Securing V2I Backscattering from Eavesdropper

Ruotong Zhao, Deepak Mishra, Aruna SeneviratneICC 2024 · DOI

TL;DR

This paper introduces a novel framework to secure Vehicle-to-Infrastructure (V2I) backscattering communications against eavesdropping by optimizing the reflection coefficients, power allocation, and vehicle trajectory. By employing an alternating optimization approach, they achieve a significant improvement in the sum secrecy rate, paving the way for more secure and efficient future intelligent transportation systems.

What Problem Does This Paper Address?

As cities become "smarter" and more connected, particularly with the advent of 6G technologies, the communication between vehicles and infrastructure (V2I) is becoming increasingly critical. This is especially true for autonomous driving systems. However, a key challenge is ensuring this communication is energy-efficient and secure. Backscattering communication offers a low-power solution for V2I, but it's vulnerable to eavesdropping, where malicious actors can intercept sensitive data. Existing research has focused on physical layer security (PLS) for vehicular networks, but there's a gap in securing V2I backscattering specifically against eavesdropping attacks from nearby vehicles. This paper aims to fill that gap by developing a secure and sustainable framework for V2I backscattering.

Methods

The core of this paper's contribution lies in its novel optimization framework. The authors tackle the complex, non-convex problem of maximizing the sum secrecy rate by decomposing it into three sub-problems:

  • Reflection Coefficient Optimization: This sub-problem is addressed using parallel computing and a fractional programming (FP) technique. The authors prove that this problem is convex, allowing for efficient global optimization.
  • Power Allocation (CW and AN): Similar to the reflection coefficient optimization, FP is used to transform the non-convex power control problem into a convex one. This allows for optimal allocation of power for the continuous wave (CW) signal and artificial noise (AN).
  • Vehicle Trajectory Optimization: This is framed as finding the optimal path for the target vehicle (TV) to maximize the secrecy rate, subject to velocity and acceleration constraints. The authors identify a finite set of candidate optimal points and use a theorem to establish conditions for global optimality.

These three sub-problems are solved iteratively using an Alternating Optimization (AO) approach. The process starts with initial assumptions for power and trajectory, optimizes the reflection coefficients, then updates power, then trajectory, and repeats until convergence. This iterative approach ensures that each sub-problem is solved optimally given the solutions of the others, leading to a near-global optimum for the overall problem.

Datasets

Not discussed in the paper. The paper relies on numerical simulations with default parameters and parameters from existing literature for evaluation.

Key Insights and Ideas

  • V2I Backscattering Security is Crucial: The paper highlights that as V2I communication becomes more prevalent, securing these low-power backscattering links against eavesdropping is paramount for safe and efficient intelligent transportation.
  • Joint Optimization is Key: Optimizing reflection coefficients, power allocation, and vehicle trajectory simultaneously is essential for maximizing the secrecy rate. Treating these in isolation leads to suboptimal performance.
  • Convexification for Tractability: The clever use of fractional programming to transform non-convex problems into convex ones is a significant methodological contribution, enabling efficient global optimization of sub-problems.
  • Trajectory Matters for Security: The vehicle's path is not just about efficiency but directly impacts the communication security, especially in the presence of eavesdroppers. The paper shows how trajectory design can be leveraged to enhance secrecy.
  • Artificial Noise (AN) is Effective: Injecting AN by the infrastructure is shown to be a viable strategy to confuse eavesdroppers, especially when the eavesdropper lacks knowledge of the transmitted signal.

Key Results

  • Significant Secrecy Rate Improvement: The proposed secure V2I backscattering framework outperforms existing benchmarks by a substantial margin. For 50 time slots, the improvement is over 4.7 times in terms of secrecy rate.
  • Fast Convergence: The alternating optimization algorithm demonstrates rapid convergence, often requiring only 2-3 iterations to reach a stable solution.
  • Impact of Parameters:
    • Increasing the number of time slots generally leads to higher secrecy rates, providing more opportunities for optimization.
  • Higher power budgets directly translate to improved secrecy rates.
  • Longer time slot durations can negatively impact the secrecy rate due to increased distances and trajectory challenges.
  • The proximity of the eavesdropper significantly influences the optimal reflection coefficient and overall secrecy. The system adapts by increasing reflection when the eavesdropper is farther away.
  • Individual Variable Contributions: Power allocation was found to have the most significant impact on system performance, followed by reflection coefficient optimization and then trajectory design.

Limitations and Future Work

  • Quasi-Static Fading Assumption: The study assumes quasi-static large-scale fading, meaning channel conditions are constant within each time slot but change between slots. Future work could explore more dynamic channel conditions.
  • One-Dimensional Trajectory Optimization: The paper optimizes a one-dimensional trajectory variable ($x_g$). Extending this to a full 2D or 3D trajectory optimization could offer further improvements.
  • Limited Eavesdropper Model: The paper considers a single eavesdropper. Future research could investigate scenarios with multiple eavesdroppers or more sophisticated eavesdropping strategies.
  • Practical Implementation: While the framework is theoretically sound, practical implementation challenges related to precise CSI acquisition and real-time optimization in dynamic vehicular environments would be an interesting area for future exploration.

Summary updated

BibTeX
@inproceedings{zhao2024securing,
  author = {Ruotong Zhao and Deepak Mishra and Aruna Seneviratne},
  title = {{Securing V2I Backscattering from Eavesdropper}},
  booktitle = {ICC},
  year = {2024},
  doi = {10.1109/icc51166.2024.10622397},
}
Figure