Research On Key Technology Of MU-MIMO In Next Generation WLAN

Due to continuous growing in the number of users and their demand for data traffic,existing wireless local area network(WLAN)has been unable to meet the needs of users.In order to improve the throughput of per unit area covered by WLAN and enhance the user experience,a new standard named IEEE 802.11 ax has been developed,which could be adapted to a variety of indoor and outdoor dense user scenarios.As a key technology of IEEE 802.11 ax,the downlink MU-MIMO technology can improve the spectral efficiency of WLAN.The key to implement the downlink MU-MIMO is designing a suitable precoding matrix through limited channel state information(CSI)to eliminate interference between users.Therefore,the accuracy and real-time of CSI feedback is essential to the performance of MU-MIMO precoding.In practical implementation,as the increase of the frequency of channel estimation and the number of users involved in MU-MIMO,the overhead of CSI feedback increases,which seriously affects the performance of the system.Therefore,how to feedback the CSI effectively is the key to implementation of IEEE 802.11 ax.Aiming at the above problems,the research of this paper is as follows:Firstly,the MIMO system model and the time coherence characteristic and channel coherence bandwidth of MIMO wireless channel is studied in this thesis.Besides,several linear precoding techniques commonly used in downlink MU-MIMO including ZF precoding,BD precoding and SVD decomposition precoding are discussed,and the related simulation is carried out to analyze the performance of these precoding technology.In order to reduce the complexity and improve the scalability of the platform,the SVD decomposition precoding is chosen in this thesis.Then,in the dense deployment scenarios of the next generation WLAN,the CSI feedback in downlink MU-MIMO system consumes too much channel resource.However,if the amount of CSI feedback is reduced,the inaccuracy of the CSI would degrade the system performance.How to reduce the amount of CSI feedback while ensuring the system performance is the problem will be solved in this paper.So,an adaptive feedback period adjustment algorithm is proposed in this thesis.With the joint processing of AP and STA and the use of the time correlation of the channel,the period of feedback is adjusted in real time.The simulation results show that the algorithm can improve the average sum rate of the system.At the same time,combined with the specific protocol process in WLAN and according to the frequency domain and time domain characteristics of wireless channel,several related factors,including the size of subcarrier packets,the number of quantization bits and the feedback delay are investigated.The influence of these factors on system performance is analyzed in theory.The results of theoretical analysis are also verified through building a link-level simulation system.Finally,a platform implementation scheme of SVD decomposition precoding based on compression feedback is designed in this thesis.In order to meet the real-time processing requirements and improve the computing speed of the SVD decomposition algorithm,a parallel bilateral Jacobi rotation algorithm is proposed,which is also implemented on the hardware platform.The results of the experiment show that the algorithm can quickly complete the process of SVD decomposition.At the same time,the pipeline architectures for the compression and decompression of precoding matrix are provided in this thesis.The consumption of the time resources and hardware resources of the designed pipeline architectures are then assessed in the hardware platform.