• DocumentCode
    558729
  • Title

    Cross-layer cluster-based data dissemination for failure detection in MANETs

  • Author

    Kidston, David ; Li, Li ; Al Mamun, W. ; Lutfiyya, Hanan

  • Author_Institution
    Canada Commun. Res. Centre, Ottawa, ON, Canada
  • fYear
    2011
  • fDate
    24-28 Oct. 2011
  • Firstpage
    1
  • Lastpage
    7
  • Abstract
    Node failures may be frequent in MANETs, but there can be many different causes for those failures. Nodes may lose power, crash, or simply move out of range of other nodes in the network. Identifying the root cause is complicated by a lack of fixed monitoring and analysis infrastructure. Past research has focused on monitoring using either ping, heartbeat, or gossip-based approaches, which can incur significant network wide overhead. This paper proposes a novel k-hop cluster based data dissemination scheme that can piggyback on routing messages for more efficient detection of failures including node disconnection. In this scheme, nodes forward their neighbour-hood observations to a per-cluster failure detector based on the observed spanning tree. Simulations show that detecting disconnected nodes using a cross-layer implementation of the data dissemination scheme is more efficient while an application layer implementation is faster. This effect is more pronounced in sparse networks.
  • Keywords
    mobile ad hoc networks; telecommunication network management; telecommunication network routing; trees (mathematics); MANET; cross-layer cluster-based data dissemination; failure detection; k-hop cluster; neighbour-hood observation; node disconnection; node failure; routing message; spanning tree; sparse network; Ad hoc networks; Data models; Detectors; Heart beat; Mobile computing; Protocols; Topology; MANET; cross-layer; data dissemination; failure detection; network efficiency;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Network and Service Management (CNSM), 2011 7th International Conference on
  • Conference_Location
    Paris
  • Print_ISBN
    978-1-4577-1588-4
  • Electronic_ISBN
    978-3-901882-44-9
  • Type

    conf

  • Filename
    6104019