Peer Data Caching Algorithms in Large-Scale High-Mobility Pervasive Edge Computing Environments

Emerging innovative edge devices like drones, self-driving cars, phones/tablets and IoT nodes are revolutionizing our daily lives. Caching data among peer edge devices enables data sharing needed in many applications. In such applications, network scalability and node mobility bring many challenges. They change the topology and the resources in the network and make the network less robust. In this paper, we propose peer data caching strategies that consider the scale and mobility of these increasingly popular edge devices. We propose a grouping method creating a layered design to reduce the number of entities in each layer. We propose inter-group and intra-group optimization problems which proactively cache data onto best places to support robust and fast data access. We develop a 7-approximation algorithm for inter-group optimization and use uncapacitated facility location problems to solve intra-group optimization. We also transform the mobility of nodes into node behaviors to reduce the impact of mobility on the network. Our extensive simulation results show that our proposed strategies can apply to large-size and high-mobility networks, while achieving satisfactory results for data access.