GEVD辅助的MMSE干扰抑制下行多用户MIMO系统

GEVD Aided MMSE Interference Suppression for the Downlink of Multi-User MIMO System

为提高多用户多输入多输出(MIMO)系统中多个用户的预编码增益,提出了一种利用广义特征值分解(GEVD)辅助的最小均方误差(MMSE)干扰抑制算法,抑制干扰和噪声的同时考虑了用户的预编码增益、最大化发射端的信漏噪比和接收端的信干噪比,并给出有信道误差情况下修正的预编码设计方法.理论分析和仿真结果表明,新算法比广义的MMSE信道反转算法具有更好的系统性能.

A general eigenvalues decomposition（GEVD） aided interference suppression algorithm is proposed based on the minimum mean-squared error（MMSE） criterion to improve the precoding gain of multi-user in multi-user multi-input multi-output（MIMO） system.It suppresses the interference and noise,meanwhile,the users pre-coding gain is taken into account at the same time.The algorithm can maximize the signal-to-leakage-and-noise-ratio at the transmitter and maximize the signal-to-interference-and-noise-ratio of each user＇s receiver.And how to modify the design in order to account for channel estimation errors is also analyzed.Simulation shows that the algorithm can achieve a better system performance than the general MMSE invension algorithm.