Multi-Objective Genetic Approach of QoS-Based Multicast Routing for Mobile Ad Hoc Networks

Résumé du livre

Messages in Multicast Ad Hoc On-Demand Distance Vector (MAODV) routing protocol are transmitted by flooding which causes traffic overhead. Further mobility breaks those paths that are established by the routing protocol and disrupts the communication, executing an increased rate of link failure. The multicast tree is also not repaired until link breakage occurs. In addition to this, the protocols relying on the source of information (i.e., location information from Global Positioning System (GPS) is unavailable in areas such as tunnels and undergrounds) will fail to operate correctly if this information is not available. Modified MAODV (M-MAODV) is implemented on MAODV which uses only the selected nodes in the mobile ad hoc network (MANET) based on Heading Direction Angle (HDA) (self-content information) to broadcast the messages. The way the nodes are selected exploits the mobility and a long-lived route is also established to a destination. The work is further extended by applying multi-objective Genetic Algorithm (GA) to optimize the path established by M-MAODV based on Quality of Service (QoS) constraints. Simulation results of Genetic-MAODV (G-MAODV) show that the performance has been further enhanced satisfying the QoS issues in addition to the establishment of long-lived route.