A Deadline-Aware Offloading Scheme for Vehicular Fog Computing at Signalized Intersection
A Deadline-Aware Offloading Scheme for Vehicular Fog Computing at Signalized Intersection
复制标题
DOI：
10.1109/percomworkshops48775.2020.9156078
复制
发表时间：
2020-03
期刊：
影响因子：
--
通讯作者：
Beichen Yang;Min Sun;X. Hong;Xiaoming Guo
中科院分区：
文献类型：
--
作者：
Beichen Yang;Min Sun;X. Hong;Xiaoming Guo
车载雾计算是利用车辆上的计算资源来提供计算服务的一种有前途的范例。同时，城市信号交叉口车辆独特的移动模式可用于提高车雾的性能。在本文中，我们通过考虑交叉口车辆的移动模式，提出了一种适合信号交叉口的卸载方案。我们的卸载方案解决了截止日期和复制成本的问题。预测移动性是进一步计算连接时间和链路带宽以在十字路口找到合适服务车辆的关键因素。该方案在给定的期限限制和预算下实现了最大的卸载成功率。我们的评估使用交通模拟器 SUMO 生成的真实交通数据。将结果与给定多个参数的一些基准进行比较。与最佳情况卸载场景相比，所提出的方案可以获得令人满意的卸载成功率。
Vehicular Fog Computing is a promising paradigm to exploit the computational resources on vehicles for providing computing services. Meanwhile, the unique mobility patterns of vehicles at the urban signalized intersection can be used to improve the performance of Vehicular Fog. In this paper, we present an offloading scheme tailored to the signalized intersection, by taking the vehicles' mobility pattern at the intersection. Our offloading scheme addresses the conditions of deadlines and the cost of replications. Predicting mobility is a key factor in further calculating the connection time and the link bandwidth for finding appropriate service vehicles at intersections. The scheme achieves maximum offloading success rate under given deadline constraints and budgets. Our evaluation uses realistic traffic data generated by traffic simulator SUMO. The results are compared with a few benchmarks given multiple parameters. The proposed scheme can achieve a satisfactory offloading success rate compared with the best-case offloading scenario.