Event-to-sink directed clustering in wireless sensor networks

  • Authors:
  • Alper Bereketli;Ozgur B. Akan

  • Affiliations:
  • Next Generation Wireless Communications Laboratory, Department of Electrical and Electronics Engineering, Middle East Technical University, Ankara, Turkey;Next Generation Wireless Communications Laboratory, Department of Electrical and Electronics Engineering, Middle East Technical University, Ankara, Turkey

  • Venue:
  • WCNC'09 Proceedings of the 2009 IEEE conference on Wireless Communications & Networking Conference
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Wireless sensor networks (WSN) are event-based systems based on the collaboration of several microsensor nodes. Due to the limited supply of energy at sensor nodes, energy-efficient configuration of WSN has become a major design goal to improve the lifetime of the network. Many clustering algorithms have been proposed as energy-efficient, however, the existing classical pre-event clustering solutions form clusters in the entire network unnecessarily that brings significant overheads in maintaining the network configuration. Unlike preevent clustering, energy-efficient operation of WSN requires the event-to-sink directed clustering notion, which forms clusters when and where they are needed and in the direction of data flow from event location to the sink. To the best of our knowledge, energy-efficient clustering in WSN has not been studied from this perspective before. In this paper, we propose a novel Eventto-Sink Directed Clustering (ESDC) protocol for WSN. ESDC realizes energy efficiency in sensor network configuration by employing two techniques: (1) clustering of the nodes only within the event-to-sink data flow corridor to avoid unnecessary cluster formation, (2) directional clustering to minimize the number of hops for data forwarding. The directional clustering process in ESDC also sets up the routing path of the event flows over the clusters. Performance results reveal that the ESDC protocol achieves the energy-efficiency objectives and outperforms the existing conventional pre-event clustering approaches.