Co-channel interference must be suppressed in Uplink multiuser MIMO system. In this paper, an algorithm based on the criterion of maximizing signal to interference and noise ratio was proposed to suppress multiuser interference, and the improved algorithm in the presence of channel estimation error was also provided. In comparison with BD algorithm, The algorithm taked into account the influence of noise, and the performance had greatly improved. Then a method of reconstructing signals was proposed to employ Alamouti space-time coding in uplink multiuser MIMO system for the reason of further improving the detection performance. Finally aimed at limitations of algorithms with the near-far effect, an optimization algorithm of grouping users was proposed to ensure that each user can achieve the acceptable capacity. Simulation shows that the proposed algorithms achieve good performance.