Source-linked AI summary
Joint Beamforming Design for RIS-Assisted Integrated Sensing and Communication Systems
Honghao Luo, Rang Liu, Ming Li, Yang Liu, Qian Liu
TL;DR
The paper addresses spectrum congestion in ISAC by studying RIS-assisted joint communication and radar design. It jointly optimizes BS transmit beamforming and RIS reflections using FP, MM, and manifold methods, and simulations show improved sum-rate and beampattern behavior with RIS and optimized design. The study’s stated future scope includes clutter, blocked targets, performance trade-offs, and fairness.
Problem
ISAC must balance communication and radar sensing under spectrum scarcity, while active beamforming alone has limited gains under severe channel degradation.
Method
The paper jointly optimizes BS transmit beamforming and RIS reflection coefficients using FP, MM, and manifold optimization within an alternating algorithm.
Results
Simulations show that RIS deployment improves communication and sensing performance, while the proposed joint design outperforms random RIS reflection design in the reported comparisons.
Takeaways & Limitations
Optimizing a RIS jointly with BS beamforming is beneficial for RIS-assisted ISAC, especially as the number of reflecting elements increases.
Abstract
from arXiv · showhide
Integrated sensing and communication (ISAC) has been envisioned as a promising technology to tackle the spectrum congestion problem for future networks. In this correspondence, we investigate to deploy a reconfigurable intelligent surface (RIS) in an ISAC system for achieving better performance. In particular, a multi-antenna base station (BS) simultaneously serves multiple single-antenna users with the assistance of a RIS and detects potential targets. The active beamforming of the BS and the passive beamforming of the RIS are jointly optimized to maximize the achievable sum-rate of the communication users while satisfying the constraint of beampattern similarity for radar sensing, the restriction of the RIS, and the transmit power budget. An efficient alternating algorithm based on the fractional programming (FP), majorization-minimization (MM), and manifold optimization methods is developed to convert the resulting non-convex optimization problem into two solvable sub-problems and iteratively solve them. Simulation studies illustrate the advancement of deploying RIS in ISAC systems and the effectiveness of the proposed algorithm.
I. INTRODUCTION
ISAC addresses spectrum congestion by sharing a platform for communication and radar sensing, while RIS adds propagation-control degrees of freedom when active beamforming faces severe channel degradation. The paper therefore jointly designs BS and RIS beamforming to improve communication and sensing performance.
- Motivation: ISAC uses a shared dual-functional waveform to perform communication and radar sensing simultaneously.This shared platform improves spectrum and hardware efficiencies.
- Motivation: Active transmit beamforming must balance communication and radar sensing, but its gains are limited under severe channel degradation.
- RIS for ISAC: RIS provides additional propagation-control degrees of freedom by intelligently tailoring the wireless environment.Its reflecting elements can create additional non-line-of-sight links.
- Paper objective: The paper jointly optimizes BS transmit beamforming and RIS reflection coefficients to maximize multiuser sum-rate under radar, RIS, and power constraints.The proposed solution combines fractional programming, majorization-minimization, and manifold methods to solve two iterative subproblems.
II. SYSTEM MODEL AND PROBLEM FORMULATION
The system uses a multi-antenna BS, an N-element RIS, multiple single-antenna users, and point-like targets, with the RIS primarily assisting downlink communication. The formulation jointly designs communication and radar transmission with RIS reflections to maximize sum-rate while matching a desired radar beampattern and respecting power and unit-modulus constraints.
- System model: The considered system has an M-antenna BS, K single-antenna users, an N-element RIS, and T point-like targets.The RIS is deployed near users, while weak RIS-reflected radar echoes from distant low-altitude targets are neglected.
- Transmit signal: The BS transmits separate communication and radar beamforming components, combined into an overall beamforming matrix W.The radar probing signal is introduced to provide additional sensing degrees of freedom.
- User reception: The RIS reflection matrix is formed from unit-modulus coefficients, and the received-user model includes BS–RIS, RIS–user, and direct BS–user channels.
- Radar sensing: Radar sensing is evaluated by comparing the designed transmit beampattern with an ideal beampattern across sampled directions.The ideal pattern is scaled by α before computing the beampattern mean squared error.
- Problem formulation: The optimization maximizes achievable user sum-rate subject to beampattern similarity, total transmit power, and RIS phase-shift constraints.The problem is non-convex because of fractional-log terms, a quartic constraint, and unit-modulus variables.
A. FP-based Transformation
The paper applies fractional programming to reformulate the sum-rate objective by introducing auxiliary variables and removing ratio terms from the logarithmic expression. Because the transformed objective still couples BS and RIS variables, block coordinate descent is used to optimize them iteratively.
- FP reformulation: Fractional programming introduces auxiliary variables c_k to extract each SINR ratio from the logarithmic sum-rate objective.The resulting reformulation is based on a Lagrangian-dual transformation.
- FP reformulation: A quadratic transform introduces auxiliary variables g_k to further handle the remaining multiple-ratio term.
- Transformed objective: After the transformations, the objective becomes more tractable with respect to the BS beamforming matrix W and RIS coefficients φ.
- Alternating optimization: Block coordinate descent alternates between optimizing W and φ because the transformed objective still couples both variable blocks.
B. Optimize W with Given φ
With φ fixed, the beamforming update replaces the quartic beampattern MSE constraint with MM-based convex upper bounds, yielding a convex optimization problem for W.
- B. Optimize W with Given φ: With φ fixed, the optimization problem for W is formulated as a subproblem.
- B. Optimize W with Given φ: The beampattern MSE is reduced to a univariate function E(W) after optimizing the auxiliary variable α.
- B. Optimize W with Given φ: MM constructs a tractable surrogate for E(W) at the current iterate Wt that serves as an upper bound.
- B. Optimize W with Given φ: The matrix Bt is decomposed into positive- and negative-semidefinite parts, separating the quadratic term into convex and concave components.
- B. Optimize W with Given φ: A first-order Taylor expansion upper-bounds the concave component, producing a convex surrogate for the beampattern MSE.
- B. Optimize W with Given φ: The resulting reformulated problem is convex and can be solved using existing convex optimization solvers such as CVX.
C. Optimize φ with Given W
With W fixed, the RIS reflection-coefficient update addresses the non-convex unit-modulus constraint using a manifold-based optimization method.
- C. Optimize φ with Given W: With W fixed, the optimization problem for the RIS reflection coefficients φ is formulated as a subproblem.
- C. Optimize φ with Given W: The RIS subproblem contains a smooth objective and a non-convex unit-modulus constraint.
- C. Optimize φ with Given W: The unit-modulus constraint forms a complex circle Riemannian manifold.
- C. Optimize φ with Given W: The paper solves the RIS subproblem with a Riemannian conjugate gradient algorithm after deriving the Riemannian gradient from the Euclidean gradient.
IV. SIMULATION RESULTS
Simulations evaluate the proposed joint beamforming design in a multi-user RIS-assisted ISAC setting. The results compare RIS configurations, transmit-power behavior, RIS size, and transmit beampatterns against specified benchmarks.
- IV. SIMULATION RESULTS: The simulation uses an 8-antenna BS, 4 single-antenna users, and an N-element RIS under distance-dependent path loss and Rician fading.User noise power is set to −80 dBm, with BS–RIS and RIS–user distances of 50 m and 4 m.
- IV. SIMULATION RESULTS: The desired beampattern contains three target directions at −35°, 0°, and 35°, each with a 10° beam width.
- IV. SIMULATION RESULTS: In Fig. 2, the proposed RIS scheme, random-phase RIS, no-RIS, and communication-only systems are compared by achievable sum-rate versus transmit power.
- IV. SIMULATION RESULTS: RIS-assisted scenarios achieve better performance than the no-RIS scenario because RIS introduces additional NLoS links that enhance downlink communication.
- IV. SIMULATION RESULTS: The proposed RIS scheme improves more than the random-phase RIS scheme, while ISAC remains separated from communication-only performance because of the communication–radar trade-off.
- IV. SIMULATION RESULTS: With more RIS elements, achievable sum-rate increases and the proposed scheme’s gap over random-phase RIS becomes larger.
- IV. SIMULATION RESULTS: The proposed transmit beampattern has higher sidelobes than the radar-only benchmark but a more preferable shape than the no-RIS scheme.
V. CONCLUSION
The study jointly optimizes BS transmit beamforming and RIS reflection coefficients for RIS-assisted ISAC, using an alternating algorithm while enforcing sensing, power, and RIS constraints. Future work will address clutter, blocked targets, performance trade-offs, and fairness.
- V. CONCLUSION: The proposed design jointly optimizes transmit beamforming and RIS reflection coefficients to maximize sum-rate under beampattern, power, and RIS phase-shift constraints.The optimization targets communication performance while preserving radar-sensing requirements.
- V. CONCLUSION: An efficient alternating algorithm was developed to solve the resulting joint optimization problem.
- V. CONCLUSION: Future work will consider clutter, blocked targets, performance trade-offs, and fairness.