MANET Performance Analysis Through Gateway and Routing Selection Schemes in Network Disaster Recovery
Mahiddin, Nor Aida
MetadataShow full metadata
In recent years, many natural and human-made disasters have occurred, destroying urban and rural areas worldwide. It can be devastating when existing telecommunication infrastructure is destroyed by disasters. Without telecommunication infrastructure, victims are left isolated from the rest of the world. A mobile ad hoc network (MANET) provides a solution to the problem of connectivity in disaster affected areas. A MANET has the potential to provide quick network connectivity that does not require any infrastructure. However, in disaster recovery areas, MANET performance can deteriorate because of network traffic congestion. The heavy traffic generated by the hand-held and mobile devices of victims and their loved ones can create a bottleneck at the MANET gateway. In this thesis, an empirical investigation of the issues influencing MANET performance is described and the results are reported. The analysis and simulation results show an efficient gateway and routing selection scheme increases network performance, including higher packet throughput, lower packet end-to-end delay, lower packet loss and better packet delivery for medium-to-high traffic loads based on node density and velocity in the disaster recovery area. Performance is improved by introducing an efficient gateway selection scheme to avoid packet congestion at each gateway and manage the load balancing. Another necessary improvement in the routing selection scheme, involving simplifying the route discovery process in MANET, is presented. The effect of node density, the velocity of a node in the disaster recovery area and node pause time, which also influence MANET gateway performance, were investigated. This thesis proposes an efficient gateway load balancing and routing selection scheme (GWRS), which considers the pause time, the density and the velocity of nodes in the disaster recovery area. A GWRS design and deployment scheme is outlined and recommendations for future work are made.