Energy Efficiency Analysis And Optimization Of Low-precision Quantized Millimeter-wave Large-scale MIMO System

The large-scale popularization of personal laptops and smart phones,and similar electronic devices not only promotes the all-round development of the digital age,but also brings about great growth for the transfer data.Therefore,the application of millimeter wave technology with ultra-high speed and ultra-wide bandwidth comes into being.Aiming at the short board with high transmission loss of millimeter wave,the introduction of massive multiple input multiple output(mMIMO)system can provide beamforming gain,which complement each other and they are both indispensable technology in many future years.However,mmwave large-scale MIMO radio struncture has the problems of large power consumption and expensive devices,this hinders the further development of the system.To overcome the mentioned problems,our research mainly studies how to reduce the power consumption and cost of mmwave widespread MIMO radio structure by introducing low-precision quantization and oversampling technology,and further designs the energy efficiency optimization scheme based on the analysis of Spectrum Efficiency,SE)and Energy Efficiency,EE)of the above-mentioned theme.Firstly,the energy efficiency of millimeter wave large-scale MIMO system with low-precision quantization and oversampling technology is theoretically deduced and simulated.In this study,firstly,the millimeter wave oversampling uplink widespread MIMO radio structure based on alldigital beamforming(DBF)architecture is modeled,and the nonlinear quantization process is approximately transformed into a linear process by Bussgang decomposition theory,and then the theoretical derivation of the spectral efficiency and energy efficiency of the system is completed by using the linear signal processing algorithm with low complexity.In order to facilitate large-scale simulation and result analysis,a visual simulation platform with separated front and back ends was built by using frameworks such as VUE and SpringBoot.With this platform,the quantitative relationship between system spectral efficiency,energy efficiency and the number of base station antennas,quantization accuracy and oversampling rate can be observed.The experiment results told us that low-precision quantization combined with oversampling technology can not only reduce the total power consumption of mmwave widespread MIMO radio structure,but also improve the equivalent quantization accuracy of ADC device of RF chain at the same time,and finally efficaciously increase the energy efficiency of the whole framework.Secondly,considering that the channel state between each antenna of the base station and the user is not the same,if the base station is equipped with analog-to-digital converters with fixed quantization accuracy for all antennas,the capacity of the communication system cannot be fully explored.Therefore,the thesis designs an all-digital beamforming architecture with adaptive ADC configuration accuracy by using Channel State Information,CSI)between users and each antenna of the base station,which can adaptively adjust the quantization accuracy of each ADC according to CSI.Furthermore,by means of the minimum mean square quantization error criterion,a heuristic quantization Bit-Mapping algorithm is proposed.The algorithm effectively avoids the complexity of exhaustive search,and can obtain the quantization bit allocation result with better performnce after a few iterations.Our tests display efficiently and show that,compared with the benchmark system with different parameter combinations,the optimization algorithm can improve the energy efficiency in different degrees,especially in large-scale antenna systems.