Passive congregation-based particle swarm optimisation for designing quadrature mirror filter bank

Abstract

In this paper, we propose two different approaches based on Particle Swarm Optimisation (PSO) for the design of Near Perfect Reconstruction (NPR) Quadrature Mirror Filter (QMF) bank. The proposed method employs Constriction Factor-based PSO (CPSO) and PSO with Passive Congregation (PSOPC) approach for the design of prototype filter. Both the PSO methods are presented to optimise FIR filter coefficients that lead to NPR filter bank. Passband ripple, stopband energy, transition-band error and amplitude distortion of the overall bank are combined to obtain a closed form objective function. Two illustrative examples are given to demonstrate the feasibility of our proposed scheme and performances are compared with other existing methods. The experimental results show that PSOPC algorithm provides better performance parameters with great accuracy and convergence.