Decoding algorithms for noncoherent trellis coded modulation

摘要: Noncoherent decoding of trellis codes using multiple-symbol overlapped observations was shown previously to achieve close the coherent performance. Optimal by Viterbi algorithm for L-symbol requires a number states which grows exponentially with L. Two novel suboptimal algorithms are presented, is same as original code, yielding complexity depending weakly on For practical values L, both substantially less complex than optimal algorithm. The first algorithm, basic decision feedback (BDFA), low scheme, based This suffer from increased error probability and propagation. A slight modification this can, in most cases, reduce these effects significantly. second uses BDFA building block. concept called "estimated future" its performance very optimum eases some additional memory requirements compared Performance analysis simulation results also given.