Compact microstrip dual-band bandpass filters design using genetic-algorithm techniques

Abstract

An optimization scheme based on hybrid-coded genetic-algorithm (GA) techniques is presented to design compact dual-band bandpass filters with microstrip lines. A representation scheme is proposed to represent an arbitrary microstrip circuit as a set of data structures. Each data structure in the set describes a simple two-port network with the corresponding connection method and electrical parameters. The optimization algorithm based on conventional GAs is then applied to simultaneously search for the appropriate circuit topology and the corresponding electrical parameters with dual-band characteristic. Two examples are designed and implemented to validate the proposed algorithm. In the first example, the 3-dB fractional bandwidth of the low and high bands is 35% and 17%, respectively. It has return losses larger than 10 dB from 2.14 to 2.96 and 5.14 to 6.06 GHz. In the second example, the 3-dB fractional bandwidth of the low and high bands is 9.9% and 7.9%, respectively. The return losses are larger than 10 dB from 3.37 to 3.64 and 5.27 to 5.62 GHz. The sizes of the proposed filters are nearly half as small as those of the filters presented before. All the studies are completed on a computer with a 2.4-GHz microprocessor, and the computing time of two examples is 6 and 3 min, respectively.