一种基于权重干扰泄露的分布式预编码算法

The use of precoding in the downlink of multi-user multiple-input multiple-output (MU-MIMO) systems can suppress co-channel interference effectively. Many schemes have been given to design precoding matrix for different optimization objective. In this paper, a distributed precoding scheme, named block diagonalization with weighted signal to leakage plus noise ratio (BD-WSLNR) for MU-MIMO systems with multiple data streams is proposed. It can nullify the intra-cell interference and mitigate the weighted leakage interference to other cell users according to their interference affordability based on channel quality information (CQI). Besides, a generalized singular value decomposition (GSVD) based matrix calculation method is introduced to improve the calculation accuracy and reduce the gap between effective SINRs of multiple data streams. Simulation results show that, compared to the existed schemes such as block diagonalization (BD), max signal to leakage plus noise ratio (max-SLNR), and block diagona-lization with other cell interference (BD-OCI), the proposed precoding scheme achieves considerable gains in the spectral efficiency of edge users, while maintaining nearly the same achievable sum-rate.