.png){kind=icon}
.png){kind=icon}
Return
Secure Wireless Networks: AI-Driven UAV Data Relay
.png){kind=icon}
PRE
AI
DOI：10.1109/TCE.2024.3412097.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The openness of wireless networks makes sensing data collection tasks vulnerable to jamming attacks. Techniques such as frequency hopping or spoofing are not suitable for multiuser data transmission scenarios because they require additional channels or come at the expense of designated users' benefits. Unmanned aerial vehicles (UAVs), with their advantage of communication coverage and channel quality, have become an effective paradigm for anti-jamming through data relay. However, current research primarily focuses on transmitting devices in stationary environments. When these devices are in motion, such as in the case of UAVs performing data relay, they may encounter challenges including degradation in channel quality and failures in coverage. In this paper, we propose employing a UAV to consistently track mobile devices(MDs), ensuring superior channel quality while simultaneously providing data relay. To address the unpredictable movement of MDs and varying jamming power, we formulate the UAV's interaction within the existing environment as a Markov Decision Process (MDP). Our approach involves joint optimization of UAV flight speed, direction, and relay decisions to minimize data transmission delay. We then propose the Deep Deterministic Policy Gradient (DDPG) to account for the non-convexity of the optimization problem. Additionally, we consider the complexity and frequency of the action space, decomposing it, and optimizing it by alternating between upper and lower-level policies. Experimental results demonstrate that the proposed scheme can quickly converge to the optimal solution, significantly reducing data transmission delays.
Keywords:
Jamming
Relays
Autonomous aerial vehicles
Data collection
Consumer electronics
Data communication
Interference
Anti-jamming
hierarchical reinforcement learning
mobile wireless network
unmanned aerial vehicle
physical layer security
Journal
IF:
10.9
Papers:
5.1K
Citations:
6.8K