Mean Field Games Theoretic for Mobile Privacy Security Enhancement in Edge Computing

Abstract

Edge computing paradigm extends the cloud service to the edge of networks and reduces service latency to edge users. Meanwhile, the characteristics of edge computing, such as mobility and heterogeneity, arise new privacy security challenges. Current research on privacy security relies on information security techniques like encryption and anonymization. In this paper, we present an application of mean field game theory for privacy leakage with large scale edge devices in edge computing because the edge devices are resource constrained. The contributions are threefold: first, we construct an individual cost function with a mean field term and discuss the evolution of the state if the number of devices is large enough. Subsequently, we elaborate the Nash equilibrium of the mean field game models which are coupled through the Hamiltonian–Jacobi–Bellman (HJB) backward and Fokker–Planck–Kolmogorov (FPK) forward equations. In addition, an approximate method is introduced to analyze the stable solution of the state. Finally, numerical examples are provided to illustrate the stability of the state and the presented strategy.