Optimized bi-connected federation of multiple sensor network segments

Sookyoung Lee, Mohamed Younis, Meejeong Lee

Research output: Contribution to journalArticlepeer-review

13 Scopus citations

Abstract

Wireless sensor networks (WSNs) serving in hostile environments are susceptible to simultaneous failure of multiple collocated nodes that cause the network to split into distinct segments. Restoring connectivity would be needed to resume full WSN operation. Similar scenario is encountered when autonomous WSNs have to collaborate for achieving a common task. Proactively provisioning spare connections would be also desirable to prevent possible partitioning caused by a single node failure in the federated topology. In this paper, we present an optimized strategy for establishing 2-vertex distinct paths between every pair of segments or individual WSNs while minimizing the number of the deployed relay nodes (RNs) and increasing the efficiency of the resulting bi-connected topology in terms of average node degree. The proposed Optimized Bi-Connected federation of multiple sensor network segments (OBiC) algorithm strives to form a single simple cycle which visits every segment exactly once. The least number of relays are populated along the steinerized edges of the formed bi-connected topology by the cycle and thus 2-vertex inter-segment disjoint paths are guaranteed. We analyze the properties of OBiC mathematically and validate its performance through extensive simulation experiments. The validation results show that OBiC yields highly-connected topologies with short inter-segment paths while employing fewer RNs than competing schemes.

Original languageEnglish
Pages (from-to)1-18
Number of pages18
JournalAd Hoc Networks
Volume38
DOIs
StatePublished - 1 Mar 2016

Keywords

  • 2-vertex connectivity
  • Fault tolerance
  • Network partitioning
  • Relay node placement
  • Topology repair
  • Wireless sensor networks

Fingerprint

Dive into the research topics of 'Optimized bi-connected federation of multiple sensor network segments'. Together they form a unique fingerprint.

Cite this