On Secure EKF-enhanced UAV-ISAC Systems

Technical Summary: Secure EKF-enhanced UAV-ISAC Systems

Problem Statement
This work addresses the challenge of securing Integrated Sensing and Communication (ISAC) systems assisted by Uncrewed Aerial Vehicles (UAVs) in the presence of mobile eavesdroppers. While existing UAV-ISAC research often assumes stationary targets or focuses solely on single-target tracking, practical scenarios involve mobile legitimate users and illegitimate eavesdroppers with uncertain locations. The core problem is to maximize the achievable Secrecy Rate (SR) by jointly designing transmit beamforming vectors and the UAV's trajectory. This optimization must operate under practical constraints, including sensing beamwidth requirements, transmit power limits, and the UAV's propulsion energy consumption, while simultaneously tracking and predicting the motion states of both legitimate and illegitimate users.

Methodology
The proposed system model features a full-duplex aerial base station (UAV) equipped with Uniform Planar Arrays (UPAs) communicating with a mobile ground user (GU) while actively jamming a mobile eavesdropper. The system utilizes the same signals for communication and sensing.

1. System Modeling: The authors model the Air-to-Ground (A2G) links as Line-of-Sight (LoS) channels. The UAV transmits a superposition of information-bearing signals for the legitimate user and dedicated sensing signals for the eavesdropper. The received Signal-to-Interference-plus-Noise Ratio (SINR) and achievable rates are derived for both users. The UAV's horizontal flight energy consumption is modeled based on aerodynamic principles.
2. EKF-Based Tracking: To handle target mobility, an Extended Kalman Filter (EKF) framework is integrated. The system estimates the state vectors (position and velocity) of both the legitimate user and the eavesdropper using sensing measurements (time delay, Doppler shift, and angles of arrival/departure). The measurement noise covariance is inversely proportional to the echo Signal-to-Noise Ratio (SNR).
3. Optimization Formulation: The objective is to maximize the SR subject to constraints on transmit power, rank-one beamforming, sensing thresholds, UAV kinematics (velocity, acceleration), and propulsion power. The problem is highly non-convex due to the coupling between trajectory and beamforming variables, as well as non-convex constraints related to the UAV's power consumption and the fractional form of the SR.
4. Algorithm Design: An efficient iterative algorithm is developed based on the Block Coordinate Descent (BCD) method. The problem is decomposed into two subproblems solved alternately:
   • Beam Optimization: With a fixed trajectory, the beamforming vectors are optimized. Successive Convex Approximation (SCA) and slack variables are used to handle non-convex constraints, transforming the problem into a Semidefinite Programming (SDP) form.
   • Trajectory Optimization: With fixed beamforming, the UAV trajectory is optimized. SCA is again employed to approximate non-convex constraints regarding flight power and the secrecy rate.
   • EKF Integration: In each time slot, the EKF updates the state estimates based on the current trajectory and beamforming, which are then used as inputs for the next iteration.

Key Contributions

• Joint Tracking and Security: Unlike prior works that often ignore eavesdropper mobility or focus on single targets, this paper investigates a secure UAV-ISAC system capable of simultaneously tracking and predicting the trajectories of multiple mobile targets (both legitimate and illegitimate).
• Explicit Security Modeling: The system explicitly incorporates a mobile eavesdropper with an uncertain location into the design, utilizing EKF-based predictions to generate targeted jamming and communication strategies.
• Algorithmic Solution: The authors propose a novel iterative algorithm integrating BCD, SCA, and EKF to solve the resulting non-convex optimization problem, yielding a high-quality suboptimal solution.
• Comprehensive Constraints: The optimization accounts for practical physical constraints, including sensing beamwidth, transmit power, and the specific propulsion energy consumption of the UAV.

Results
Numerical simulations validate the proposed scheme against benchmarks, including a fixed-location UAV scenario.

• Trajectory and Tracking: The proposed scheme demonstrates that the UAV effectively flies toward the legitimate user to enhance link quality while maintaining tracking and jamming of the eavesdropper. The EKF-based predictions closely align with true trajectories, showing convergence over time.
• Secrecy Rate: The proposed joint optimization achieves a significantly higher Secrecy Rate compared to the benchmark where the UAV remains stationary. The benchmark scheme suffers from filter divergence and performance degradation as targets move away from the fixed UAV.
• Parameter Sensitivity:
  • Interference Residual ($\psi$): Lower interference residual levels (better cancellation) yield higher SR, though the impact diminishes when the beam is already tightly focused on the eavesdropper.
  • Transmit Power ($P_{max}$): SR increases with transmit power but eventually saturates due to the simultaneous improvement in the eavesdropper's SNR.
  • Propulsion Power ($P_{hor}^{max}$): Higher available propulsion power allows the UAV to approach the legitimate user faster, thereby improving the SR.

Significance and Claims
The paper claims that by integrating ISAC with Physical-Layer Security (PLS) and EKF-based target tracking, UAVs can fully exploit their inherent advantages of flexible deployment and high mobility. The work bridges the gap between existing studies that overlook communication security or assume static targets. The proposed framework enables the design of targeted communication and jamming strategies under constrained resources, facilitating more secure and efficient transmissions. The authors note that the EKF-based framework can be extended to aerial-user scenarios and future work may focus on robust measurement models against clutter and multi-UAV coordination.