Research On Multi-user Clustering And Channel Estimation Algorithm For Massive MIMO Two-tier System Based On Local Discriminant Projection

The development of the communication industry can be said to lead the development of the times,and the development of 5G networks occupies a vital position.With the investment of a large number of resources,the use of 5G networks has developed rapidly.In the process of using 5G,5G has obvious advantages,such as reducing end-to-end delay;the overall network speed is greatly improved.Then 5G brings us the transmission rate beyond the fiber,beyond the real-time capabilities of the industrial bus and the connection of the whole space.Mobile networks are digitizing the entire industry and becoming the underlying productivity.In order to improve the network speed,reduce the delay,and improve the user service quality,the cell density in the wireless network is continuously increasing,and the cell radius is continuously reduced.Therefore,in the future 5G research,small cells are densely and densely deployed in a massive Multiple Input Multiple Output(MIMO)system to form a massive MIMO two-tier system,which can solve the problems faced in the future development of wireless communication systems.The specific content and conclusions are as follows:With the rapid growth of high-speed wireless data access services and users,for the deployment of macro base stations and the placement of small cell base stations in ultra-dense massive MIMO two-tier systems,a clustering scheme suitable for actual scenarios is studied,thus ensuring improve the throughput of the system with a certain communication quality.Firstly,for the scenario of massive MIMO two-tier systems for high-density users,this paper proposes a user grouping algorithm based on local discriminant projection using angle and distance as the characteristics of similarity measure.The algorithm improves the user grouping effect,reduces the computational complexity of the high-dimensional channel matrix,and reduces the cell-interference.The simulation results show that the proposed algorithm improves system capacity and throughput.Secondly,in order to reduce user inter-group interference and inter-user interference,this paper proposes a minimum Mean Square Mean Square Error Estimation(MMSE)precoding.Its outer precoding uses statistical channel state information(CSI)to reduce inter-group and inter-layer interference,while internal precoding uses instantaneous effective channel information to mitigate intra-group interference.The simulation analysis shows that the proposed two-layer MMSE precoding performance is better than other precoding schemes.Finally,for the acquisition of channel state information of complex interference channels in massive MIMO two-tier systems,this paper proposes a three-stage data-assisted channel estimation method.In order to implement the data assist scheme,it is assumed that there is no error in the system and no delay,if the data detection and decoding data sequence is completed in the small cell base station and transmitted to the macro base station through the wired backhaul.Since the channel of the macro base station after user grouping is sparse,and thus the downlink channel at the macro base station utilizes the decoded uplink data and the known training sequence,an optimal block orthogonal matching is proposed.Channel estimation algorithm of Optimal Block Orthogonal Matching Pursuit algorithm(OBOMP).The simulation results show that the data assisted method proposed in this paper can effectively improve the channel estimation accuracy.