First Glimpse on Physical Layer Security in Internet of Vehicles: Transformed From Communication Interference to Sensing Interference

Zhiyong Feng, Kaixuan Li, Kan Yu, Xiaowu Liu, Dingyou Ma, Qixun Zhang, Zhiyong Feng 0001, Dong LiIEEE Trans. Communications 2025 · DOI

TL;DR

This paper introduces a novel physical layer security (PLS) approach for the Internet of Vehicles (IoV) by leveraging sensing interference, rather than communication interference, to protect data. By jointly optimizing vehicle trajectory and transmission power, the authors achieve enhanced security and reliability in integrated sensing and communication (ISAC) systems, outperforming traditional PLS methods.

What Problem Does This Paper Address?

The Internet of Vehicles (IoV) relies heavily on Integrated Sensing and Communication (ISAC) for applications like enhanced driving safety and traffic efficiency. However, the open nature of wireless communication in IoVs makes confidential data transmission vulnerable to eavesdropping. Traditional Physical Layer Security (PLS) methods often rely on generating artificial noise (AN), which incurs extra power costs. Furthermore, in ISAC systems, communication and sensing functions often share frequency bands, leading to coupled interference that complicates existing PLS solutions. This paper addresses the challenge of securing IoV communications within ISAC frameworks by proposing a novel PLS strategy that utilizes the inherent sensing interference from other vehicles, rather than generating additional interference, to protect data.

Methods

The core of this paper's contribution lies in transforming the PLS paradigm from relying on communication interference to utilizing sensing interference. The authors propose a novel metric, the Transmission Reliability and Sensing Accuracy-based Secrecy Rate (TRSA_SR), to evaluate system performance. They then formulate an optimization problem to maximize this metric by jointly designing:

  • Communication Duration: Determining the optimal time slots for transmitting confidential information, specifically when eavesdroppers are likely to be within the sensing interference range of other vehicles (Carols).
  • Transmission Power: Optimizing the power allocated for transmitting confidential data.
  • Straight Trajectory: Adjusting the movement path of the legitimate transmitter (Alice).

To tackle the non-convex nature of the optimization problem, they employ a Block Coordinate Descent (BCD) method combined with Successive Convex Approximation (SCA). This involves decomposing the problem into two sub-problems: transmission power optimization and trajectory optimization, which are then iteratively solved. An Alternating Optimization (AO) algorithm is proposed to achieve a globally optimal solution.

Datasets

Not discussed in the paper. The evaluation is based on numerical simulations with parameters derived from existing literature and realistic IoV scenarios.

Key Insights and Ideas

  • Sensing Interference as a Security Tool: The most significant insight is the successful transformation of sensing interference, typically considered a nuisance, into a valuable resource for enhancing physical layer security in IoVs. This offers a power-efficient alternative to traditional AN-based PLS.
  • TRSA_SR Metric: The introduction of the TRSA_SR metric is crucial. It provides a holistic measure that simultaneously considers communication reliability, sensing accuracy, and security, which is essential for ISAC systems.
  • Joint Optimization is Key: The paper highlights that simply optimizing one aspect (like trajectory or power) is insufficient. The synergistic effect of jointly optimizing transmission power, trajectory, and communication duration is critical for maximizing the TRSA_SR.
  • Sensing Interference Effectiveness: The results demonstrate that when traditional PLS methods fail to provide positive secrecy rates, the proposed sensing interference-based method can achieve significant secrecy rates, underscoring its effectiveness.

Key Results

  • The proposed SI-PLS with ST-TP method achieves a secrecy rate of 3.92 bps/Hz in simulations, outperforming traditional PLS methods that yield non-positive secrecy rates.
  • The TRSA_SR metric is shown to be sensitive to noise power, with higher noise generally decreasing reliability but improving security.
  • Increasing sensing power enhances security but can degrade reliability.
  • The proposed AO algorithm converges effectively, achieving a stable TRSA_SR after approximately 5 iterations.
  • The paper demonstrates that the proposed scheme significantly reduces COP and SOP compared to traditional PLS methods, by an average of 82% and 41% respectively, under certain conditions.

Limitations and Future Work

  • Simplified Scenario: The current analysis often assumes a single Eve and a single Carol for mathematical tractability. The paper notes that the method can be extended to multiple interferers.
  • Perfect CSI Assumption: The analysis relies on the assumption of perfect Channel State Information (CSI) for both legitimate and eavesdropping channels. In real-world scenarios, CSI acquisition is imperfect and can be challenging.
  • Future Work: The authors suggest extending the framework to more complex scenarios, including non-line-of-sight (NLOS) channels, multiple eavesdroppers and interferers, and more sophisticated vehicle mobility models. Investigating the impact of different radar sensing beamwidth models and exploring adaptive strategies for dynamic environments are also promising directions.

Summary updated

BibTeX
@article{feng2025first,
  author = {Zhiyong Feng and Kaixuan Li and Kan Yu and Xiaowu Liu and Dingyou Ma and Qixun Zhang and Zhiyong Feng and Dong Li},
  title = {{First Glimpse on Physical Layer Security in Internet of Vehicles: Transformed From Communication Interference to Sensing Interference}},
  journal = {IEEE Trans. Communications},
  year = {2025},
  doi = {10.1109/tcomm.2025.3611638},
}
Figure