College of Computer Science, Inner Mongolia University, Hohhot 010021, China.
College of Computer Science and Technology, Inner Mongolia Normal University, Hohhot 010022, China.
Sensors (Basel). 2020 Mar 29;20(7):1900. doi: 10.3390/s20071900.
IoT and 5G technologies are making smart devices, medical devices, cameras and various types of sensors become parts of the Internet, which provides feasibility to the realization of infrastructure and services such as smart homes, smart cities, smart medical technology and smart transportation. Fog computing (edge computing) is a new research field and can accelerate the analysis speed and decision-making for these delay-sensitive applications. It is very important to test functions and performances of various applications and services before they are deployed to the production environment, and current evaluations are more based on various simulation tools; however, the fidelity of the experimental results is a problem for most of network simulation tools. PiFogBed is a fog computing testbed built with real devices, but it does not support the testing of mobile end devices and mobile fog applications. The paper proposes the piFogBedII to support the testing of mobile fog applications by modifying some components in the piFogBed, such as extending the range of end devices, adding the mobile and migration management strategy and inserting a container agent to implement the transparent transmission between end devices and containers. The evaluation results show that it is effective and the delay resulting from the migration strategy and container agent is acceptable.
物联网和 5G 技术使智能设备、医疗设备、摄像头和各种类型的传感器成为互联网的一部分,这为智能家居、智能城市、智能医疗技术和智能交通等基础设施和服务的实现提供了可行性。雾计算(边缘计算)是一个新的研究领域,可以加速这些对延迟敏感的应用程序的分析速度和决策速度。在将各种应用程序和服务部署到生产环境之前,测试它们的功能和性能非常重要,目前的评估更多地基于各种仿真工具;然而,对于大多数网络仿真工具来说,实验结果的准确性是一个问题。PiFogBed 是一个使用真实设备构建的雾计算测试平台,但它不支持移动终端设备和移动雾应用程序的测试。本文通过修改 piFogBed 中的一些组件,如扩展终端设备的范围、添加移动和迁移管理策略以及插入容器代理来实现终端设备和容器之间的透明传输,提出了 piFogBedII 来支持移动雾应用程序的测试。评估结果表明,该方法是有效的,并且迁移策略和容器代理产生的延迟是可以接受的。