ETC-IoT: Edge-Node-Assisted Transmitting for the Cloud-Centric Internet of Things
ETC-IoT: Edge-Node-Assisted Transmitting for the Cloud-Centric Internet of Things
复制标题
ETC-IoT：以云为中心的物联网的边缘节点辅助传输
DOI：
10.1109/mnet.2018.1700164
复制
发表时间：
2018
期刊：
影响因子：
9.3
通讯作者：
Hara Takahiro
中科院分区：
文献类型：
--
作者：
Zhao Wei;Liu Jiajia;Guo Hongzhi;Hara Takahiro
云系统的出现可以为以云为中心的物联网应用程序在平台管理和服务质量改进方面带来显着的好处。边缘或雾计算，其中边缘节点主要自愿贡献其计算资源，由于其能够克服可扩展性问题并减轻传统以云为中心的物联网架构中的云计算服务器的负载而引起了极大的关注。然而，由于高数据速率设备在物联网中已经无处不在，并且需要将大量数据传输给最终用户，因此通过从商业云租用服务来提供这样的平台是昂贵且低效的。因此，本文描述了一种在以云为中心的物联网架构（ETC-IoT）中进行边缘节点辅助数据传输的新方法，以克服云中带宽消耗过大的问题。它通过扩展以云为中心的物联网架构中的边缘计算的现有范例来探索边缘节点的带宽资源。具体地，中心云根据设计的副本分发策略将物联网数据的副本发送到边缘节点。然后，拥有副本的边缘节点可以通过使用其带宽资源来响应来自其相关联的本地最终用户的多个请求，而如果边缘节点不能容纳请求，则中央云补充任何不足以满足最终用户的需求。
The promising emergence of cloud systems can bring significant benefits to cloud-centric IoT applications in terms of both their platform management and improvements to quality of service. Edge or fog computing, where edge nodes mainly contribute their computing resources voluntarily, has been attracting great attention for their ability to overcome scalability problems and to alleviate the load of cloud computing servers in the traditional cloud-centric IoT architecture. However, since high-data-rate devices have become ubiquitous in IoT and huge amounts of data need to be transmitted to end users, providing such a platform by renting a service from commercial clouds is costly and inefficient. Thus, this article describes a novel approach for edge-node-assisted data transmission in the cloud-centric IoT architecture (ETC-IoT) to overcome the problem of overwhelming bandwidth consumption in the cloud. It explores the bandwidth resources of edge nodes by extending the existing paradigm of edge computing in the cloud-centric IoT architecture. Specifically, the central cloud sends the replicas of IoT data to edge nodes according to designed replica distribution strategies. Then an edge node owning the replicas can respond to multiple requests from its associated local end users by using its bandwidth resources, while the central cloud supplements any shortfall to satisfy end-user demands if edge nodes cannot accommodate requests.