Performance Analysis And Optimization Of Reconfigurable Intelligence Surface-assisted Communication Systems

5G commerce will increase the capacity of wireless communication network thousands of times in the future.Although traditional MIMO systems can meet the business needs such as wireless connection everywhere,they have to face the challenges of highly complex network,increasing energy consumption,high-cost hardware and so on.Intelligent Hypersurface(RIS)has attracted much attention due to its unique low cost and low energy consumption characteristics.It is expected to break through the uncontrollability of wireless channel and build a new intelligent programmable wireless environment.Since the birth of RIS,people have been studying its basic performance limitations and the impact of various system parameters on its realizable performance.In order to quantify the advantages and limitations of RIS in different network scenarios,a more accurate approximation analysis framework has been developed to perform performance analysis to show the advantages and limitations of RIS in different network situations.The main work of this thesis is as follows:First,in the single RIS-assisted wireless communication system model,the accuracy of the traversal reachability ratio(EAR)upper and lower bounds,interrupt probability(OP)closed solution derived under random phase shift is verified by Monte Carlo numerical simulation.To solve the problem of random phase shift restriction performance,an optimal phase shift design scheme is proposed to optimize RIS phase shift.The matlab simulation proves that this scheme compares with EAR under random phase shift.At low signal-to-noise ratio(SNR=10d B)when other configurations are the same,there is an increment of about 11.7%,while at high SNR=60d B,the increment is up to 40.6%,the advantage is obvious.The simulation also shows that OP under this scheme has obvious advantages over random phase shift,RIS-free system,traditional two-hop system,etc.It can decrease faster at low signal-to-noise ratio so that OP can be ignored to ensure quality of service(Qo S).Secondly,a location index scheme is proposed to make RIS obtain dynamic uniform linear array and uniform planar array(UPA)with high complexity.Based on this,an EAR closed-form solution is deduced.It is found that the EAR of the system can be controlled by configuring the size of the reflective element with a scale factor\alpha.Simulations show that an EAR of 31.138 times can be obtained with \alpha set to the same deployment condition.In addition,in order to study the spatial correlation of the configured reflectors,a spatial correlation matrix is proposed.The spatial correlation of the reflectors can be characterized by the simulation of the eigenvalue situation and the analysis of Crohneck model.Thirdly,in the multi-RIS system,the gamma distribution is used to approximate the channel distribution,and the statistical characteristics of the channel are deduced by the method of moments.The matlab function fitting simulation proves that the gamma approximation is accurate.Based on this,the EAR upper bound,lower bound,high signal-to-noise ratio slope and the approximate closed-form solution of OP for multi-RIS systems are derived.The simulation shows that the EAR upper bound of multi-RIS systems is higher than that of single RIS systems.