Secure Status Updates for Internet of Drones: A Deep Q-Learning-Based Antenna Selection Approach

• Published in: 2024 International Wireless Communications and Mobile Computing (IWCMC) pp. 461 - 466
• Main Authors: Xiao, Yuquan, Du, Qinghe, Lu, Chen, Wang, Yizhuo
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27.05.2024
• Subjects: Age of information; deep reinforcement learning; Downlink; internet of drones; Minimization; Q-learning; Receiving antennas; status update; Surveillance; transmit antenna selection; Transmitting antennas; Wireless networks
• ISSN: 2376-6506
• DOI: 10.1109/IWCMC61514.2024.10592538

Status update applications are very common in the internet of drones (IoD), where some of status update information are privacy-sensitive. How these information can be efficiently and securely transmitted remains as an open challenging issue. Motivated by this concern, we in this paper consider using the transmit antenna selection (TAS) technique to guarantee secure status updates over IoD downlink networks, in which only the drones corresponding to the highest channel gain on each antenna can become the candidate receivers, and then the base station (BS) selects one antenna to transmit the associated status update information with wiretap coding to the corresponding candidate. The mobility of drones causes the channel gain is time-varying, and thus one natural question arises, i.e., what is the optimal antenna selection scheme in terms of the long-term network-wide freshness performance. To answer this question, the weighted sum of average age-of-information (AoI) minimization problem is formulated, and we propose a deep Q-learning-based antenna selection approach to solve it. It is worth mentioning that, in light of the try-error mechanism of Q-learning, no prior knowledge of the wireless environments is required for our proposal in contrast to the traditional schemes. Finally, the numerical results verify that our proposal can further reduce the weighted sum of AoI as compared with the state-of-the-art max-weight scheme.