Grand Challenges of Online Flight Control Learning and Privacy-preserving Data Capture in UAVs-enabled Internet-of-Things
Ref: CISTER-TR-240102 Publication Date: 10, Jan, 2024
Grand Challenges of Online Flight Control Learning and Privacy-preserving Data Capture in UAVs-enabled Internet-of-Things
Ref: CISTER-TR-240102 Publication Date: 10, Jan, 2024Abstract:
Applications of unmanned aerial vehicles (UAVs) for data collection are a promising means to extend Internet of Things (IoT) networks to remote and hostile areas and to locations where there is no access to power supplies. The adequate design of UAV velocity control and communication decision making is critical to minimize the data packet losses at ground IoT nodes that result from overflowing buffers and transmission failures. Due to the broadcast nature of wireless channels, data communications between the UAVs and the ground IoT nodes are vulnerable to eavesdropping attacks. In this talk, we discuss a new deep-graph-based reinforcement learning framework, which trains the real-time continuous actions of the UAV in terms of the flight speed, heading, and the offloading schedule of the IoT nodes. Moreover, we study a channel-based secret key generation in UAVs-enabled IoT, where received signal strength at the UAVs and the IoT nodes is quantized to generate the time-varying secret keys. A dynamic programming-based channel quantization scheme is developed to minimize the secret key bit mismatch rate of the UAVs and the IoT nodes by recursively adjusting the quantization intervals.
Following Dr. Kai Li's presentation, there will be a panel discussion with Dr Tahmina Zebin and Dr Hongying Meng.
Kai LiThe Grand Challenges Series, Department of Computer Science, Brunel University London (BUL).
London, United Kingdom.
Record Date: 13, Jan, 2024