Conditional Single Tree Search Sphere Decoding
Xinyu Mao,Weiliang Fan,Zhijun Wang,Dou Li,Haige Xiang
DOI: https://doi.org/10.1109/VTCSpring.2015.7145757
2015-01-01
Abstract:Information iteratively updating between the detection and the decoding in multiple-input multiple-output (MIMO) systems is a promising method to access the channel capacity. However, the complexity of detection algorithms is very high even if no iteration is considered. Single tree search is a very powerful detection algorithm in the iterative detection of MIMO, which combines tree searches of all byte into a single tree search (STS). But inherited from the complexity of nonlinear detection algorithm, the STS algorithm is still too complicated in many scenarios. In this article, we proposed that byte whose values of log likelihood ratio (LLR) are large enough are good enough for detecting and need no further updating. We set threshold in detection. A tree where LLR values of all byte are larger than it is free of update. As a result, a lot of calculation of search can be reduced, and the complexity can be reduced lot while the performance loss is very small. When BER=10-3, threshold=0.1 or 1, the performance loss is about 0.03dB or 0.16dB. When SNR=22dB, the complexity of the proposed algorithm is about 26.6% or 20.02% of the original algorithm.