A hybrid clustering routing protocol based on machine learning and graph theory for energy conservation and hole detection in wireless sensor network. (June 2019)

• Record Type: Journal Article
• Authors: Masoud, Mohammad Z; Jaradat, Yousef; Jannoud, Ismael; Al Sibahee, Mustafa A
• Is Part Of: International journal of distributed sensor networks. Volume 15:Number 6(2019)
• Publication Date: 2019-06
• Subjects: Hole detection -- energy consumption -- wireless sensor network (WSN) -- graph metrics -- soft clustering -- clustering routing protocol; Sensor networks -- Periodicals; Intelligent agents (Computer software) -- Periodicals; Multisensor data fusion -- Periodicals; 681.2
• DOI: 10.1177/1550147719858231
• Languages: English
• ISSNs: 1550-1329

Abstract

In this work, a new hybrid clustering routing protocol is proposed to prolong network life time through detecting holes and edges nodes. The detection process attempts to generate a connected graph without any isolated nodes or clusters that have no connection with the sink node. To this end, soft clustering/estimation maximization with graph metrics, PageRank, node degree, and local cluster coefficient, has been utilized. Holes and edges detection process is performed by the sink node to reduce energy consumption of wireless sensor network nodes. The clustering process is dynamic among sensor nodes. Hybrid clustering routing protocol–hole detection converts the network into a number of rings to overcome transmission distances. We compared hybrid clustering routing protocol–hole detection with four different protocols. The accuracy of detection reached 98%. Moreover, network life time has prolonged 10%. Finally, hybrid clustering routing protocol–hole detection has eliminated the disconnectivity in the network for more than 80% of network life time.