Bit-error-rate-minimizing channel shortening using post-FEQ diversity combining and a genetic algorithm

Abstract

In cyclic prefixed wireless communication systems, the bit error rate (BER) is the primary design concern. However, most equalizers for these systems optimize proxy metrics that are easy to use but are suboptimal in terms of BER. In this work, a channel shortener and frequency domain equalizer (FEQ) are developed for cyclic prefixed systems, which directly attempt to minimize the BER. To do this, first we modify the FEQ structure of a single input multiple output (SIMO) system such that the diversity combining is after the FEQ (whereas usually it is before the FEQ); and extend adaptive equalization algorithms to the new structure. This greatly mitigates frequency nulls, improving the BER. Second, a Genetic Algorithm, which is an optimization method based on the principles of natural selection and genetics, is used to choose the channel shortener to directly minimize the BER. The new approaches lower the BER by 10 2 in simulations.