Caching Scheme with Edge Nodes for Mobile Cyber Physical Systems

Abstract

To improve user's quality of experience (QoE) in Mobile Cyber Physical Systems (MCPS), caching layered contents at a group of edge nodes that are close to mobile users has been advocated as an efficient solution. However, due to the complexity of trust management and the limited caching capacity of edge nodes, designing an efficient caching scheme for edge nodes becomes a challenge. In this paper, we propose a novel caching scheme for edge nodes in MCPS. Firstly, layered contents are cached at edge nodes so that a flexibility of content quality can be achieved. Secondly, the trustworthiness of edge nodes are evaluated and updated, based on mobile users' feedback after they obtain the cached contents. Thirdly, mobile users' demands in different social communities are analyzed based on the unconstrained nonlinear programming and optimized with the gradient descent algorithm. A max-min based caching scheme is presented to select a group of trustworthy edge nodes to allocate caching resources for mobile users. Finally, we evaluate the proposed work in simulations and results show that the presented scheme not only improves the utilities of edge nodes but also increases the QoE of mobile users.