期刊
PEER-TO-PEER NETWORKING AND APPLICATIONS
卷 16, 期 1, 页码 71-86出版社
SPRINGER
DOI: 10.1007/s12083-022-01385-6
关键词
Resource allocation; Task offloading; Latency minimization; Edge computing; VANET
This study proposes a multi-vehicle side communication and edge computing collaboration framework to maximize the use of communication and computation resources and reduce task computation latency. It presents low-complexity solution and an improved optimization algorithm for matching and resource allocation. Evaluation shows that the method effectively meets the latency requirements and improves task execution efficiency.
There are numerous real-time and low latency application scenarios of the Internet of Vehicles (IoV), such as autonomous driving. The efficient use of limited computing and communication resources to perform IoV tasks is a hot topic in current research. Many idle vehicles (IVs) are parked around driving busy vehicles (BVs) on urban roads. This paper envisions a multi-vehicle side communication and edge computing collaboration framework with all vehicles acting as edge nodes to reduce BV task computation latency and maximize the use of each vehicle's communication and computation resources. We simulate the matching and resource allocation problem between BVs and IVs. The optimization goal is to minimize the latency, and the energy consumption is comprehensively considered. For the one-to-one matching case of BVs and IV, a new low-complexity reformulation linearization method solution is proposed. To solve the one-to-many matching problem between BVs and IVs, an improved biogeography-based optimization (IBBO) algorithm is used. Finally, the performance of the proposed task offloading and allocation method is evaluated by average task execution delay, the task computation time of BVs and IVs, energy consumption, and other metrics. The results show that for one-to-one and one-to-many matching, the proposed method can effectively guarantee the latency requirements of BVs. Compared with existing methods, our method in this paper can improve task execution efficiency by 118% while reducing the average task execution latency by 54.2%.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据