Research On Key Technologies Of Downlink Of Wireless Communication Private Network Based On Massive MIMO

With the maturity of 5G technology,the application of wireless communication private network based on massive MIMO will become a new research demand.The coverage range of the base station of 5G private network is much larger than that of ordinary commercial 5G cellular mobile communication system.It works in suburban areas,islands,mountainous areas and other applications and supports high mobility.As a result,the delay extension of the channel is larger and the maximum doppler frequency deviation is larger.According to the demand characteristics of 5G private network,SDMA precoding and detection algorithm for massive MIMO downlink is studied and determined in this paper,and IP core is formed through FPGA hardware implementation.In this paper,massive MIMO downlink detection algorithm is firstly studied,and formula derivation,simulation experiment and performance analysis and comparison of detection algorithms such as ZF,MMSE and lmmse-isdic are carried out.In the centralized massive MIMO scenario,a new iterative detection structure is proposed based on the lmmse-isdic detection algorithm,which improves the overall performance of the system through the interaction of soft in soft out detector,soft in soft out decoder and equivalent channel estimator and other soft information.Simulation results show that the performance of the new iterative detection algorithm is better than that of the original algorithm.Secondly the research based on the mass of the MIMO wireless private network down link SDMA precoding,the BD precoding algorithm,yearly income and GMI users formula deduction,simulation and performance analysis and comparison of the results showed that the yearly income is the most suitable for massive MIMO air separation of multiple access wireless private network base station side downward SDMA precoding,satisfies the requirement of wireless private network system design specifications.The paper then studies the downlink detection algorithm of the massive mu-mimo wireless private network system using GZI precoding.At this point,the equivalent channel matrix dimension of the massive mu-mimo(32x32,16 users)base station and the single user is 2x2.Equivalent 2 x2 MIMO channel,on the MMSE-PIC with LDPC decoding,SD,and the combination of LDPC decoding two iterative algorithm formula derivation,EVA channel scenarios are given ZF and MMSE,MMSE-SIC,MMSE-PIC,SSD,SSDKB,ML detection algorithm simulation and performance analysis,results show that the sphere decoding algorithm could be used in 5 g private network system to optimize detection performance.At last,the paper adopts Vivado HLS tool to select KUC105 FPGA chip to develop the IP core of SDMA precoding and terminal side detection algorithm at base station side of massive MIMO private network.First,the IP kernel implementation of the basic matrix operations used in downlink,such as matrix multiplication,matrix QR decomposition,matrix SVD decomposition and matrix inverse,is given.Then,according to the principle of SDMA precoding algorithm on the base station side of downlink,the IP kernel of SDMA precoding algorithm on the base station side of downlink is realized.The resource occupation and Vivado HLS simulation results are given.Finally,according to the principle of downlink terminal side detection algorithm,the IP core of downlink terminal side detection algorithm is realized,and the results of resource occupation and Vivado HLS simulation are given.