Research On Resource Allocation Technologies For Intelligent Reflecting Surface Aided Wireless Systems

Intelligent reflecting surface(IRS)originates from software-defined metamaterials and consists of multiple reflecting elements,where each element can be controlled via a smart software controller to induce a certain phase shift on the incident signal,thus collaboratively enhancing the wireless channel quality between transmitter and receiver.IRS with low power consumption and lightweight characteristics can enable the construction of smart and green radio environments for the future wireless networks.In the current researches on IRS,the performance indicator of system is single and instantaneous,which will pose challenge to the management of resource allocation with the increasing communication requirements as well as the complex and changing network environment.Therefore,this thesis studies the long-term performance of the time-varying uplink and downlink communication systems,and the pricing problem of IRS collaborative communication for multiple objectives of network optimization.The structure of this thesis is introduced as follows:In Chapter 1,the research background and significance of the IRS-aided communication system is introduced,the research status of IRS is summarized,the main work of this thesis are described.In Chapter 2,dynamic resource allocation technology for IRS-aided downlink communications is investigated.The goal is to decrease the time-averaged power consumption of this dynamic system,via optimizing the active beamforming at the base station(BS)and reflecting coefficient at IRS.By applying Lyapunov concept-based driftplus-penalty(DPP)algorithm,the long-term optimization problem is transformed into short-term sub-problems related to each other.Subsequently,the Lagrange dual transformation and quadratic transformation are applied to derive the solutions for active beamforming and reflecting coefficient in the closed form.Simulation data demonstrate the effectiveness and convergence of the proposed algorithm.In Chapter 3,dynamic resource allocation technology for IRS-aided uplink communications is investigated.To maximize the time-averaged sum-rate of the dynamic system,the power allocation at users and the reflecting coefficient at IRS are jointly optimized.Through theoretical analysis,the closed-form solutions of power and reflecting coefficient can be obtained.Finally,simulation results illustrate that the proposed algorithm can allocate resources reasonably and improve the fairness of users.In Chapter 4,the collaborative resource allocation technology of IRS-aided wireless networks is studied.According to Stackelberg game model,the process of IRS pricing and collaborative resource allocation in multi-entity,multi-constraint and multi-objective networks is modeled,and the effectiveness and importance of IRS pricing are proved by simulation results.In Chapter 5,the thesis is summarized and the future research direction and work are given.