Fast Converging Cuckoo Search Algorithm to design symmetric FIR filters

Abstract

Elimination of noise from transmitted signals inevitably incorporated during transmission persisted important task for the researchers from the preliminary days of Signal Processing. Among different techniques proposed regarding this purpose, use of digital filters has become most effectual in multiple ways. Feasible design, lower hardware costs have made the Finite Impulse Response (FIR) filters popular. Among different techniques, using perfect set of filter coefficients to implement FIR filters is most acceptable. Process of determining appropriate set of filter coefficients is characterized as an optimization problem with the objective of minimizing error function. Error function is conceptualized as an approximation function signifying deviancy between the designed filter responses from the ideal filter responses. Present work proposes Fast Converging Cuckoo Search Algorithm to obtain optimized filter coefficients to implement lowpass, highpass and bandpass FIR filters. Responses of the implemented filters using the proposed method are compared with the responses obtained by the filters designed using conventional Cuckoo Search Algorithm and Parks McClellan algorithm.