Simulation results show that the proposed scheme outperforms others both in performance and complexity, and is very suitable for the implementation of high speed decoders.