Theoretical Limits and Practical Detection Schemes for Channels Affected by Class-A Impulse Noise

Abstract

We consider digital communications over channels affected by Class-A impulse noise. First, by means of information theoretic arguments, the performance limits of such systems are derived and discussed. Then, practical communication schemes employing powerful channel codes with iterative decoding are presented. Together with an ideal detection scheme which perfectly knows the statistics of the impulse noise, we describe a sub- optimal scheme which ensures a limited performance degradation even when a reliable statistical characterization of the noise is not available. Finally, by means of computer simulations, we show the effectiveness of the described schemes, whose performance is fairly close to the derived theoretical limits.