Title
Efficient Edge-Cloud Publish/Subscribe Broker Overlay Networks to Support Latency-Sensitive Wide-Scale IoT Applications.
Abstract
Computing services for the Internet-of-Things (IoT) play a vital role for widespread IoT deployment. A hierarchy of Edge-Cloud publish/subscribe (pub/sub) broker overlay networks that support latency-sensitive IoT applications in a scalable manner is introduced. In addition, we design algorithms to cluster edge pub/sub brokers based on topic similarities and geolocations to enhance data dissemination among end-to-end IoT devices. The proposed model is designed to provide low delay data dissemination and effectively save network traffic among brokers. In the proposed model, IoT devices running pub/sub client applications periodically send collected data, organized as a hierarchy of topics, to their closest edge pub/sub brokers. Then, the data are processed/analyzed at edge nodes to make controlling decisions promptly replying to the IoT devices and/or aggregated for further delivery to other interested edge brokers or to cloud brokers for long-term processing, analysis, and storage. Extensive simulation results demonstrate that our proposal achieves the best data delivery latency compared to two baseline schemes, a classical Cloud-based pub/sub scheme and an Edge-Cloud pub/sub scheme. Considering the similar Edge-Cloud technique, the proposed scheme outperforms PubSubCoord-alike in terms of relay traffic ratio among brokers. Therefore, our proposal can adapt well to support wide-scale latency-sensitive IoT applications.
Year
DOI
Venue
2020
10.3390/sym12010003
SYMMETRY-BASEL
Keywords
Field
DocType
broker-based publish,subscribe,topic similarity,geolocation awareness,IoT applications,latency-sensitive,distributed pub,sub systems
Combinatorics,Software deployment,Latency (engineering),Computer network,Dissemination,Hierarchy,Relay,Overlay network,Mathematics,Cloud computing,Scalability
Journal
Volume
Issue
Citations 
12
1
0
PageRank 
References 
Authors
0.34
0
4
Name
Order
Citations
PageRank
Van-Nam Pham161.78
Vandung Nguyen26210.10
Tri D. T. Nguyen300.34
Eui-Nam Huh41036113.46