Privacy-Preserving Decentralized Edge Caching in 5G Networks

Yiming Zeng,Yaodong Huang,Zhenhua Liu,Ji Liu,Yuanyuan Yang
DOI: https://doi.org/10.1109/CLOUD53861.2021.00032
2021-01-01
Abstract:How to serve mobile users in rural areas by 5G networks is challenging due to the sparse distribution of base stations and poor connection to the cloud. Existing solutions focus on transmission frequency implementation such as frequency multiplexing. In this paper, we consider a decentralized caching scheme for two reasons. first, caching contents in the edge is an effective approach to reduce the transmission latency and improve the quality of service for mobile users. Second, the decentralized caching allows base stations to serve mobile users without any coordination of the cloud. Meanwhile, data privacy in the edge is critical for individual users. This paper aims to jointly determine the caching and routing policy in rural areas of 5G networks in a decentralized manner and simultaneously design a proper privacy-preserving mechanism. We tackle the challenges in two progressive steps. First, we design a decentralized algorithm with the convergence guarantee. Furthermore, we enhance the developed decentralized algorithm with a privacy-preserving mechanism based on (local) differential privacy and prove its privacy guarantee. We conduct extensive numerical simulations based on real-world traces to evaluate the proposed algorithms. Results highlight significant performance improvements compared to existing baselines.
What problem does this paper attempt to address?