Sensing for Communication: RIS-Assisted ISAC Coordination Gain Enhancement With Imperfect CSI

TL;DR

This work tackles coordination gains in ISAC under imperfect CSI by proposing a RIS-assisted SACE (R-SACE) framework that operates an ISAC period followed by a PC period. By jointly optimizing time allocation, RIS phase shifts, and BS beamforming, and transforming probabilistic CSI constraints into a non-probabilistic form, the authors develop a fixed-point iterative (FPI) algorithm using quadratic-transform techniques. The results show AST gains over RIS-only and sensing-only baselines, demonstrating RIS-enabled interference suppression and sensing-informed coordination that improve systemic throughput in practical fading environments. The approach offers a principled, computationally efficient pathway to enhance ISAC coordination gains in 6G settings with imperfect CSI, with potential extensions to broader RIS configurations and multi-user scenarios.

Abstract

Integrated sensing and communication (ISAC) has the potential to facilitate coordination gains from mutual assistance between sensing and communication (S&C), especially sensing-aided communication enhancement (SACE). Reconfigurable intelligent surface (RIS) is another potential technique for achieving resource-efficient communication enhancement. Therefore, this paper proposes an innovative RIS-assisted SACE (R-SACE) mechanism with the goal of improving the systemic communication performance of the ISAC system in practical scenarios where the channel status information (CSI) is imperfectly known. In the proposed R-SACE mechanism, a dual-functional base station (BS) provides downlink communication services to both the communication user and the dynamically changing target that is detected using the communication signals. RIS assists in both sensing and communications of the BS. A typical scenario is investigated in which either or both the direct and RIS-assisted reflected communication links are available depending on sensing results. The average systemic throughput (AST) over the entire timeline of the R-SACE mechanism is maximized by jointly optimizing both temporal and spatial resources under the probabilistic constraint and the sensing performance, transmission power, and communication interference constraints. The non-convex probabilistic mixed optimization problem is transformed and then solved by the proposed fixed-point iterative (FPI) algorithm. Simulation results demonstrate that the proposed FPI algorithm and R-SACE mechanism outperform the baseline algorithms and communication enhancement mechanisms in achieving higher systemic communication performance.

Figures (12)

• Figure 1: System model of the proposed R-SACE mechanism.
• Figure 2: Illustration of the time protocol of the proposed R-SACE mechanism.
• Figure 3: A comparison of the convergence iterations and running time of different algorithms. (a) Convergence iteration. (b) Running time.
• Figure 4: The AST obtained with perfect and imperfect CSIs changing with $\tau_1$.
• Figure 5: The sensing performance obtained at ISAC period changing with $\sigma^2$.