Research On Wireless Transmission Method In Spatial Non-stationary Massive MIMO Channels

With the rapid development of wireless communication technology,various novel services have been proposed,and the 6th generation mobile communication technology(6G)oriented to 10 years later has attracted more and more people’s attention.Extra-large scale massive multiple-input multiple-output(XL-MIMO)technology equipped with ultra-large antenna arrays has become one of the key technologies of 6G due to its advantages such as good cell coverage and high regional throughput.However,in practical applications,when the size of the antenna increases,spatial nonstationarity will occur,and the user can only see part of the antenna array of the base station,which is called the visibility region(VR).At the same time,due to the different positions of users and scatterers,each user has its own specific VR,and the VRs of different users can be overlapping or non-overlapping.Currently,most of the literature studies non-overlapping scenarios of user VR.In order to explore the ubiquitous wireless transmission method in user VR overlapping scenarios,this paper conducts research on the XL-MIMO wireless transmission method based on overlapping VR.The specific contributions are as follows:First,for a scenario where there are only two user VRs and they overlap,a VR transmit power allocation method in XL-MIMO system with the same user VR is proposed.Consider a communication system composed of two single-antenna user groups and an XL-MIMO array.The user VRs in the same user group are the same,but the two groups of user VRs overlap each other.Using the spatial non-stationarity and near-field properties of XL-MIMO systems,a simplified channel model is constructed.According to the VR distribution of user groups,the antenna arrays are divided into overlapping VRs and non-overlapping VRs,and each VR adopts regularized zeroforcing(RZF)precoding to communicate with the users it serves,and analyze the system ergodic sum rate.Based on statistical channel state information(CSI),the large-scale system approximation results of the system ergodic sum rate are deduced by using large-dimensional random matrix theory.By maximizing this approximation,a power allocation method based on alternating optimized overlapping VR and non-overlapping VR is proposed.The simulation results show that the derived approximate rate can well approximate the ergodic sum rate,and the proposed power allocation method and precoding scheme can effectively improve the system performance.Secondly,a user power allocation method based on overlapping VRs is proposed for scenarios in which each user VR is different and overlaps with each other.Consider a communication system composed of multiple single-antenna users and an XL-MIMO array,where the VRs of the users overlap in pairs.A simplified XL-MIMO channel model is constructed.According to the VR distribution of users,the antenna array is divided into multiple sub-arrays,and each VR uses RZF precoding to communicate with the users it serves,and analyzes the system ergodic sum rate.Based on statistical CSI,the large-scale approximation results of the system ergodic sum rate are derived by using large-dimensional random matrix theory.By maximizing this approximation,a closedform solution for the optimal power for each user is derived.The simulation results show that the approximate rate derived has high accuracy,and the closed-form solution of the proposed user optimal power can effectively improve the system performance.