Abstract
In this work, a covert offloading framework is established for unmanned aerial vehicle (UAV) system. Specifically, one UAV-enabled mobile edge computing (MEC) server is deployed to assist the offloading for multiple ground users, while the offloading behavior might be exposed and detected due to the existence of a malicious warden. To enhance the covertness, each user splits a part of its power to transmit jamming signal, and the relationship between users' power split ratio and the warden's minimum detection error probability (DEP) is investigated. Then, an efficient energy minimization algorithm is designed by optimizing users' power split, computing resource allocation and deployment of the UAV-MEC server jointly, subject to specific covertness, power and transmission rate constraints. Finally, simulation results are provided to demonstrate the effectiveness of our covert offloading design.