Spectrum allocation in cognitive radio networks using multi-objective differential evolution algorithm

In the existing literature the forced termination probability is analyzed after the completion of spectrum allocation (SA) process. Since the forced termination probability depends on the allocation results, it is necessary to take the termination probability into account during the allocation process. In this paper, a two dimensional Markov model is used for analyzing the spectrum access. The Markov process assumes the mean arrival time of primary and secondary users and calculates the forced termination probability. In the current work, the forced termination probability is considered as one objective function along with three network utility functions namely Max-Sum-Reward, Max-Min-Reward and Max-Proportional-Fair to improve the quality of service. Finally the spectrum allocation process is formulated as a multi-objective optimization problem consisting of the above mentioned four objective functions and solved by using multi-objective differential evolution (MODE) algorithm. The performance of MODE algorithm is compared with nondominated sorting genetic algorithm II (NSGA-II) for solving the SA problem. The simulation results show that MODE performs better compared to NSGA-II algorithm in terms of timing complexity and pareto optimal solutions.