Research On Downlink User Schedulingalgorithm In LTE-Advanced CoMP System

As the evolution version of 3GPP LTE,the LTE-Advanced system targets for a higher system throughput and spectral efficiency than LTE R8 system and aims at meeting with performance guideline proposed by ITU for IMT-Advanced system. As the cell continues to get smaller and frequency reuse factor becomes 1,the cell performance is significantly limited by the inter-cell interference (ICI). Since the existing interference coordination mechanisms in LTE system fail to effectively resolve the problem of low spectrum efficiency of cell edge users caused by interference between different cells,LTE-Advanced introduced a effective inter-cell coordination technique named Coordinated Multi-Point. In this transmission mode,by employing joint processing or coordinated scheduling among multiple cells,the interference signals among them are turned into useful ones and the adjacent cell interference could be depressed. Meanwhile, the received signal quality of cell edge users is improved. Thus cell edge throughout and average cell throughout are both enhanced.Currently,CoMP,which is being widely discussed and standardized,can be classified into joint processing and coordinated scheduling/coordinated beamforming according to transmission schemes. From another aspect,CoMP are itemized into CoMP-SU-MIMO and CoMP-MU-MIMO according to the user number scheduled at a special time slot or physical resource block.This thesis mainly focuses on JP transmission of CoMP,mainly on downlink user scheduling algorithm of CoMP-SU-MIMO and CoMP-MU-MIMO. A joint proportional fair scheduling algorithm is proposed in CoMP-SU-MIMO system. Meanwhile,an adaptive user pairing algorithm with variable user number in the coordinated clustering is brought forward in CoMP-MU-MIMO system . The paired users are chosen on the basis of sum capacity maximization criterion,channel correlation coefficient minimum criterion and user fairness criterion respectively. The corresponding simulation results and algorithms analysis are demonstrated.