Research Of Cooperating Set Selection And Feedback Technology In COMP Of LTE-A

With the increasing demand of data rate of mobile terminals,3GPP has launched a research of the Long Term Evolution of UMTS. In order to meet the requirement of IMT-Advanced,3GPP has also started the smooth evolution from LTE to LTE-A. As a newly introduced technology in LTE-A, CoMP has attracted lots of attention from the first time to be put up to now. Applying this technology in a real network can make use of the cooperation between base stations to alleviate or eliminate the inter-cell interference, thus improving the throughput of cell-edge users. The advantage of CoMP has been broadly discussed and proved, but the performance can get even worse if the deployment is not suitable. These problems include the disadvantage of improper selection of cooperating sets and the decrease uplink performance caused by big burden of feedback information. This paper focuses on these two problems and proposes improved schemes compared with traditional solution. The main works of the paper includes:Firstly, this paper proposes three strategies of cooperation sets selection based on SLNR. These schemes are designed to minimize the data exchange between cooperating base stations while maintaining the QoS of each user. The detailed procedure included three stages. In the first stage, SLNR is introduced in the selection of cooperation sets and beamforming vectors are designed based on the rule of maximizing the SLNR. In the second stage, downlink user scheduling schemes are presented based on the beamforming vector calculated in the first stage and greedy algorithm. This scheduling strategy is used to guarantee the SINR of each user. In the third stage, three clustering methods which mean the selection of cooperation sets are given based on the results of the first two stages. Three clustering methods are SLNR-based, SLNR-relative and greedy algorithm respectively. The simulation results have shown that all the clustering methods based on SLNR have performed better than the traditional schemes based on channel strength, both in system throughput and the data exchange. The greedy algorithm performs best in the three schemes,.Secondly, this paper gives some research on the feature of feedback schemes in single cell and multiple cells. In the single cell, the explicit feedback based on SVD and codebook selection are discussed respectively. Two codebook schemes are introduced, including the4-antenas codebook of3GPP Release8and8-antenas codebook of Sumsung. The research results show that explicit feedback has the best performance and the highest complexity, and implicit feedback is more suitable in practical network. In the multiple cells, a self-adaptation feedback encoding scheme based on CQI is proposed to decrease the amount of feedback information. It assigns different bits for different base stations based on the CQI. The simulation results shows that this self-adaptation can reduce the feedback bits significantly at the price of small performance decrease.