Improved Meta-Heuristic Technique to FIR Filter Design and Application

The main aim of signal separation with the FIR function is considerably extended by the requirements of communication networks to improve the transfer of information from transmitter to the receiver without increasing hardware difficulty. To achieve this, there is a design requirement for a certain type of digital filter with specific frequency response characteristics concurrently to develop the efficient hardware techniques is also essential. The advantage of FRM based FIR filter is that its structure comprises of sub filters that has a smaller number of non-zero coefficient compared to a minimax optimum design. This paper work envisage an integrated optimization approach known as Brain Storm-Grey Wolf Optimizer (BSGWO) algorithm for the development of Frequency Response Masking (FRM) based design of FIR filter. In GWO, three wolves are named alpha, beta and delta respectively based on their social ranks. Among three wolves, alpha is placed in first position in social ranking. BSGWO make use of selected alpha wolf in the mutation or new individual generation process of BSO to speed up the convergence of optimization algorithm. By leveraging the characteristics of the highest ranked alpha, BSGWO aims to guide the search towards promising regions of the solution space, potentially enhancing the efficiency and effectiveness of the optimization process. The least square algorithm is used initially to find the filter coefficients of the original filters and the proposed algorithm is used to obtain optimal filter coefficients. In both the cases, fitness, the number of components, and the Mean Absolute Error (MAE) will be measured and analysed to prove the filter performances. This proposed algorithm offers a minimum fitness with improved convergence speed, and smallest MAE of 0.05, 0.0155, respectively in comparison with the existing heuristic algorithms.