Recursive semi-blind decoding of a space-time block code over frequency-selective channel

In this paper, a recursive semi-blind decoding method for Alamouti space-time block code over frequency-selective channel is proposed. The method combines subspace-based blind equalization with the structure of Alamouti space-time block code, and recursively recovers the information symbols from a temporally oversampled received signal on multiple receive antennas without channel state information. With the aid of a few training symbols, the proposed recursive algorithm has two fortes: First, because different transmit antennas employ distinct training sequences, it can resolve the ambiguity inherent in Alamouti code when employing blind equalization method; Second, on each recursion, by using the previously detected symbols, the number of unknown symbols in the processing window can be reduced, consequently, the computation complexity is decreased.