User Partition Based Inter-cell Interference Suppression In Massive MIMO Systems

As one of key technologies of next generation 5G communication systems,the massive MIMO technology significantly improves spectrum efficiency and data link reliability by making full use of spatial degree of freedom(Do F).Inter-cell interference is the main factor which affects the performance of massive MIMO multi-cell multi-user systems.The existences of pilot contamination(PC),high feedback overhead and high computational complexity make inter-cell interference more complex,which lead the inter-cell interference suppression more challenging.User partition is an effective method to alleviate inter-cell interference by processing the central user and the edge user separately,which improves the performance of the edge user.In this paper,to improve the system performance of the threedimensional(3D)massive MIMO multi-cell multi-user systems,we investigate the inter-cell interference suppression through user partition scheme based on vertical channel information and scheduling.The details are as follows:The existing user partition scheme in 3D massive MIMO scene can not make full use of Do F in vertical direction and have high feedback overhead.To tackle this problem,a user partition scheme based on vertical channel information is proposed.The method divides the users whose vertical information is similar to that of the users in the neighboring cell into edge users.The soft pilot reuse(SPR)scheme is used to allocate same pilot resources to central users in all cells,and orthogonal pilot resources are applied for edge users to reduce PC problem.The modified channel estimation scheme for central users is proposed to improve the accuracy of channel estimation in time division duplex(TDD)systems.Two-stage precoding is carried out for cell center users during downlink transmission,and according to the signal-to-leakage-and-noise ratio(SLNR)of edge users in adjacent cells,edge users are reallocated to the base station(BS)with the minimum SLNR,so as to reduce the interference to other edge users.Simulation results show that the proposed scheme can achieve more accurate channel estimation,reduce the number of edge users when the number of vertical antennas is large,thus reducing pilot overhead and improving system and speed.Therefore,compared with other schemes,this scheme can be applied to three-dimensional large-scale MIMO systems.The simulation results show that this algorithm can achieve more accurate channel estimation,reduce the number of edge users when the number of vertical antennas is large,thus reducing pilot overhead and improving system sum rate.Therefore,compared with other user partition schemes,this scheme is more suitable for 3D massive MIMO systems.Considering the problem of the existing multi-cell user scheduling scheme is not suitable for massive MIMO frequency division duplex(FDD)system due to the high feedback overhead and high computational complexity,a multi-cell user scheduling scheme based on joint spatial division multiplexing(JSDM)is proposed,which partition the center/edge users into groups according to the channel covariance matrix.For each group of dimension-reduced center users,the greedy scheduling with low complexity is carried out to reach the maximum sum-rate,while for each group of edge users,sequential scheduling method is adopted to realize inter-cell interference suppression.Finally,inter-cell interference and multi-user interference is eliminated by precoding.Simulation results show that the proposed algorithm can reduce the feedback overhead and computational complexity on the premise of improving system performance.