Performance Research On Massive MIMO Systems Based On Intelligent Reflecting Surfaces

With the development of smart wireless devices,the requirements for data transmission rates continue to increase,and people’s commercial demand for the 5th generation mobile communication technology(5G)is in full swing.As one of the key technologies of 5G communication,large-scale multiple-input multiple-output(MIMO)is equipped with large-scale antenna arrays at the transmitting end and the receiving end to increase the array gain,thereby improving the user’s communication quality.However,when users have poor channel conditions or want to achieve higher system performance,they can use the assistance of Intelligent Reflecting Surface(IRS)to intelligently control each component on the IRS and adjust the reflected signal to achieve higher system performance.Therefore,this thesis combines massive MIMO technology with IRS,and conducts research on user security performance and fairness performance.The first part studies the security performance of massive MIMO systems.First,a solution that only optimizes the precoding vector at the transmitter is studied.With the help of the CVX toolkit,iteratively optimizes the precoding vector at the transmitter according to the Dinkelbach algorithm,and analyzes the achieved security rate through simulation.The simulation results show that optimizing the precoding vector at the transmitting end can effectively improve the security rate of the target user compared to the maximum ratio transmission.Then,after introducing the artificial noise vector,the algorithm of jointly optimizing the precoding vector and artificial noise vector at the transmitting end is discussed,and the maximum security rate of the target user and the minimum power of the transmitting end are simulated respectively.The simulation results show that the algorithm of jointly optimizing the precoding vector and artificial noise vector at the transmitting end can further improve the privacy rate of users.Among them,artificial noise is more obvious in scenarios where there are more eavesdropping users and fewer antennas at the transmitting end.Finally,considering the system combining massive MIMO technology and IRS,two optimization problems of maximizing the security rate of the target user and minimizing the transmit power of the transmitter are proposed respectively based on alternate optimization algorithms.The simulation results show that after adopting IRS,the target user can achieve more than double the security performance.The second part studies the fairness performance of massive MIMO-NOMA system.IRS is applied to the massive MIMO-NOMA system based on hybrid precoding,and an IRS-assisted alternate optimization algorithm is proposed.Specifically,first,a user is selected as the cluster head for each beam according to the cluster head selection algorithm,and the simulated precoding is designed according to the selected cluster head.After that,the users are grouped according to the correlation of the users’ equivalent channels.Then,the user with the strongest equivalent channel gain in each beam is selected to design the digital precoding.Finally,by alternately optimizing the power allocation and phase shift matrix,the user with the weakest spectrum efficiency among users is maximized on the premise of satisfying user fairness.The simulation results show that the proposed IRS-assisted alternate optimization algorithm can significantly improve the spectrum efficiency of the weakest user compared with the massive MIMO-NOMA system scheme without IRS assistance.