为实现认知多输入多输出( MIMO)系统中主次用户的高效频谱共享，提出一种基于部分投影的干扰抑制算法，该算法将主用户的空间传输特性引入次用户的设计中。分析了干扰的结构化特征，指出泄漏到主用户空闲维度上的次用户信号并非有害干扰。通过将次用户发射信号投影到干扰信道低维子空间的零空间上即可消除干扰。理论分析和仿真表明，算法在有效消除对主用户干扰的同时，次用户传输速率和算法适用性相比传统算法都有显著提升。

Partial projection based interference suppression algorithm is proposed for cognitive Multiple-Input Multiple-Output(MIMO) system to achieve effective spectrum sharing between primary user and cognitive user. In this algorithm, spatial transmission characteristics of the primary user are introduced to the design of secondary user’s transmission. Based on analysis of structural characteristics of the interference, it is proved that the leakage signal from secondary transmitter to the idle dimensions of primary user is not harmful. Then the secondary signals are projected to the null space of low-dimensional subspace of interference channel between primary and secondary user to eliminate interference. Theoretical analysis and simulation results show that the algorithm effectively eliminates interference to the primary user; the secondary user’s transmission rate and the applicability of the algorithm have been significantly improved compared to the traditional method.