Decoding Metric Study for Turbo Codes in Very Impulsive Environment

In this paper, we study the decoding metric of turbo codes in very impulsive environments. Channel impairments, for example in ad hoc networks, contain not only the thermal noise, but also the network interference generated by other nodes, which exhibits an impulsive nature. We use stable distributions to model this interference due to their heavy tail property. In order to adapt to this type of interference, we propose the p-norm as a decoding metric to replace the conventional Euclidean distance. Simulation results show that the p-norm allows a significant error correction improvement compared with other metrics in very impulsive cases, and its performance appears to be close to the optimal decoder.