An approach based on soft-statistics to the joint adaptive equalization/decoding of TCM data for multipath channels

Adaptive detection based on soft-statistics of Trellis-Coded-Modulated (TCM) data transmitted over multipath-degraded fast time-variant channels is considered. Starting from some system-considerations about the equalization/decoding of multipath radio mobile links, a fast-adaptive receiver suitable for cellular applications is presented where the tasks of channel-estimation, equalization and decoding are carried out in a combined optimized form. The last two operations are jointly accomplished by a Symbol-by-Symbol (SbS) Maximum A posteriori Probability (MAP) detector which delivers soft-statistics in the form of suitably defined A Posteriori Probabilities (APPs). These APPs are used for channel-estimation so that the resulting channel-estimator is optimal (in a Minimum Mean Square Error (MMSE) sense) and is able to exploit the redundancy introduced by the code. As a matter of fact, it exhibits an enhanced tracking capability when the impulse-response of the channel is fast time-variant.