Capacity and Optimal Resource Allocation for IRS-assisted Multi-user Communication Systems
share
This paper investigates intelligent reflecting surface (IRS)-assisted multi-user wireless communication systems, where an access point (AP) sends independent information to multiple users with the aid of one IRS. To determine the fundamental rate limits of these systems, we jointly optimize the IRS discrete phase-shift matrix and resource allocation for both orthogonal multiple access (OMA) and capacity-achieving non-orthogonal multiple access (NOMA) transmission schemes. The ergodic and delay-limited rate/capacity regions are characterized by invoking rate-profile techniques. For ergodic rate/capacity regions, as the formulated non-convex average sum rate maximization problems satisfy the time-sharing condition, we derive the optimal solutions for both transmission schemes using the Lagrange duality method. For OMA transmissions, the optimal solution unveils that each user should be alternatively served with its effective channel gain maximized by dynamically adjusting the IRS phase. For the NOMA transmission, the optimal transmission strategy still follows the alternative transmission scheme but among different user groups. For delay-limited rate/capacity regions, the instantaneous sum rate maximization problem is decomposed into a series of resource allocation subproblems. For OMA transmissions, we derive the optimal resource allocation by utilizing the Lagrange duality method. For NOMA transmissions, we show that the subproblem can be optimally solved by successively checking problem feasibility with the bisection method. The optimal solutions indicate that the IRS phase should be fixed throughout the transmission to maintain the instantaneous achievable rate of all users. We further propose a Hadamard codebook based scheme for IRS phase adjustment, which serves as a lower bound on the optimal performance gains. Finally, numerical results are provided to verify our proposed designs.
READ FULL TEXT