PaperlayerView exact revision on arXiv

Source-linked AI summary

Enabling Secure Wireless Communications for FARIS-Aided Systems

Hong-Bae Jeon, Yonghwi Kim, Hyung-Joo Moon, Kai-Kit Wong

arXiv:2608.27837v1eess.SP

TL;DR

The paper addresses secure downlink transmission when fixed or passive surfaces cannot adequately adapt to adverse wireless conditions. It jointly designs beamforming and FARIS reconfiguration through alternating optimization under practical power constraints, and reports consistently superior secrecy performance, including robustness to unfavorable eavesdropper geometries.

  • Problem

    Fixed-position RIS has limited adaptability, while passive fluid-RIS remains constrained in signal enhancement and robustness under severe path loss or blockage.

  • Method

    The paper jointly optimizes transmit beamforming, active reflection coefficients, and fluid port selection using an alternating optimization framework that decomposes the nonconvex design into tractable subproblems.

  • Results

    The proposed FARIS scheme consistently outperforms the benchmarks and maintains the highest secrecy rate across the tested range of Eve’s horizontal positions.

  • Takeaways & Limitations

    FARIS combines active amplification and fluid port selection to reshape legitimate and eavesdropping channels under practical power constraints, including unfavorable geometries.

Abstract

from arXiv · show

This paper investigates secure downlink transmission assisted by a fluid active reconfigurable intelligent surface (FARIS), which enables both active reflection and dynamic port selection, offering enhanced flexibility for physical-layer security. We formulate a secrecy rate maximization problem that jointly optimizes the transmit beamformer, active reflection coefficients, and fluid port configuration under practical power constraints. To efficiently handle the resulting highly nonconvex problem, we develop a tailored alternating optimization (AO) framework that decomposes the original joint design into tractable subproblems, where each admits an efficient solution while preserving the system constraints, enabling an effective joint optimization of beamforming and FARIS reconfiguration. Numerical results demonstrate that the proposed FARIS-assisted design consistently outperforms the benchmarks. The results further highlight the robustness of FARIS against unfavorable eavesdropping geometries, confirming its potential as a powerful enabler for secure communications in challenging environments.

I. INTRODUCTION

The introduction motivates FARIS as a flexible physical-layer security architecture that combines active reflection with dynamic port selection. It formulates joint beamforming and FARIS configuration as the basis for secrecy enhancement and develops an alternating optimization solution.

  • Wireless broadcast transmission exposes confidential signals to eavesdropping, motivating physical-layer security as an information-theoretic complement to cryptography.
  • Conventional RIS uses fixed locations, limiting adaptation to dynamic and adversarial wireless environments.
  • Fluid-RIS adds position reconfigurability and spatial degrees of freedom, but passive operation limits enhancement under severe path loss or blockage.
  • FARIS combines controllable amplification with dynamic port selection, enabling compensation for path loss and resilience against adverse channel conditions.
  • The paper jointly optimizes transmit beamforming, FARIS port selection, and reflection or amplification coefficients under transmit and radiated power limits.
  • An alternating optimization framework decomposes the coupled design into tractable subproblems, while numerical results show consistent improvement over benchmark schemes.

II. SYSTEM MODEL

The system model considers secure AP-to-user downlink transmission through a fluid active surface with correlated candidate ports, active reflection, and power constraints. Secrecy rate is maximized through joint beamforming, reflection, and port-selection optimization solved by alternating optimization.

  • An N-antenna AP communicates with a single-antenna Bob in the presence of a single-antenna Eve, assisted by an M-element FARIS.
  • Finite element spacing induces spatial correlation across the FARIS aperture, modeled by a Jakes correlation matrix.
  • Each FARIS element is a fluid-antenna port that is either active for controllable amplitude and phase modification or off with no reflection.
  • At each transmission block, only Mo < M ports are activated through a selection matrix whose columns are distinct canonical basis vectors.
  • The active reflection vector assigns each selected port a bounded gain and phase, and the effective FARIS operator incorporates port selection and spatial correlation.
  • Because FARIS introduces thermal noise and hardware power consumption, feasible designs must satisfy transmit, radiated, amplifier, and circuit power constraints.
  • The secrecy rate is the nonnegative difference between Bob’s and Eve’s log2 SINRs, and the design jointly optimizes beamforming, reflection, and port selection.
  • The resulting problem is a mixed-integer nonlinear program, addressed by alternating optimization over beamforming, reflection, and port selection.

A. Optimization of w: Transmit Beamforming

For fixed FARIS reflection and port configuration, the transmit-beamforming subproblem is transformed into a convex semidefinite program by lifting the beamformer and relaxing its rank constraint.

  • A. Optimization of w: Transmit Beamforming: For fixed (v, SMo), the beamforming objective uses the effective channels and noise powers ηk.The resulting constraints include transmit-power and FARIS-power limits.
  • A. Optimization of w: Transmit Beamforming: The lifted variable W = ww* converts the beamforming problem into a matrix optimization with trace constraints and rank(W) = 1.The constraints enforce both transmitter power and radiated FARIS power.
  • A. Optimization of w: Transmit Beamforming: Dropping the rank-one constraint yields a convex semidefinite relaxation that can be efficiently solved by CVX.The relaxed solution is converted back to a feasible beamformer using Gaussian randomization.

B. Update of v: Reflection and Amplification Coefficients

The reflection and amplification update lifts the coefficient vector into a matrix variable, applies fractional-objective transformations and majorization, and recovers feasible coefficients by randomization.

  • B. Update of v: Reflection and Amplification Coefficients: For fixed (w, SMo), the reflection update represents the FARIS-assisted terms and power expressions using lifted matrices involving v and V = vv*.The formulation uses identities involving K ⊙ V to express amplified signal power.
  • B. Update of v: Reflection and Amplification Coefficients: The lifted reflection problem maximizes the secrecy ratio subject to an active-power constraint and per-coefficient amplitude bounds.The lifted variable satisfies V = vv* before relaxation.
  • B. Update of v: Reflection and Amplification Coefficients: An augmented matrix X absorbs the linear cross-term, converting the channel expressions and amplitude constraints into trace and diagonal constraints.This produces γk(X) through tr(QkX) and tr(ĈkX).
  • B. Update of v: Reflection and Amplification Coefficients: The quadratic transform introduces auxiliary variables yk, while a first-order upper bound handles the eavesdropper term's difference-of-concave structure.The resulting surrogate is a tractable convex problem for fixed auxiliary variables and current X.
  • B. Update of v: Reflection and Amplification Coefficients: The relaxed solution X⋆ is converted into a feasible active reflection vector v⋆ through Gaussian randomization.Alternating updates of X, v, the eavesdropper surrogate, and {yk} produce a non-decreasing objective until convergence.

C. Update of SMo: FARIS Port Configuration

The port-configuration update uses cross-entropy optimization to search over fixed-size port subsets through sampled configurations, elite selection, and probability updates.

  • C. Update of SMo: FARIS Port Configuration: Given w and v, the cross-entropy method optimizes SMo by representing each configuration Γ as a subset of Mo ports.Each configuration satisfies Γ ⊆ {1, · · ·, M} and |Γ| = Mo.
  • C. Update of SMo: FARIS Port Configuration: At each iteration, K port configurations are sampled without replacement from a probability mass function over the M ports.The sampled subsets are evaluated by their secrecy-ratio objective.
  • C. Update of SMo: FARIS Port Configuration: The top Ke = ⌈ρK⌉ samples with the largest objective values form the elite set used to update port-selection probabilities.The update counts each port's occurrence in elite configurations.
  • C. Update of SMo: FARIS Port Configuration: Probability smoothing updates the sampling distribution, and the Mo ports with the largest converged probabilities define S⋆Mo.The CEO iterations stop when the probability distribution converges.

D. Overall Algorithm and Complexity Analysis

The complete algorithm alternates beamforming, reflection, and port updates, with a non-decreasing objective and complexity dominated by semidefinite programs, randomization, and CEO sampling.

  • D. Overall Algorithm and Complexity Analysis: Algorithm 1 combines the alternating updates of w, v, and SMo into the proposed optimization framework.The objective is non-decreasing in each iteration, guaranteeing convergence to a stationary point.
  • D. Overall Algorithm and Complexity Analysis: Updating w solves an N × N semidefinite program with complexity O(ISDPN^4.5), plus O(NrandN^2) when Gaussian randomization is used.ISDP denotes solver iterations and Nrand denotes the number of randomization trials.
  • D. Overall Algorithm and Complexity Analysis: Updating v solves a semidefinite problem with an (Mo + 1) × (Mo + 1) variable and complexity O(Icvx(Mo+1)^4.5), followed by Gaussian randomization.The reflection-update complexity depends on the convex-solver iterations and randomization candidates.
  • D. Overall Algorithm and Complexity Analysis: CEO port selection uses ICEO iterations and K samples per iteration, with each sample requiring O(MoN + M^2o) computation.Its total cost is O(ICEOK(MoN + M^2o)).
  • D. Overall Algorithm and Complexity Analysis: Across IAO alternating iterations, the total complexity combines the beamforming SDP, reflection SDP, Gaussian randomization, and CEO sampling costs.The overall approximation is expressed as IAO(ISDPN^4.5 + NrandN^2 + Icvx(Mo+1)^4.5 + ...).

IV. SIMULATION RESULTS

Simulations evaluate FARIS under varying transmit power, surface dimensions, candidate-port density, and eavesdropper location. FARIS consistently achieves the strongest secrecy performance, including unfavorable eavesdropping geometries.

  • Optimization Procedure: The FARIS-Aided Secrecy Maximization algorithm alternates beamformer optimization, reflection-vector recovery, and CEO-based port selection until objective convergence.The procedure updates w, v, and S_Mo sequentially within an alternating optimization loop.
  • Simulation Setup: The simulations use fixed AP, user, and FARIS placements with Rician AP–FARIS fading, Rayleigh other links, and 25 dB direct-path blockage loss.The FARIS has M = 64, M_o = 25, and g_max = 40 dB; the AP uses N = 4 antennas and P_AP = 25 dBm.
  • Transmit Power: As P_AP increases, FARIS provides the largest secrecy gain, while No-RIS produces the lowest secrecy rate.ARIS slightly outperforms FRIS because of active amplification; RIS trails FRIS because it lacks active gain and fluid elements.
  • Surface Elements: Increasing M_o monotonically improves FARIS and FRIS secrecy rates, with FARIS retaining the highest rate and a steeper growth trend than FRIS.ARIS eventually outperforms FRIS, RIS gains more modestly, and No-RIS remains flat.
  • Candidate Ports: Increasing M_x steadily improves FARIS secrecy performance because CEO exploits a denser candidate-port grid and finer spatial sampling within the same aperture.FRIS improves more moderately, whereas fixed-element ARIS and RIS remain nearly unchanged.
  • Eavesdropper Location: As Eve moves from x_Eve = 48 to 60 m, all schemes lose secrecy rate as Eve approaches Bob, but FARIS remains highest across the full range.FARIS adapts active gain control and port selection to mitigate leakage in unfavorable geometries; No-RIS is most vulnerable.

V. CONCLUSION

The paper concludes that FARIS expands physical-layer security design by jointly using active amplification and fluid port selection. Its secrecy gains arise from geometry-aware reconfiguration under practical power constraints, including unfavorable geometries.

  • Conclusion: FARIS expands PLS design by jointly leveraging active amplification and fluid port selection for geometry-aware secrecy shaping beyond RIS and FRIS.The conclusion identifies FARIS as a promising enabler for 6G physical-layer security.
  • Conclusion: Secrecy enhancement depends on selectively amplifying and spatially reconfiguring propagation paths, not solely on array gain.This enlarges the SINR disparity between the legitimate user and eavesdropper under practical power constraints.
  • Conclusion: Effective FARIS operation requires holistic co-design of beamforming and element configuration while accounting for port correlation and amplification-induced noise.
Loading 2608.27837v1…