Stochastic List Generator for Iterative MIMO Detection

Abstract

An iterative maximum-a-posteriori (MAP) multiple-input multiple-output (MIMO) detector is presented. We take note that to develop a low-complexity detector one should first obtain a list of candidate samples of the transmitted data symbols that closely match with the received signal. Here, for the list generation, we expand on a recently proposed stochastic sampling method. Two methods are developed and demonstrated. The first method, called single list stochastic list generator (SL-SLG), generates a list at the first iteration of the turbo loop, i.e., without the benefit of any a priori knowledge, and used throughout the iterative detection process. The second method, called update list stochastic list generator (UL-SLG), updates the list after each iteration using the a priori information provided by the channel decoder. The effectiveness of these stochastically generated lists are benchmarked against the celebrated method of K-best. Extensive computer simulations, using real-world MIMO channels, reveal that the proposed method outperforms the K-best method, when the system parameters are set for the same list size. It is also noted that whereas the list generation method in K-best follows a sequential approach, the stochastic sampling method proposed in this paper is tailored towards a parallel implementation, which, helps in reducing the detector latency significantly.