Robust turbo decoding in single-carrier systems over memoryless impulse noise channels

Abstract

Communication systems subjected to strong impulse noise are prone to performance degradation when the impulse occurrence is neglected in the decoding process: turbo decoders are likely to exhibit error propagation because the decision-making is dictated by excessive samples corrupted by impulses when the conventional decision metric, which is based on the assumption of additive white Gaussian noise, is used. A sophisticated decoder is widely known to yield a substantial performance improvement by exploiting impulse statistics; however, impulse statistics are difficult to model accurately and are generally not time invariant, likely leading the receiver complexity to be prohibitive. Based on our previous work - which proved the robustness of a soft limiter despite a lack of impulse statistics, this study incorporated a clipping operation into a single-carrier turbo-coded system in memoryless impulse noise channels. Further, the performance of the proposed decoder was evaluated according to the bit error rate to demonstrate the performance gain derived from various clipping thresholds, each of which reflects a different degree of impulse information to which the proposed decoder can leverage.