Research On User Selection With Precoding In MIMO Systems

MIMO is a key technology generally adopted by new mobile communication systems which need higher spectral efficiency with the requirement of high speed data service. Compared with single user MIMO, multi-user MIMO can obtain extra multi-user diversity gain and draws much attention. As core technologies to obtain multi-user diversity gain, user selection and precoding have become hot topics.Having analyzed the idea of SLNR, ZFBF and BD, the paper researches on user selection technology based on precoding from the perspective of joint optimization of precoding and user selection.1. SLNR precoding user selection technology mainly discusses the topic of user with single antenna. The paper gives a user selection algorithm based on SLNR precoding aimed at solving the problem of user with multi-antenna. The algorithm splits the joint selction problem of antennas and users to selection of antennas and users in sequence through reducing demensions, which overcomes the shortcoming that SLNR is great influenced by joint channels'dimensions. The algorithm is applicable to the mode of user with multi-antenna, which makes it more practical.2. Aimed at reducing the complexity in interference elimination precoding algorithm, the paper gives a user selection algorithm based on codebook clustering algorithm. The algorithm takes the advantage of the influence of channel correlation between selected users and alternative users on the system performance. It deletes alternative users that have a strong correlation with selected users and narrows down the search space during the iterative procedure, which reduces the complexity of the algorithm. Analysis result shows that the complexity improves remarkably compared to classic greedy algorithm. Moreover, the more users it selected, the less complicated is the algorithm.3. Under the condition of limited feedback, the paper gives a channel codebook generation algorithm based on Lloyd, which exploits Lloyd vector coding principle that sets chordal distance as the criteria of distortion judgment. With small quantization error, the algorithm has a better sum capacity performance than RVQ algorithm. Then the paper gives a user selection algorithm based on minimum sum capacity loses depending on the square of channel quantization error. Through compressing CQI's value range, it reduces the complexity of base station compared with SINR estimation algorithm and improves the sum capacity at high SNR. The paper designs several user selection algorithms with low compexith and amount of feedback under different application scenarios. In order to further verify the validity of proposed algorithms, a user selction algorithm simulation platform is developed and performances of some proposed algorithms in this paper are tested on the platform.