DocumentCode :
2372872
Title :
Virtual Brokers for Large-Scale Publish/Subscribe in Wireless Sensor Networks
Author :
Liu, Yang ; Seet, Boon-Chong ; Al-Anbuky, Adnan
Author_Institution :
Dept. of Electr. & Electron. Eng., Auckland Univ. of Technol., Auckland, New Zealand
fYear :
2010
fDate :
11-13 Dec. 2010
Firstpage :
240
Lastpage :
246
Abstract :
Brokers are critical elements for many publish and subscribe (pub/sub) systems. They are dedicated devices with large computation and storage capacities, but they could suffer from scalability problems when comes to actual large-scale implementations in wireless sensor networks. Inspired by the recent virtual MIMO concept for wireless sensor networks, this paper proposes the idea of `virtual brokers´ for large-scale pub/sub systems in sensor networks, and presents in detail the design of a three-stage algorithm for forming a virtual broker around multiple co-located sensor nodes using only network connectivity information. Using computer modeling, we conduct an investigation of our virtual broker formation algorithm and preliminary results demonstrate the potential of our algorithm for scalable formation of virtual brokers.
Keywords :
MIMO systems; message passing; middleware; storage allocation; wireless sensor networks; computer modeling; critical elements; dedicated devices; large-scale implementations; large-scale pub/sub systems; large-scale publish/subscribe; multiple co-located sensor nodes; network connectivity information; publish and subscribe systems; scalability problems; storage capacity; three-stage algorithm; virtual MIMO concept; virtual broker formation algorithm; virtual brokers; wireless sensor networks; connectivity-based Voronoi diagram; connectivity-based k-means clustering; distributed virtual broker; publish; subscribe; wireless sensor network;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Embedded and Ubiquitous Computing (EUC), 2010 IEEE/IFIP 8th International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-9719-5
Electronic_ISBN :
978-0-7695-4322-2
Type :
conf
DOI :
10.1109/EUC.2010.42
Filename :
5703522
Link To Document :
بازگشت