Research On Key Technologys Of Reconfigurable Intelligent Surface Based On Software Radio Platform

As a key technology of 5G,large-scale multiple-input multiple-output(MIMO)uses spatial gain to greatly improve the performance of the receiving terminal.However,as the number of antenna arrays increases,the cost and power consumption of the system also increase rapidly.In recent years,Reconfigurable Reflective Surface(RIS)has realized low-cost intelligent reconstruction of wireless signal propagation environment by integrating a large number of low passive reflective units on a planar array and has become an efficient and novel wireless communication technology.However,most research of the smart reflective surface beamforming is direct theoretical adjustment and control the phase information of the smart reflective surface arrays.How to complete the beamforming of the smart reflective surface in the actual platform,and the research on the wireless communication model based on the smart reflective surface have become one of the important challenges in realizing the reconstruction of the wireless network environment.On the other hand,the software radio platform is widely used in wireless communication by virtue of its high flexibility and software control.In this thesis,a systematic analysis of the intelligent reflective surface based on the software radio platform is carried out.Meanwhile,a wireless communication platform of the intelligent reflective surface is built and a study of its function and performance is carried out.The thesis first considers the wireless communication system model based on the intelligent reflective surface,and analyzes the physical layer structure and system flow in detail.The thesis has done in-depth research on the frame structure design of wireless communication system,synchronization based on PN sequence,cyclic redundancy check and channel model,and studied the beamforming system architecture based on intelligent reflective surface according to its model.The software radio platform is implemented based on field-programmable gate array(FPGA),with high flexibility and operability,which is widely used in wireless communication test scenarios due to its low cost and other advantages.We use the VIVADO+FPGA platform,establishes a wireless communication model through FPGA,and simulates and tests the entire system communication process in the VIVADO compilation environment.Then,the thesis introduces and analyzes the wireless communication hardware platform model based on intelligent reflective surface in detail.According to the performance analysis of each hardware platform module,the overall design of the wireless communication platform is given.We introduce the board of FPGA platform,AD9361 chip and radio frequency link,etc.,and test the communication frame error rate and constellation diagram of our hardware part.According to the characteristics of intelligent reflective surface beamforming,the thesis uses FPGA platform to control the azimuth and horizontal angle of intelligent reflective surface and analyzes the scanning performance and transmission rate based on the angle domain.Meanwhile,the uart protocol is used to control the horizontal angle and azimuth angle of the intelligent reflecting surface,and the signal transmission spectrum and anteena main lobe performance based on the intelligent reflecting surface are tested.After completing the construction of the entire system platform,the thesis tested the signal transmission performance of the wireless communication system and the video transmission function under the user datagram protocol(UDP).Finally,the indoor positioning and adaptive beam alignment algorithms of intelligent reflective surface platforms based on deep learning algorithms are researched.Through the analysis of the specific technical parameters of the intelligent reflective surface platform based on the deep learning algorithm,a new structure of energy array is proposed and the indoor positioning function using the deep learning algorithm is completed.At the same time,according to high-resolution energy array and high-resolution energy array,the accuracy of indoor positioning under different accuracy is quantitatively analyzed.Furthermore,the thesis combines back-propagation(BP)neural network and receiver cyclic redundancy check detection to realize the adaptive beam alignment function.The test results show that the intelligent reflective surface platform based on the deep learning algorithm designed in this thesis meets the expected requirements in indoor positioning.