Evolutionary design of analog FIR filters with variable time delays for optically controlled microwave signal processors

This paper presents the application of genetic algorithms to the design of analog FIR filters with variable time delays — specific examples for tunable optically controlled microwave signal processors. Besides the FIR filter coefficients as the standard design parameters, the time delays can additionally be optimized. Because of physical constraints, specific restrictions of the design parameters, however, have to be obeyed. In order to make use of the additional freedom of optimizing the time delays and to observe the design restrictions, non-standard design techniques arc needed. To this end, this paper studies the applicability of genetic algorithms to analog FIR filter design. Experimental results and a comparison to a standard design algorithm arc given that demonstrate the excellent properties of the proposed evolutionary design technique.