| 摘要: |
| 工业设备接入网络实现生产自动化的过程中数据量级快速增长，而边缘层设备资源有限，无法完成全部任务请求。针对边缘层设备合理高效处理端设备任务请求的问题，提出了一种基于多跳计算卸载方法的物联网边缘网关（Internet of Things Edge Gateway，IoTEG）框架。该框架要求数据优先在网关侧处理以降低时延和保护隐私。首先，该框架根据端设备任务流特点将其分为时敏和非时敏两类。其次，设计了任务轮转调度处理机制，对任务流按时延要求高低进行处理。最后，设计了基于实时网络资源、实时本地资源和任务类型的最优联合计算卸载策略。实验结果表明，IoTEG框架能有效提高任务卸载的成功率，并能够高效处理不同类型的任务。 |
| 关键词: 工业物联网（IIoT） 边缘网关 调度机制 计算卸载 智能边缘计算 |
| DOI：10.20079/j.issn.1001-893x.230216001 |
|
| 基金项目:国家重点研发计划(2019YFB2101902) |
|
| Intelligent Edge Computing for Industrial IoT Gateway |
| WANG Jiawei,ZHAO Xiaoyan,ZHANG Zhaohui,WANG Yihao |
| (1.School of Automation and Electrical Engineering,University of Science and Technology Beijing,Beijing 100083,China;2.Beijing Engineering Research Center of Industrial Spectrum Imaging,Beijing 100083,China) |
| Abstract: |
| In the process of industrial equipment accessing the network to realize production automation,the data volume increases rapidly,while the resources of the edge layer equipment are limited and cannot complete all task requests.For the problem of reasonably and efficiently processing end device task requests by edge devices,an Internet of Things edge gateway (IoTEG) framework based on multi-hop computing offload method is proposed.The framework requires data to be processed preferentially on the gateway side to reduce latency and protect privacy.First of all,the task flow is divided into time-sensitive and non-time-sensitive by the framework according to the characteristics of end device tasks.Secondly,the task rotation scheduling processing mechanism is designed to process the task flow according to the delay requirement.Finally,an optimal joint computing offloading strategy based on real-time network resources,real-time local resources and task types is designed.The experimental results show that the IoTEG framework can effectively improve the success rate of task offloading and can efficiently handle different types of tasks. |
| Key words: industrial Internet of Thing(IIoT) edge gateway scheduling mechanism compute offloading intelligent edge computing |
