Research On The Cell Selection And User Scheduling Algorithm In CoMP System

Coordinated multi-point(CoMP)is a key technique which mitigates inter-cell interference by the cooperation of adjacent cells.The selection of cooperative cells does affect the system performance.To improve the service quality,multiple factors should be taken into account in cooperative cells selection system.Meanwhile,the spectrum resource is limited,which means finite users could be served in one slot.As a consequence,user scheduling should be considered to make the most of user diversity.The selection of cooperative cells considering multiple factors and user scheduling based on the coordinated beamforming have been studied.The traditional coordinated beamforming precoding algorithms,including Zero Force,Minimum Mean Square Error,Block Diagonalization and max SLNR algorithms were studied.Then modified max SLNR algorithm based on SINR was studied,which took interference signal as well as leakage signal into consideration,and the optimal beamforming vector was found by interation.The performance of five algorithms was provided by system simulation.Coordinated beamforming precoding algorithms could suppress the ICI using channel information,which is benefit to the system performance especially the cell edge performance.Cooperative cells consist of the service cell and assist cooperative cells.When the service cell transmitted data to users,the signal from assist cooperative cells would be interference if there is no coordinated beamforming.Therefore,the cooperative cells channel information is the key factor of cell selection.What's more,the load of service cell is another important factor of service cell selection,especially in the scene of user distribution is not uniform.The optimal cell could be found by multiple objective algorithm,which could improve user data rate and user satisfy rate.User scheduling selects the optimal set of users from the overall instead of selecting the optimal user from each cell independtly in coordinated beamforming system.The principles of user scheduling in one cell are still used here,that is,system sum rate and user fairness,which shiftwith each other.User scheduling algorithms trade off between the two principles.A modified proportional fairness algorithm based on Differential Evolution is studied here,which could achieve favorable system performance rapidly as the excellent optimizing performance of Differential Evolution algorithm.