Complexity reduced soft MIMO detetion using single tree search

Abstract

In MIMO systems, soft iterative detection and decoding can produce the near capacity performance. One of the promising detection techniques known as sphere decoder can play an important role in order to meet the requirements of achieving near optimal performance. The single tree search (STS) is based on the sphere decoding which can produce near optimal performance in iterative detection and decoding. The main hindering in STS method is that it is computationally complex. The complexity increases as we increase the iterations. In this paper, we propose to reduce complexity of the STS method by limiting the candidates of calculating soft information to those bits whose a priori information provided by turbo decoder is less reliable. Simulation results show that the proposed method can reduce the complexity with negligible performance degradation compared to the conventional full search and STS methods.