RIS-Assisted JCAS: Communication & Sensing

• RIS-assisted JCAS systems are characterized by real-time passive tunability, enabling simultaneous data transmission and radar sensing over shared channels.
• They leverage joint beamforming and RIS configuration to optimize dual objectives like communication rate and sensing accuracy via iterative algorithms.
• Simulation and experimental studies demonstrate enhanced channel controllability, reduced self-interference, and improved detection accuracy with moderate RIS phase resolution.

A Reconfigurable Intelligent Surface (RIS)-assisted Joint Communication and Sensing (JCAS) system integrates programmable metasurface technology with wireless transceivers, enabling simultaneous radio sensing and data transmission over the same spectrum and physical infrastructure. This approach realizes a major paradigm for 6G, where the RIS’s passive and real-time tunability facilitates spatial, spectral, and information-function coupling between the communication and sensing domains, augmenting the channel’s controllability, improving accuracy, and providing a tractable optimization surface for multi-objective JCAS design. The fundamental architecture encompasses a multi-antenna base station, distributed or planar RIS(s), multiple communication endpoints, and at least one sensing target; operational modalities include monostatic, bistatic, and even full-duplex radar schemes.

1. System Architectures and Physical Channel Models

The canonical RIS-assisted JCAS model is governed by the following blockwise composition (Chepuri et al., 2022, Wijekoon et al., 13 Jan 2026, Liu et al., 2022):

• Base Station (BS): Equipped with $M_t$ transmit antennas (and optionally $M_r$ receive antennas for radar), forms active digital beamformers $\mathbf{w}$ and/or matrix $\mathbf{W}$ for data and probing.
• Reconfigurable Intelligent Surface (RIS): Array of $N$ passive elements, each programmable with unit-modulus phase shift $\phi_n$, described by the diagonal matrix $$\Phi = \mathrm{diag}\{e^{j\phi_1},\dots,e^{j\phi_N}\}$$.
• Communication Users and Sensing Targets: Single or multiple users, single or multiple targets; channels modelled as Rician/LoS for mmWave/THz, with BS–RIS $\mathbf{H}_{BR}$, RIS–User $h_{RU}$, and RIS–Target $h_{RT}$ (steering vectors or spatially correlated models).
• Mutual Coupling: Practical RIS layouts require consideration of mutual-element EM coupling, captured via the scattering matrix $S$ and coupling-aware reflection operator $\Theta=(\Phi^{-1}-S)^{-1}$ (Wijekoon et al., 13 Jan 2026).
• Signal Paths: JCAS is instantiated via dual/interleaved functional waveforms $x(t) = \mathbf{w}_c s_c(t) + \mathbf{w}_s s_s(t)$, enabling simultaneous data and probing.

The end-to-end effective communication channel (single-user) reads

$$h_{\mathrm{eff}}^T = h_{RU}^T \Theta H_{BR} + h_{BU}^T,$$

while the radar (sensing) channel is a composite function of BS/RIS steering, target location and optional direct/reflected paths, adopting monostatic ($$y_s = \gamma_1 A_1 w s + \gamma_2 H_{R \rightarrow BS} \Theta A_2 \Theta H_{BR} w s$$) or bistatic ($$y_s = \gamma_3 A_3 w s + \gamma_4 A_4 \Theta H_{BR} w s$$) observation models (Wijekoon et al., 13 Jan 2026, Chepuri et al., 2022).

2. Joint Design of Beamforming and RIS Configuration

RIS-assisted JCAS optimization is inherently multi-objective, designed to balance communication rate and sensing accuracy. The problem is formulated as (Wijekoon et al., 13 Jan 2026, Li et al., 6 Nov 2025, Liu et al., 2022):

$$\max_{\mathbf{w},\,\Phi} \;\; \alpha\,\mathrm{FI}(\mathbf{w},\Phi) + (1-\alpha)\,\mathrm{MI}(\mathbf{w},\Phi)$$

subject to

• Power: $\|\mathbf{w}\|^2 \leq P_{\mathrm{max}}$,
• Phase: $|\phi_n| = 1, \forall n$ (unit modulus),
• Additional constraints: e.g., sensing CRB/PEB, radar SINR, uplink rate thresholds.

Performance metrics include:

• Fisher Information (FI) for radar parameter estimation, e.g., angles,
• Mutual Information (MI) or communication rate, e.g., $$I(\mathbf{w},\Phi) = \log_2(1 + |h_{\mathrm{eff}}^T \mathbf{w}|^2/\sigma_c^2)$$,
• Beampattern Similarity or Chordal-Distance: For array-based designs, the similarity of transmit beampattern to a prescribed radar or sensing shape (Luo et al., 2022, Wang et al., 2022).

Key approaches:

• Alternating Optimization (AO): Iterative blockwise optimization of $\mathbf{w}$ (active beamforming) and $\Phi$ (passive phase) (Wijekoon et al., 13 Jan 2026, Chepuri et al., 2022, Xing et al., 2022).
  • AO with projected gradients or manifold updates for unit-modulus constraints ($\phi_n \to \phi_n/|\phi_n|$).
• Fractional Programming and Majorization-Minimization (MM): Fractional SINR and quartic beampattern constraints convexified via surrogate step, yielding QCQP or SDP subproblems (Luo et al., 2022, Li et al., 6 Nov 2025).
• Semidefinite Relaxation (SDR): Relaxing phase constraints to convex hull, randomization for recovery (Li et al., 6 Nov 2025, Xing et al., 2022).
• Low-Complexity Partitioned Design: Partition RIS phase into a communication-maximizing component and small sensing perturbation, with closed-form perturbation via regularized least-squares or beam power allocation (Li et al., 6 Nov 2025).

3. Full-Duplex JCAS, Self-Interference, and Robust Algorithms

Recent advanced schemes incorporate full-duplex (FD) operation at the BS, enabling simultaneous transmit/receive for uninterrupted radar sensing and uplink communication (Sheemar et al., 2023, Guo et al., 2023, Guo et al., 2023):

• FD Node: $M_b$ Tx/$N_b$ Rx antennas, interacts with near-field RIS (size $RC$), users and targets.
• Self-Interference (SI): Explicit model $H_{b,b}$, comprising near-field LoS and diffuse components, mitigated by passive RIS phase optimization.
• CRB-Constrained Solutions: Joint optimization minimizes radar power and target estimation CRB, subject to communication MSE or sum-rate maximization.
• WMMSE and Penalty-Dual-Decomposition (PDD): Advanced AO combines weighted MMSE for sum-rate and SI metrics, MM/PDD for nonconvex quadratic RIS subproblem with closed-form majorization and trust-region solutions.

Simulation and theory establish that joint SI-aware design of BS beamformers and RIS configuration can passively cancel over 10–15 dB of SI power, allowing near communication-only sum-rate at tight radar CRB thresholds with moderate RIS phase resolution (2–3 bits) (Sheemar et al., 2023, Guo et al., 2023, Guo et al., 2023).

4. Performance Trade-offs, Resource Allocation, and Complexity

JCAS design is governed by a fundamental trade-off between communication efficiency and sensing accuracy, quantified by Pareto frontiers between rate and CRB/FI (Wijekoon et al., 13 Jan 2026, Li et al., 6 Nov 2025, Chepuri et al., 2022, Liu et al., 2022):

• Tuning Trade-off: Weight $\alpha$ ($0 \leq \alpha \leq 1$) controls priority; increasing $\alpha$ favors sensing (FI/CRB improves, MI/rate decays), and vice versa.
• RIS Mutual Coupling: Denser sub-element spacing (e.g., $\lambda/4$) increases mutual coupling, which can be exploited for simultaneous improvement of MI and FI via subspace rotation and expansion (Wijekoon et al., 13 Jan 2026).
• RIS Size and Partitioning: More elements ($N\uparrow$) yield higher channel gain, broader beampattern control, lower CRB, and better multi-user/multi-target performance (Chepuri et al., 2022, Li et al., 6 Nov 2025, Sankar et al., 2021).
• Resource Allocation: Power, time-frequency scheduling, and proportional beam gain distribution can be optimized to favor desired sensing/comm. endpoints (Liu et al., 2022).
• Algorithm Complexity: SVD-based closed-form partitioning scales $\mathcal{O}(KN^2)$, SDR methods scale up to $\mathcal{O}(N^6)$, AO-based blockwise updates converge in 5–20 iterations at modest cost (Li et al., 6 Nov 2025, Guo et al., 2023, Chepuri et al., 2022).

5. Multi-User, Multi-Target, and Practical Deployment Scenarios

RIS-assisted JCAS architectures are scalable to multi-user and multi-target environments, as well as specialized scenarios (Sankar et al., 2021, Qian et al., 2023, Zhu et al., 2023, Chepuri et al., 2022):

• RIS Partitioning/Hierarchical Codebook: Adaptive allocation of RIS elements (sensing vs communication) enables rapid beam training for localization, target search and communications, reducing transmission overhead with modest spectral efficiency loss (cf. classic MIMO) (Sankar et al., 2021).
• Multi-Target Beampattern: Partitioning and perturbation of RIS phases enable proportional control of beam gains, enforcing fairness and precise target discrimination (Li et al., 6 Nov 2025).
• Sensing-Aided Beamforming: Use of local coarse-grained sensing (e.g., MUSIC/ESPRIT, device angle-of-arrival, echo location) enables low-overhead two-phase protocols for blind-zone positioning and throughput maximization (Qian et al., 2023).
• Forward-Only Sensing Control and SDR Validation: Prototype platforms with embedded active sensor arrays validate autonomous DOA/RF-ID recognition and robust interference suppression with over-the-air SDR experiments, showing 10–20 dB SINR enhancements and sub-degree DOA accuracy (Luo et al., 31 Mar 2025).
• Joint Imaging and Uplink JCAS: RIS phase optimization and Bayesian echo-decoupling (factor graph + sparse Bayesian learning) support simultaneous uplink communication and region-of-interest imaging, with discrete/continuous phase models and low-coherence matrix designs (Zhu et al., 2023).

6. Future Directions and Open Challenges

Major avenues for further RIS-JCAS research span architecture, theoretical limits, and robust algorithmics, as follows (Liu et al., 2022, Li et al., 2024, Wijekoon et al., 13 Jan 2026):

• Deployment Strategies: Optimal placement and adaptive control of multiple/distributed RISs for spatial diversity, coverage, and dynamic environments, including UAV-mounted RIS and near-field operation.
• Advanced Architectures: Integration of active RIS (with built-in amplification), hybrid analog/digital beamforming at BS, and RIS-embedded receiver arrays.
• Theoretical Analysis: Joint CRB-capacity region characterizations, tensor-based ambiguity-function estimation for high-dimensional echo data, and realistic physical channel modeling under multipath and EM coupling.
• AI-Driven Control: Deep learning and reinforcement learning for online phase adaptation, partial CSI handling, and non-differentiable hardware constraints.
• Cross-Layer ISCC Integration: Extending to joint sensing-communication-computation networks, with resource orchestration for data transmission, radar sensing, and edge computing.

RIS-assisted JCAS presents a coherent framework for dual-functional wireless networks, integrating beamforming, propagation control, resource allocation, and robust algorithmics, and is a foundational technology for high-efficiency, simultaneous communication and environment sensing in advanced 6G architectures.