Reduction Factor Approach to Improve the Performance of Soft Output Viterbi Algorithm in AWGN Channel

Abstract

Turbo Codes are considered the paramount channel coding technique due to its significant performance with guaranteed Quality of Service (QOS) that is highly suitable for advanced wireless applications. Complexity of these codes, particularly their decoders are considered a major hindrance for practical implementation. Turbo decoder uses Maximum A posteriori Probability (MAP), Logarithmic MAP (Log-MAP) and Soft Output Viterbi Algorithm (SOVA) for decoding. The MAP decoding algorithm is optimal but computationally intense and hence they cannot be implemented in real time systems. So, simplified and sub-optimal algorithms like Log-MAP and SOVA are used for practical implementation. Sub-optimal performance of SOVA is slightly attributed to over optimistic estimation of reliability values that can be reduced by Reduction Factor (RF) approach. This paper proposes the use of two RF values corresponding to each component decoder to reduce over optimistic effect in AWGN channel. RF values giving least Bit Error Rate (BER) for each $E_{b}/{N}_{0}$ are chosen using Matlab. The proposed SOVA with RF is analyzed in AWGN channel and the results provide superior performance of proposed algorithm with low BER and nominal complexity than SOVA and MAP algorithms respectively.