Adaptive bi-modal decoder for binary source estimation with two observers

In this paper, an implementation friendly distributed version of parallel concatenated convolutional codes (D-PCCC) with a novel bi-modal iterative decoder is proposed to estimate a binary data source observed by two sensors. The convergence of the iterative decoder is analyzed by defining a modified EXIT chart. It is shown that iterative method is not always the best decoding algorithm for D-PCCC with two encoders. The system quality conditions in terms of channel SNR and sensors´ observation accuracy required for usefulness of iterative algorithm is found. This provides a frame by frame mode selection criterion for decoder. The decoder chooses an appropriate iterative or non-iterative decoding mode that yields higher BER performance in both modes and reduces decoding complexity by avoiding destructive or useless iterations.