Hidden Markov Model Based User Selection Mechanism for MIMO Beamforming

• Main Authors: Chien-Chung Lai, 賴建中
• Other Authors: Hsing-Piao Lin
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/36w5hm

碩士 === 國立臺北科技大學 === 電腦與通訊研究所 === 98 === The core technology research in Multi User - Multiple Input Multiple Output (MU-MIMO) is user selection mechanism and beam forming technology. It mainly utilizes the messages through feedback channel to know the spatial relationship among users and to find out best user set. Then, base station will calculate the suitable beam pattern for these users by those channel messages. However, some channel messages like channel fading are impossible to feedback. Even if feedback those messages are possible, the quantization technology would be adopted. Then, the performance of user selection mechanism would decrease even beam forming technology. Thus, this thesis proposes one technology which does not depend on these unreliable messages, and it adopts fixed beam to co-schedule users and beam patterns. In time-vary situation, in order to solve the performance of beam switching speed which does not affect system performance. This thesis introduces the Hidden Markov Model (HHM) which is a probability model. By estimating transition state of user’s channel and predicting transition degree, it can obtain the degree of channel’s change. Then, the user selection mechanism can select the users which have higher time correlation. Hence, the system capacity would not change rapidly, when the same user set and beam weights are communicating for some time slots. Thus, the user selection mechanism and beam forming technology are not necessary to change every slot, even reduce the feedback information.