Failure-Tolerant Task Offloading for Vehicular Fog Computing
Chaogang Tang, Huaming Wu, Chunsheng Zhu, Shuo Xiao, Haifeng Jiang, Dong Zeng · 2022
Vehicular fog computing (VFC) as one of the newly emergent computing paradigms has attracted widespread attention recently, which brings the computing and storage resources of cloud computing to the mobile terminal devices – smart vehicles. Despite the advantage of VFC, tasks are more susceptible to failures, owing to the limited computing capabilities of vehicles as well as the unstable wireless communication links during service provisioning. However, current works seldom pay attention to it. In this paper we strive to design failure-tolerant task offloading strategy in VFC, by considering different kinds of failures during task offloading. We also pose strict latency requirement on the response delay for each task in the optimization period. The goal of this paper is to minimize the average response latency for all the offloaded tasks in the optimization period by allocating the offloading requests to the fog vehicles. The optimization problem is formulated as one variant of multiple knapsack problem that is NP-hard. To solve this problem in polynomial time, we put forward a heuristic greedy algorithm to solve it. The extensive simulation is conducted to validate its efficiency and effectiveness and the results show its advantages compared to the baseline approaches.