Optimization of Matched and Mismatched Filters in Short Range Pulse Radars using Genetic Algorithm

Matched and mismatched filters are considered important parts of a radar signal processing unit. In this paper, we present an approach to optimize the matched filters and mismatched filters in short range pulse radars. For radar, the matched filter coefficients are the complex conjugates of transmitted code. We used binary phase codes as transmitted pulse. The disadvantage of binary phase codes is having high sidelobe levels in the output of correlation function. Thus, we decided to use optimization algorithms for finding binary phase codes with minimum peak sidelobe levels (MPS). After that, we succeeded in producing mismatched filter coefficients (Mis-co) for each code using floating point genetic algorithm (FGA) and we could generate and test the filter coefficients with maximum peak to sidelobe level ratio (PSR). For testing the filter, we plotted ambiguity function for each set of coefficients and tested the filter with Doppler shift.