Novel design of FIR digital filters with arbitrary magnitude and phase specifications based on Chaotic Imperialist Competitive Algorithm

An efficient nonlinear global optimization approach is utilized to design digital finite-impulse response filters with arbitrary complex frequency responses in the least-squares sense. The adopted numerical optimization technique is based upon the well-known Imperialist Competitive Algorithm (ICA). Hear we enhanced the performance of the ICA with use of improved chaos iteration circle map. This implementation known as Chaotic ICA (CICA). Several examples and comparisons to the existing methods are presented to illustrate the effectiveness and flexibility of CICA method.