Optimization of FIR filter parameters using forest optimization algorithm.

This paper demonstrates the usage of forest optimization technique for the determination of optimal parameters for finite impulse response (FIR) filter. Inputs are selected as design specifications; an attempt to apply the forest optimization-based algorithm for the complex nonlinear, constrained optimization task of designing the filter has been made. Optimization of filter parameters is done by exploiting the natural evolution of trees in a forest. Local seeding is employed for local searching and global seeding concept ensures that only global solution is reached without falling into local optimum. Unlike the conventional windowing techniques generally employed, proposed technique realizes the ideal frequency characteristics meeting the specified targets related to pass band filter and stop band. Simulation results are presented. [ABSTRACT FROM AUTHOR]