On Pursuit of Privacy Preservation for Dependable Offloading in VECON: An Optimization Perspective
Xuehan Li, Tao Jing, Ruinian Li, Xiaoxuan Wang, Hengyu Yu, Yan Huo · 2022 IEEE Globecom Workshops (GC Wkshps) · 2022
The Internet of Vehicles (IoVs) networks are gaining significant advantages as a result of the emergence of Vehicular Edge Computing and Networks (VECONs). For VECONs, efficient and secure task offloading is essential for users to obtain dependable services. Nevertheless, few efforts are directed towards security issues such as privacy exposure risk during offloading, which seriously affect modeling and optimization of offloading dependability. This paper presents a new measure of offloading dependability entitled cost from Energy consumption, Delay and Privacy exposure risk (cEDP). A dependable offloading framework is proposed, in which tasks are reasonably divided and the Fully Homomorphic Encryption (FHE) algorithm is implemented to mitigate privacy exposure risks. A self-learning based decentralized computation and privacy preservation offloading (SDCPO) algorithm is presented to solve the optimal offloading problem dependably and decentrally. Numerical results show that over the centralized offloading schemes, SDCPO has better convergence performance with lower cEDP, which verifies the dependability of the offloading scheme and algorithm.