标题： 基于IRS的多用户集成感知与通信的Cramér-Rao界最小化 显示英文标题

标题： Cramér-Rao Bound Minimization for IRS-Enabled Multiuser Integrated Sensing and Communications

摘要： 本文研究了一个智能反射表面（IRS）启用的多用户集成感知与通信（ISAC）系统，该系统包括一个多天线基站（BS）、一个IRS、多个单天线通信用户（CUs）以及一个位于BS非视距（NLoS）区域的目标。 IRS被部署以不仅协助BS到CUs的通信，还基于BS-IRS-目标-IRS-BS链路的回波信号实现BS对NLoS目标的感知。 我们考虑两种类型的目标，即扩展目标和点目标，分别要求BS估计完整的目标响应矩阵和相对于IRS的目标到达方向（DoA）。 为了为感知提供完整的自由度，我们考虑BS除了通信信号外还发送专用的感知信号。 相应地，我们模型了两种类型的CU接收器，即Type-I和Type-II CU接收器，分别不具备和具备取消感知信号干扰的能力。 在每种设置下，我们联合优化BS的发射波束成形和IRS的反射波束成形，以最小化目标估计的Cramér-Rao界（CRB），同时满足CUs的最小信干噪比（SINR）约束和BS的最大发射功率约束。 我们通过使用交替优化、半定松弛和连续凸近似技术，提出了有效的算法来解决高度非凸的SINR约束CRB最小化问题。 数值结果表明，所提出的设计比其他基准方案实现了更低的估计CRB，并且当CUs的数量大于一时，Type-II CU接收器的感知信号干扰取消是有益的。 显示英文摘要

摘要： This paper investigates an intelligent reflecting surface (IRS) enabled multiuser integrated sensing and communications (ISAC) system, which consists of one multi-antenna base station (BS), one IRS, multiple single-antenna communication users (CUs), and one target at the non-line-of-sight (NLoS) region of the BS. The IRS is deployed to not only assist the communication from the BS to the CUs, but also enable the BS's NLoS target sensing based on the echo signals from the BS-IRS-target-IRS-BS link. We consider two types of targets, namely the extended and point targets, for which the BS aims to estimate the complete target response matrix and the target direction-of-arrival (DoA) with respect to the IRS, respectively. To provide full degrees of freedom for sensing, we consider that the BS sends dedicated sensing signals in addition to the communication signals. Accordingly, we model two types of CU receivers, namely Type-I and Type-II CU receivers, which do not have and have the capability of canceling the interference from the sensing signals, respectively. Under each setup, we jointly optimize the transmit beamforming at the BS and the reflective beamforming at the IRS to minimize the Cram\'er-Rao bound (CRB) for target estimation, subject to the minimum signal-to-interference-plus-noise ratio (SINR) constraints at the CUs and the maximum transmit power constraint at the BS. We present efficient algorithms to solve the highly non-convex SINR-constrained CRB minimization problems, by using the techniques of alternating optimization, semi-definite relaxation, and successive convex approximation. Numerical results show that the proposed design achieves lower estimation CRB than other benchmark schemes, and the sensing signal interference cancellation at Type-II CU receivers is beneficial when the number of CUs is greater than one.