SOVA decoding in Symmetric Alpha-Stable Noise

Soft-Output Viterbi Algorithm (SOVA) is one type of recovery memory-less Markov Chain and is used widely to decode convolutional codes. Fundamentally, conventional SOVA is designed on the basis of Maximum A-Posteriori Probability (APP) with the assumption of normal distribution. Therefore, conventional SOVA fails miserably in the presence of symmetric alpha stable noise S S α which is one form of stable random processes widely accepted for impulsive noise modeling. The author studies and has improved the performance of conventional SOVA by introducing Cauchy function into path-metric calculation. Substantial performance improvement was gained from Mento Carlo Simulation for SOVA based turbo codes. Keywords-component: turbo codes, Soft Output Viterbi Algorithm (SOVA), Maximum A-Posteriori (MAP) Estimation, Cauchy metric, Symmetric Alpha Stable ( S S α ) Distribution, non-gaussian CHANNEL.