Performance improvement of Max-Log-MAP Turbo decoding algorithm using optimized scaling factor for image transmission and retrieval

In order to achieve error correction nearer to Shanon´s limit, Turbo decoder imparts either Soft Output Viterbi Algorithm (SOVA) or Maximum Posteriori Probability (MAP). The feasibility of most expected path through the Trellis is mainly determined by Max-Log-MAP, a Soft Input Soft Output (SISO) algorithm and gives a better performance over MAX-LOG-MAP (MLMAP) algorithm. To improve the decoder´s performance of MLMAP algorithm is usage of appropriate Scaling factor (SF). This enhancement in performance is achieved by reducing Bit Error Rate (BER) can be made real by fixing an random SF for inner decoder s2 and an refined SF for outer decoder s1. By comparison among various SF values, an optimized Scaling value is obtained which is so an empirical value. This empirical value usage hikes the performance of MLMAP decoding algorithm in terms of BER. Mathematical relationship between SF and Eb/N0 is also proposed. The M-MLMAP algorithm showed a gain of 0.75dB over MLMAP algorithm at BER of 2×10-5 for Rayleigh fading channel. The improved M-MLMAP algorithm is used for image transmission and retrieval. The proposed algorithm gave superior performance than the preceding algorithms for wireless image transmission.