Performance Enhancement of Modified SOVA Decoding Algorithm for Turbo Codes

Turbo coding (TC) has been adopted as a channel coding scheme for several 3G mobile systems, in particular GPP(third Generation Partnership Project) and 4G standards for high data rates. Turbo decoder uses any one of the decoding algorithms, Maximum A posteriori Probability (MAP), or Soft Output Viterbi Algorithm (SOVA) because it produces error correction near to Shannon’s limit. SOVA is an extension of classical Viterbi algorithm which is incorporated with a-priori information. The performance of SOVA algorithm is improved by scaling the extrinsic information exchanged between two decoders. Modified SOVA (MSOVA) algorithm is achieved by fixing an arbitrary value for inner decoder (S2) and an optimized value for the outer decoder (S1). A simple but effective technique to enhance the performance of MSOVA, called Enhanced SOVA (ESOVA) is proposed. In ESOVA, the arbitrary scaling factor S2 is optimized and is made to adapt with received Eb/N0. The use of an emphatically determined optimal scaling factor improved the performance of the decoding algorithm in terms of Bit Error Rate (BER). A comprehensive analysis of the proposed algorithm, by considering different channel conditions and code rates are presented. Appropriate mathematical relationship between scaling factor and Eb/N0 is also proposed.