Linear iterative turbo-equalization (LITE) for dual channels

We examine a point-to-point communications scenario in which two or more separate, but known, channels are available for data transmission. While sending the same data across multiple channels provides channel diversity, we introduce additional temporal diversity by permuting the order of the data prior to transmission over one or more of the channels. As a receiver, we introduce a low complexity iterative equalization algorithm, inspired by iterative decoders for turbo-codes, which we call linear iterative turbo-equalization (LITE). The LITE algorithm contains one minimum mean square error linear equalizer for each channel and passes soft-information between the different equalizers in the form of a prior over the transmitted data. The linear equalizers differ from conventional equalizers by incorporating this prior in the minimization. Through simulations, we compare the empirical performance of the LITE algorithm to that of conventional linear and decision feedback equalizers, as well as maximum likelihood decoding for the set of channels. Our simulations demonstrate that the LITE algorithm can achieve equalization performance comparable to maximum likelihood decoding with computational complexity comparable to that of linear equalization.