Edge Computing-based Fault-Tolerant Framework

Abstract

With the evolution of vehicular networks, the Intelligent Transportation System (ITS) has emerged as a promising technology for autonomous road transport. For a successful deployment of ITS, security and reliability are the most challenging factors to be tackled to ensure Vehicle-to-Infrastructure (V2I) and Infrastructure-to-Infrastructure (I2I) communications. Due to unreliable communications in vehicular networks, implementing fault-tolerant techniques for the Road Side Unit (RSU) infrastructure is an imperial need. Within this context, the contributions of this paper are twofold: (i) we propose a distributed fault-tolerant framework for V2I and I2I communications based on edge computing to resolve hardware- and network connectivity-based failures. The fault tolerance issue is addressed by employing open messaging standards as a subscription-based data replication solution at the edge. We also adopt Kubernetes for the fault-tolerant management, combined with high-availability mechanism, allowing automatic reconfiguration of the data processing pipeline; and (ii) we implement a demonstrator system for vehicular networks-based smart mobility to assess fault tolerance capabilities. The experimental results show that our proposed framework dynamically tolerates RSU-related failures during the vehicular communication phase.