in OFDM-based cognitive radio networks because of the existence of a large number of decision variables to be optimised for multi-carrier systems. For decision making, the cognitive engine in OFDM-based cognitive radio networks uses optimisation algorithms. However, in order to reduce the complexity and obtain a resource allocation in a reasonable time, cognitive radio networks use artificial intelligence techniques and in particular metaheuristics. We analyse in this article the performances of two recent metaheuristics namely gravitational search algorithm and flower pollination algorithm in OFDM-based cognitive radio networks. Simulation results show that FPA has surpassed GSA in terms of Fitness. In contrast, GSA outperformed FPA in terms of execution time. On the other side, FPA and GSA outperform genetic algorithms in terms of solution quality (fitness) with improvements reaching 10% and 7% respectively and prove their efficiencies in order to support three modes of transmission of the cognitive user.