Optimization of Spectrum Sensing Period in Cognitive Radio Networks Based on GA
Yang Liu, Ying Cui, Ou Li · Advanced materials research · 2013
In Cognitive Radio Networks (CRN), an important issue associated with MAC-layer sensing is how often the availability of licensed channels should be sensed. To solve this issue in multi-channel environment, we address how to maximize the discovery of spectrum opportunities by setting a set of reasonable sensing-periods. Based on alternating renewal theory we develop an channel-usage pattern estimation method, and derived the states transition matrix by discrete-time Markov process. At last, we proposed an objective function of spectrum sensing efficiency, and solved the multi-objective optimization problem by improved Genetic Algorithms (GA). Our simulation results demonstrate the efficiency of the objective function and the significant performance of the improved GA. The sensing-period vector we derived discovers more than 80 percent of the analytical maximum of discoverable spectrum-opportunities in six-channel environment, which is shown to discover up to 30% more than the method that sensing each channel with the same period.