A class of soft decoding algorithms

Good decoding algorithms are known only for some families of channel codes. In the general case, the only method of decoding is to search for the codeword closest to the received vector. Proximity is normally measured either with respect to Hamming distance or Euclidean distance. We propose two alternative distance functions for minimum distance and bounded distance decoding. The new distance functions are shown to produce decoding algorithms with complexities lower than those based on Hamming distance, while offering better performance.