Intelligent reflecting surface (IRS) offers a promising and revolutionizing solution to enhance spectrum and energy efficiency by leveraging massive low-cost passive elements that are able to reflect the incident signal with adjustable amplitude and phase. Different from the prior works focusing on MISO scenario, this paper considers the IRS-aided single-user mmWave massive MIMO systems. We formulate and solve new problems to maximize the spectral efficiency of the systems by jointly optimizing the active beamforming at both the base station and user equipment, as well as the passive beamforming at the IRS. To tackle this problem, we first give the closed-form solutions of active beamforming and passive beamforming in MISO scenario. Then, we divide the optimization problem into three sub-problems, and perform alternate optimization solutions in MIMO scenario. In addition, we analyze the performance of the proposed algorithms in more practical scenarios where each element of IRS can only take discrete phase-shift values and base station has imperfect CSI. Simulation results show that IRS-aided MIMO systems effectively improves the spectral efficiency and also verifies the superior performance of our proposed algorithms.

智能反射面 (IRS) 提供了一种有前途的革命性解决方案，通过利用能够以可调幅度和相位反射入射信号的大量低成本无源元件来提高频谱和能源效率。与之前关注 MISO 场景的工作不同，本文考虑了 IRS 辅助的单用户毫米波大规模 MIMO 系统。我们通过联合优化基站和用户设备的有源波束成形以及 IRS 的无源波束成形，制定并解决新问题，以最大限度地提高系统的频谱效率。为了解决这个问题，我们首先给出了MISO场景下主动波束成形和被动波束成形的封闭形式解决方案。然后，我们将优化问题分解为三个子问题，并在 MIMO 场景中执行替代优化解决方案。此外，我们在更实际的场景中分析了所提出算法的性能，其中 IRS 的每个元素只能采用离散的相移值，并且基站具有不完善的 CSI。仿真结果表明，IRS 辅助的 MIMO 系统有效地提高了频谱效率，也验证了我们提出的算法的优越性能。