Optimization of control parameter of Differential Evolution algorithm for efficient design of FIR filter

Modern signal processing applications invoke various evolutionary algorithms in different areas such as aerodynamic shape optimization, pattern recognition, digital filter design, automated mirror design etc. Presently, Differential Evolution (DE) algorithm has proved to be quite efficient in these areas. Digital filters of various kinds can be designed using this particular evolutionary technique. In this paper, we have studied the effect of an important control parameter namely, the Weighting Factor, on the convergence speed of the DE algorithm, associated with the efficient design of low-pass FIR filter being used subsequently as a pulse-shaping filter in a QPSK modulated system. Finally, the optimized value of the Weighting Factor for this specific design problem has been suggested. Experimentally measured Eye diagrams confirm the optimized value.