Optimizing Physical Layer Security for Mobile Edge Computing with Prospect Theory
Yichao Chen,Cheng Zeng,Xin Zhou,Liang Huang,Shibo He
DOI: https://doi.org/10.1109/iccc62479.2024.10681994
2024-01-01
Abstract:Mobile edge computing (MEC) is gaining popularity, as it can improve mobile application performance by offloading local computation to edge servers. However, due to the openness of wireless communication, the confidentiality of MEC data transmission, especially the uploading, is endangered by eavesdropping attacks, whose risks have not been thoroughly depicted, evaluated and managed in the literature. In this paper, we investigate the application of friendly jamming (FJ) from base station (BS) and a nearby user to improve physical layer security (PLS) for MEC services. Further, we propose an adjustable multilevel assessment framework for user to evaluate confidentiality risks, and model the impact of its subjective biases under uncertainty based on Prospect Theory (PT). Then, we formulate an operator’s net income maximization problem by jointly considering effective data rate, user satisfaction, antenna usage, upload/FJ power consumption and FJ incentive cost. To solve the highly complex problem, we propose a two-stage control scheme, which first optimizes the secrecy capacity and then maximizes the effective data rate that achieves user satisfaction. By simulation, we evaluate the secrecy performance of the scheme under various conditions, and compare it with the one based on expected utility theory (EUT).