An Efficient Hybrid Sphere Decoding Algorithm for MIMO Systems

For MIMO data detection methods, sphere decoding (SD) algorithms have been shown to achieve similar performance to the maximum-likelihood detection (MLD) algorithm with acceptable computational complexity. Based on different search strategies, SD can be divided into two parts: depth-first search SD and breadth-first search SD. In this work, by combining the advantages of these two search schemes, we propose a novel hybrid sphere decoding to effectively reduce the complexity. Besides, a layer reordered method is adopted to improve the performance. Compared with other SD algorithms, the proposed one has not only good performance but also lower complexity.