Joint ML Channel Tracking and Decoding for MIMO Channels

In this paper, a joint tracking and decoding algorithm is proposed for MIMO channels. This algorithm is based on a super trellis structure which calculates new estimations of the channel matrix; moreover to calculation of a new metric for each node of the trellis. This algorithm for each branch of the trellis diagram of convolutional encoder, calculate matrix estimation by ML channel tracking algorithm. The proposed algorithm is simulated for 1/2 rate convolutional code and 2x2 MIMO channel for different values of Doppler frequencies and E_b/N₀. Bit error rate (BER) and channel tracking error are considered as comparison criteria and it is shown that its performance is good and channel tracking curves have always negative slope even in blind mode of the operation. The more simulations with more efficient channel models will be presented in full version of the paper.