Dual Target-Mounted RISs-Assisted ISAC Against Eavesdropping and Malicious Interference

1. Problem Statement

The paper addresses a critical security gap in Integrated Sensing and Communication (ISAC) systems within 6G networks. While ISAC unifies radar sensing and data communication, it faces dual security threats from Unmanned Aerial Vehicles (UAVs):

1. Eavesdropping: A hostile UAV target attempts to intercept communication signals between a Base Station (BS) and legitimate users.
2. Malicious Interference: The same hostile UAV is equipped with a Reconfigurable Intelligent Surface (RIS) that launches random or optimized interference attacks to disrupt legitimate user communications.

The Core Challenge:
Existing literature often treats sensing targets merely as potential eavesdroppers or uses external RISs to aid communication. This paper tackles a more complex scenario where:

• One UAV is legitimate (carrying a friendly RIS to enhance communication).
• The other UAV is malicious (carrying a hostile RIS to jam communications and eavesdrop).
• The BS has imperfect Channel State Information (CSI) regarding the eavesdropper's location and channel, while the malicious RIS may possess perfect knowledge of the BS's beamforming and CSI in a worst-case scenario.

2. Methodology

The authors propose a Dual Target-Mounted RIS-assisted ISAC framework and develop a Semi-Definite Relaxation (SDR)-based two-stage optimization algorithm to maximize the system's secrecy rate while maintaining sensing performance.

System Model

• Transmitter: A BS with a Uniform Linear Array (ULA) transmitting a unified signal containing both communication data ($x_c$) and sensing waveforms ($x_s$).
• Targets:
  • Legitimate UAV (L-UAV): Equipped with a Legitimate RIS ($\Omega_L$) to reflect signals constructively to users.
  • Eavesdropper UAV (E-UAV): Equipped with a Malicious RIS ($\Omega_M$) to reflect signals destructively (jamming) and intercept data.
• Channel Model: Rician fading is used for communication links; sensing links use Line-of-Sight (LoS) models with angle estimation errors.

Optimization Strategy

The problem is formulated to maximize the Sum Secrecy Rate (SR) of legitimate users subject to power constraints and sensing SINR requirements. Due to the non-convex nature of the problem, it is decomposed into two stages:

1. Stage 1: Transmit Beamforming Optimization ($W$)

   • Goal: Optimize the BS transmit beamforming matrix to maximize the minimum sensing SINR for both UAV targets while satisfying power constraints.
   • Technique: The problem is reformulated using SDR. The BS prioritizes the legitimate UAV but ensures a minimum sensing threshold for the eavesdropper to maintain tracking capability.
   • Solver: Solved using the CVX toolbox (SeDuMi solver) followed by Eigenvalue Decomposition (EVD) to recover the beamforming matrix.

2. Stage 2: Legitimate RIS Phase Optimization ($\Omega_L$)

   • Goal: Optimize the phase shifts of the legitimate RIS to maximize the achievable rate of legitimate users, thereby increasing the secrecy rate.
   • Technique: The problem is converted into a Quadratically Constrained Quadratic Programming (QCQP) problem. SDR is applied to relax the unit-modulus constraints.
   • Solver: Solved via CVX, with Gaussian approximation or EVD used to ensure a rank-one solution.

3. Worst-Case Analysis

   • The paper also models a scenario where the malicious RIS has perfect knowledge of the BS beamforming and CSI. An additional optimization problem is formulated to maximize the interference power at the users, serving as a robustness benchmark.

3. Key Contributions

• Novel System Architecture: Proposes a unique ISAC framework where RISs are mounted directly on the sensing targets (one friendly, one hostile), creating a dual-threat environment.
• Dual-Threat Mitigation: Simultaneously addresses eavesdropping (signal interception) and active jamming (malicious RIS interference) under imperfect CSI conditions.
• Two-Stage SDR Algorithm: Develops an efficient algorithm to jointly optimize BS beamforming and legitimate RIS phase shifts, decoupling the sensing and communication objectives.
• Comprehensive Performance Metrics: Evaluates the system not just on communication secrecy rate, but also on sensing performance using Signal-to-Interference-plus-Noise Ratio (SINR) and the Cramer-Rao Bound (CRB) for Angle-of-Departure (AoD) estimation.
• Robustness Validation: Analyzes the system under a "worst-case" scenario where the adversary has perfect knowledge, demonstrating the algorithm's resilience.

4. Simulation Results

Extensive Monte Carlo simulations were conducted with the following key findings:

• Secrecy Rate (SR): The proposed scheme significantly improves the sum secrecy rate compared to benchmarks. Increasing the number of elements on the Legitimate RIS ($N_L$) leads to substantial gains due to higher passive beamforming resolution.
• Impact of Malicious RIS: Increasing the size of the Malicious RIS ($N_M$) degrades the secrecy rate, but the proposed algorithm maintains positive secrecy rates even under large $N_M$ and low transmit power.
• Worst-Case Scenario: Even when the malicious RIS optimally coordinates its interference using perfect CSI, the system achieves a positive secrecy rate, proving the robustness of the design.
• Sensing Performance:
  • SINR: The system successfully maintains sensing SINR for both targets. As the eavesdropper moves further away, its sensing SINR drops (due to path loss), while the legitimate target's SINR improves due to reduced interference.
  • CRB: The Cramer-Rao Bound for AoD estimation of the eavesdropper decreases (improves) with higher transmit power, validating the system's ability to track hostile targets accurately.
• Trade-off: There is a critical trade-off between transmit power ($P_T$) and secrecy rate. Excessively high $P_T$ can increase interference and eavesdropper SINR, potentially degrading performance, necessitating optimal power selection.

5. Significance and Conclusion

This paper makes a significant contribution to the field of 6G Physical Layer Security. By integrating ISAC with target-mounted RISs, it demonstrates how future networks can simultaneously perform high-precision sensing and secure communication in hostile environments.

The study highlights that:

1. Dual RIS deployment is a viable strategy to counteract sophisticated UAV-based threats.
2. SDR-based optimization effectively handles the non-convex nature of joint beamforming and phase shift design.
3. Robustness is achievable even when adversaries possess superior channel knowledge, provided the legitimate system optimizes its resources (RIS elements and beamforming) correctly.

The work paves the way for future research into dynamic scenarios involving moving targets and the integration of AI-driven security mechanisms for sensing signals.