Theoretical Limits and Practical Detection Schemes for Markovian-Gaussian Channels

We consider digital communications over channels impaired by impulse noise. A two-state Markov channel model is adopted, which, unlike the memoryless models generally considered in the literature, allows to describe the typical bursty nature of the impulse noise. First, by means of information-theoretical arguments, we compute and discuss the ultimate performance limits of such systems. Then, we derive an algorithm for optimal maximum-a-posteriori symbol detection, and explain how to exploit it in receivers employing iterative detection/decoding. Finally, we report simulation results showing the effectiveness of the proposed schemes, as well as an excellent agreement with the computed theoretical limits.