Real-coded genetic algorithm parameter setting for cognitive radio adaptation

Cognitive radio (CR) is an emerging promising technology for future wireless communication networks. For a given status of dynamic wireless channel environment, the radio uses intelligence to optimize user's QoS by adapting the transmission parameters of the radio. Transmission parameters adaptation in a dynamic multicarrier environment has been previously studied using genetic algorithms (GA). The main goal was to select the optimal transmission parameters. When genetic algorithms are applied to complex problems, such as in multicarrier (MC) one, their execution time, to find a solution, increases. In this paper, we propose a new algorithm based on a real-coded genetic algorithm. Several tests have been done to tune the different parameters of the algorithm. The new algorithm is compared to the standard GA, based on De Jong's standard parameter settings and binary coding, used for optimizing QoS for cognitive radio. The tests carried on show that our algorithm gives better performance in terms of convergence speed.