Semi-blind interference cancellation with distributed training

Some of the wireless standards, e.g. IEEE 802.16-2004, allow successive transmission of a number of consecutive slots, where each of them contains a training (pilot) interval. If a number of neighboring cells asynchronously transmit similar frames, then the resulting interference environment becomes similar to a distributed training scenario. In this paper, a distributed training scenario is addressed by means of semi-blind (SB) processing at the multiple-antenna receiver. A second-order statistics identifiability is analyzed and a SB algorithm is proposed for the situation, where the data and training intervals contain different sets of interference components. Its performance is assessed by means of comparison to the non-asymptotic maximum- likelihood (ML) benchmark initialized by the developed identification algorithm. It is demonstrated that the SB solution significantly outperforms the conventional training-based algorithm and approaches the benchmark in the case of low number of available data symbols. Using all the considered second-order solutions as initializations for the higher-order statistics iterative algorithm is also addressed.