Space-Partition-Driven Learning Network for Efficient MIMO Detection

• DOI: 10.1016/j.procs.2021.04.091
• Publication Date: 2021-06-12T20:32:59Z
• Authors: Dadan Xu; Qinghe Du
• Keywords: 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

With the development of communication systems, MIMO order low-cost equipments have also increased. However, classical algorithms either too high computational complexity or poor performance. Some existing deep learning-based detection networks achieved excellent results in high-order MIMO, but they unbearable performance low-order MIMO. Based on idea dividing sample space, we propose MLNet. Compared with Det Net and MMNet, MLNet has made a huge leap time-varying channels. At same time, four orders magnitude less computation than neural network algorithms. We training method, which is not only useful for MLNet, improves DetNet.