On the effectiveness of using Genetic Algorithm for spectrum allocation in cognitive radio networks

In this paper, an extensive study is performed to understand the effect of the various parameters used in Genetic Algorithms (GA) to solve the spectrum allocation problem in cognitive radio networks. Two utilization functions, namely, Mean-Reward, and Max-Proportional-Fair, are used to evaluate the performance of the GA under different parameters values. Extensive simulation shows that, the parameters of the GA can be fine-tuned to achieve up to 90% and 62% improvements in speed and error, respectively, as compared to the GA used in the literature.