Low-Complexity Joint Power Allocation and Trajectory Design for UAV-Enabled Secure Communications With Power Splitting

Kaidi Xu, Ming-Min Zhao, Yunlong Cai, Lajos HanzoIEEE Trans. Communications 2021 · DOI

TL;DR

This paper proposes a novel power splitting strategy for UAV-enabled secure communication systems to maximize secrecy rate. They develop a low-complexity iterative algorithm that jointly optimizes the UAV's trajectory and power allocation, achieving significant speedups over existing methods while maintaining comparable performance.

What Problem Does This Paper Address?

Unmanned Aerial Vehicles (UAVs) offer flexible and rapid deployment for communication services, but their inherent mobility and open wireless channels make them vulnerable to eavesdropping. Traditional security measures like encryption can be computationally intensive. This paper addresses the challenge of enhancing physical layer security for UAV-enabled communications by simultaneously transmitting confidential information and artificial noise (AN) to jam eavesdroppers. The core problem is to optimize the UAV's flight path (trajectory) and how it allocates its transmit power (including the split between information and AN) over time to maximize the rate at which legitimate users can receive information securely, subject to constraints on the UAV's speed, energy, and power.

Methods

The authors tackle this complex optimization problem by proposing a novel iterative algorithm that combines three established techniques:

  • Block Coordinate Descent (BCD): This method breaks down the overall problem into smaller, more manageable subproblems (power allocation and trajectory optimization) that are solved iteratively.
  • Concave-Convex Procedure (CCCP): This technique is used to handle non-convex parts of the optimization problem by approximating them with convex functions, making them amenable to standard optimization solvers.
  • Alternating Direction Method of Multipliers (ADMM): This algorithm is employed to efficiently solve the resulting convex subproblems, particularly the trajectory optimization, by introducing auxiliary variables and iteratively updating primal and dual variables.

The key innovation lies in the specific application and combination of these methods to achieve a "nearly closed-form" solution for each update step, leading to significantly reduced computational complexity.

Datasets

Not discussed in the paper. The paper focuses on a simulation-based evaluation of the proposed algorithm.

Key Insights and Ideas

  • Power Splitting as a Dual-Purpose Tool: The paper demonstrates that splitting the UAV's transmit power into two parts – one for confidential messages and another for artificial noise (AN) – is a powerful strategy. This not only directly transmits information but also actively disrupts eavesdroppers, enhancing physical layer security.
  • Joint Optimization is Crucial: Optimizing the UAV's trajectory and power allocation jointly is shown to be more effective than optimizing them independently. The UAV can strategically position itself to maximize its distance from eavesdroppers while simultaneously transmitting information and jamming signals.
  • Low-Complexity Algorithm Design: The paper's core contribution is a sophisticated iterative algorithm that cleverly decomposes the problem and uses approximations and ADMM to achieve solutions with significantly lower computational cost than traditional convex optimization solvers like CVX. This makes the approach practical for real-time applications.
  • Trajectory Dictates Security: The simulations reveal that the UAV's trajectory plays a critical role in achieving high secrecy rates. By flying closer to the legitimate user and further from the eavesdropper, the UAV can significantly improve security, especially when AN is being transmitted.

Key Results

  • Significant Speedup: The proposed algorithm achieves a running time that is at least 30 times lower than conventional methods using CVX for similar problems. This speedup is attributed to the algorithm's ability to derive nearly closed-form solutions for each update step.
  • Comparable Performance: Despite the significant reduction in complexity, the proposed algorithm achieves average secrecy rates that are almost identical to those obtained by using CVX. In some cases, it even slightly outperforms CVX due to precision differences in handling logarithmic functions.
  • Improved Secrecy Rate: The proposed power splitting and joint optimization approach consistently outperforms benchmark schemes (fixed trajectory, naive power splitting, and no AN) in terms of average secrecy rate across various scenarios, including different Bob-Eve distances and flight times.
  • Scalability: The running time of the proposed algorithm scales much more favorably with the number of time slots (T) compared to CVX-based methods, indicating better scalability for longer flight durations.

Limitations and Future Work

  • 3D Trajectory Extension: The paper mentions that the proposed algorithm can be extended to 3D trajectory optimization by introducing auxiliary variables for altitude, which would be a natural next step.
  • Assumptions: The model assumes Line-of-Sight (LoS) communication, which is common for UAVs but might not hold in all scenarios. It also assumes a known eavesdropper location.
  • Parameter Tuning: While the algorithm is efficient, the performance might still be sensitive to the choice of certain parameters (e.g., penalty parameters in ADMM), which would require careful tuning.
  • Real-World Deployment: The paper focuses on simulation results. Future work could involve experimental validation of the proposed algorithm in real-world UAV communication systems.

Summary updated

BibTeX
@article{xu2021low,
  author = {Kaidi Xu and Ming-Min Zhao and Yunlong Cai and Lajos Hanzo},
  title = {{Low-Complexity Joint Power Allocation and Trajectory Design for UAV-Enabled Secure Communications With Power Splitting}},
  journal = {IEEE Trans. Communications},
  year = {2021},
  doi = {10.1109/tcomm.2020.3042462},
}
Figure