This paper studies the harvested energy maximization in an intelligent reflecting surface (IRS) aided multiuser multiple-input multiple-output (MIMO) simultaneous wireless information and power transfer (SWIPT) system in which the users can exploit power-splitting (PS) techniques for information decoding and energy harvesting (EH) simultaneously. A design problem is mathematically formulated as a joint optimization problem of transmit precoding (TPC) matrices at the base station (BS), the phase shifts at the IRS, and EH PS factors at the users to maximize the total harvested energy at the users while guaranteeing the quality of service requirements in terms of minimum achievable user rates and minimum harvested energy at each user. Considering the nonlinear EH models of the practical EH circuits at the users and unit-modulus constraints of the IRS phase shifters, the design problem becomes a highly nonlinear and non-convex optimization problem of the coupled matrix variables. To tackle the mathematical challenges in seeking the optimal solutions, we adopt alternating optimization to decompose the original design problem into two subproblems. The first subproblem is to jointly determine the TPC matrices at the BS and PS factors at the users while the second subproblem is to obtain the phase shifts at the IRS. Since each subproblem is still a non-convex optimization problem, we employ the minorization-maximization method to devise lower bound concave functions for the nonlinear EH and user rate functions and find the appropriate convex inner sets of the feasible sets to transform the optimization problems into convex ones. Verified by simulation results that the convergence of the proposed iterative algorithm is guaranteed and the total harvested energy is significantly improved as the multiuser MIMO SWIPT system is aided by the IRS with optimal phase shifts.

本文研究了智能反射面 (IRS) 辅助多用户多输入多输出 (MIMO) 同步无线信息和功率传输 (SWIPT) 系统中的收集能量最大化，其中用户可以利用功率分配 (PS) 技术信息解码和能量收集（EH）同时进行。设计问题在数学上被表述为基站 (BS) 处的传输预编码 (TPC) 矩阵、IRS 处的相移和用户处的 EH PS 因子的联合优化问题，以最大化用户处的总收集能量，同时在每个用户的最低可实现用户费率和最低收集能量方面保证服务质量要求。考虑到实际 EH 电路在用户端的非线性 EH 模型和 IRS 移相器的单位模数约束，设计问题变成了耦合矩阵变量的高度非线性和非凸优化问题。为了解决寻求最优解的数学挑战，我们采用交替优化将原始设计问题分解为两个子问题。第一个子问题是联合确定 BS 处的 TPC 矩阵和用户处的 PS 因子，而第二个子问题是获得 IRS 处的相移。由于每个子问题仍然是一个非凸优化问题，我们采用最小化-最大化方法为非线性 EH 和用户速率函数设计下界凹函数，并找到可行集的适当凸内集，将优化问题转化为凸问题。通过仿真结果验证，由于具有最佳相移的 IRS 辅助多用户 MIMO SWIPT 系统，因此保证了所提出的迭代算法的收敛性并且总收集能量显着提高。