Suboptimum MLSE for severe echos using metric forecast

Abstract

Because of its complexity, the Viterbi algorithm is unsuitable for severe intersymbol interference (ISI) in the presence of extreme echos. Due to erroneous path discarding suboptimum breadth-first algorithms like the M-algorithm fail at poor signal-to-noise ratio because all youngest path symbols are weakly valued by the slight amount of energy at the beginning of the channel impulse response. The method of metric forecast estimates the metric which a path will achieve, when the current youngest symbol arrives at a main energy section of the impulse response and thus reduces the problem in path selection. Simulation results show that a simple decision feedback forecast algorithm is sufficient.<>