Smart Grid and Innovative Frontiers in Telecommunications. Third International Conference, SmartGIFT 2018, Auckland, New Zealand, April 23-24, 2018, Proceedings

Research Article

Efficient Fault Identification Protocol for Dynamic Topology Networks Using Network Coding

  • @INPROCEEDINGS{10.1007/978-3-319-94965-9_23,
        author={Hazim Jarrah and Peter Chong and Nurul Sarkar and Jairo Gutierrez},
        title={Efficient Fault Identification Protocol for Dynamic Topology Networks Using Network Coding},
        proceedings={Smart Grid and Innovative Frontiers in Telecommunications. Third International Conference, SmartGIFT 2018, Auckland, New Zealand, April 23-24, 2018, Proceedings},
        proceedings_a={SMARTGIFT},
        year={2018},
        month={7},
        keywords={Self-diagnosis Dynamic networks Dynamic fault RLNC},
        doi={10.1007/978-3-319-94965-9_23}
    }
    
  • Hazim Jarrah
    Peter Chong
    Nurul Sarkar
    Jairo Gutierrez
    Year: 2018
    Efficient Fault Identification Protocol for Dynamic Topology Networks Using Network Coding
    SMARTGIFT
    Springer
    DOI: 10.1007/978-3-319-94965-9_23
Hazim Jarrah1,*, Peter Chong1,*, Nurul Sarkar1,*, Jairo Gutierrez1,*
  • 1: Auckland University of Technology
*Contact email: Hjarrah@aut.ac.nz, peter.chong@aut.ac.nz, nurul.sarkar@aut.ac.nz, jairo.gutierrez@aut.ac.nz

Abstract

This paper considers the problem of fault identification in dynamic topology networks using the time-free comparison model. Here, we introduce an efficient self-diagnosis protocol that can identify faulty nodes in dynamic networks. This protocol can correctly diagnose various fault types including permanent, dynamic, and soft faults. The protocol consists of a testing stage and a disseminating stage. During the testing stage, each node identifies the state of a part of nodes using the time-free comparison model. Afterward, nodes share their views employing a random linear network coding (RLNC) technique in the disseminating stage. The design of the disseminating stage is crucial for diagnosis efficiency. Using RLNC obviates the need for disseminating the views individually, and hence it reduces the number of messages required to diagnose the network. The OMNeT++ simulation has been used to evaluate the performance of the proposed protocol regarding the communication complexity. Results show that the proposed protocol is robust, scalable and energy-efficient.