Improved Self-adaptive Differential Evolution Based Throughput Maximization of Energy Harvesting Cognitive Radio Network

Evolutionary computation is a popular optimization technique for its significant feature in achieving a global optimal solution. Optimization problem obeys different characteristics for different optimized variables. The proper choice of control parameters and mutation strategy largely affects the optimal result. To improve the convergence performance and its applicability in engineering problems, an Improved Self-Adaptive DE algorithm with discrete mutation control parameters (ISADE) is proposed. In ISADE, each variable of individual search agents has its own crossover and mutation control strategy. The performance of ISADE is tested on a set of Quadric and Griewank standard mathematical benchmark functions. Energy harvesting enabled CRN (EH-CRN) in centralized cooperation scheme is investigated and throughput maximization problem is formulated. The results of the proposed ISADE-based throughput maximization of EH-CRN are obtained and are compared with some of the well-known and popular state-of-the-art algorithms in the literature. The results showed significant improvement in terms of convergence characteristics and optimal results in comparison to most of the reported standard optimization algorithms.