An Efficient QMF Bank Design Using Differential Search Algorithm

To design a perfect reconstruction, quadrature mirror filter is an area of concern because of its usefulness in several digital signal processing applications. Their performance is measured in terms of ripple measure (amplitude distortion), stop-band error, and pass-band error for perfect reconstruction. In this manuscript, perfect reconstruction is obtained by optimizing the objective function, which is a weighted sum of mean-square pass-band error, mean-square stop-band error, and ripple measure. Using a population-based meta-heuristics differential search optimization algorithm, the objective function is optimized to obtain the optimal filter coefficients for the filter bank’s low pass prototype filter. The suggested algorithm results have compared to previously report evolutionary optimization algorithms. The performance attributes have been significantly improved over previously published filter bank design methodologies.