Asexual Reproduction-based Adaptive Quantum Particle Swarm Optimization algorithm for dual-channel speech enhancement

In this paper, we propose an improved particle swarm optimization algorithm, called Asexual Reproduction-based Adaptive Quantum Particle Swarm Optimization (ARAQPSO), for dual-channel speech enhancement. The foundation of a particle optimization algorithm is to intelligently generate and modify the initial randomized solutions. The proposed algorithm is based on Adaptive Quantum Particle Swarm Optimization (AQPSO) technique. Particles that search the problem space have the ability to reproduce asexually, where the fertility of particles is proportional to their fitness. The proposed algorithm applies an adaptive local search around the fitter particles that result in a comprehensive search in prosperous regions of the problem space. Experimental results indicate that the algorithm outperforms AQPSO, SPSO, and the gradient-based NLMS algorithm in the sense of SNR-improvement.