Replica fault-tolerant scheduling with time guarantee under energy constraint in fog computing

Abstract

Fog computing offers stronger local computing power and reduces data transmission load, making it an ideal solution to meet the energy-saving and efficient requirements of intelligent connected vehicle applications. As intelligent networked vehicles and vehicle–road collaboration technologies advance rapidly, optimizing scheduling under fog computing architecture has become a prominent research area. However, existing studies primarily concentrate on parallel task scheduling with low energy consumption or high real-time performance, failing to address the requirement for high reliability in intelligent networked vehicle scenarios. To achieve time and reliability optimization for vehicular applications in fog computing architecture, this paper proposes a fog computing task scheduling algorithm and explores its extension using replication techniques. Subsequently, the algorithm underwent evaluation utilizing randomly generated directed acyclic graph models as well as real-life automotive application instances. The experimental findings indicate that in comparison to existing methods, the proposed algorithm exhibits a notable improvement in reliability while ensuring time optimization, thereby demonstrating a distinct level of advancement and practicality.