Performance Improvement of MU-MIMO System by Optimizing the K-Factor for the K-Mean User Grouping Algorithm

Abstract

In this paper, downlink multiuser-MIMO system with large number of transmitting antennas at the base station and R user terminals each having single antenna is considered. According to this design, an access point communicates with large number of users in the Rayleigh fading scenario. Due to large number of users, it becomes difficult to accommodate all of them in the system simultaneously. So, a user grouping technique known as K-mean clustering is used, such that a group of users with similar conditions at that particular time are served together. While making groups, the interference is surely reduced but the number of users being served at a time also reduces. So, it is necessary to make out the balance such that the performance of the system is maintained while accommodating maximum number of users. So, optimum number of user groups needs to be formed. The results show that when groups are increased from two till four sum rate increases but when five groups are made the sum rate decreases to a point but, is still higher than two groups.