Securing Task Computation Offloading for Mobile Edge Computing in Untrusted Relay Networks

Abstract

This paper aims to secure task offloading and service caching in relay-assisted heterogeneous mobile edge computing networks. The relay nodes might be untrusted in such heterogeneous networks due to their limited computational power. Therefore, a friendly jamming technique is used to confuse the untrusted relay nodes. Hence, an optimization problem is formulated to minimize the system latency and energy consumption under secrecy capacity constraint. As the joint problem of the service caching and task offloading is an NP-hard problem, this work deploys a low-complexity Lyapunov drift-plus-penalty optimization technique based on the Gibbs sampling algorithm [5]. The simulation results show the superiority of the proposed framework over the state-of-the-art in terms of high secrecy capacity and low computational complexity. When the secrecy capacity threshold increases, the secrecy capacity is enhanced by approximately 2.99 %, while the system latency and energy consumption increase by 55.85 % and 4.76 %, respectively, compared to the literature.