In this paper a proactive scheduling algorithm is proposed based on associative interference of spatial subchannels for MU-MIMO downlink (broadcast) channel. The strategy converts user scheduling into subchannel selection issue. With comprehensive consideration of candidate subchannel transmission gain, along with mutual interference among candidate and selected subchannels, as well as those to be selected potentially. A set of subchannels with less mutual interference are achieved. Simulation results show that by choosing proper associative interference parameters, the proposed algorithm can achieve good tradeoff between computational complexity and transmission performance, and improve system sum rate effectively.