Study On Ad Hoc Network On Demand Multipath Distance Vector Routing Protocol

Ad hoc network is a collection of wireless mobile nodes without any fixed base station infrastructure and centralized management. Due to the limited transmission range of the nodes in ad hoc network, when data packets are sent from a source node to a destination node, several other nodes are usually used as transmission relays. Therefore, routing protocol is indispensable in ad hoc network.At present, routing protocol is one of the resarch focuses and noduses in ad hoc network, various routing protocols have been proposed. In this paper, ad hoc on-demand multipath distance vector routing protocol (AOMDV) is studied. The two important attributes, loop freedom and disjoint paths, and details of AOMDV are described in detail. The key steps of adding AOMDV to NS2 are produced. Simulation environment is set to simulate AOMDV and AODV (Ad hoc On-demand Distance Vector Routing). The following four performance metrics, packet loss percentage, average end-to-end delay of data packets, route discovery frequency and routing overhead, are used to evaluate the effectiveness of AOMDV relative to AODV.An optimization scheme of AOMDV is also presented to set up the omitted reverse paths caused by the "route cutoff problem. The key step of the optimization scheme lies in: whenever receiving a RREP, the source node will compare the IP address of first-hop node of the RREP's corresponding RREQ with the IP address of its neighbor node from which it receives the RREP. If they are not equal, it means that one reverse path has been omitted. If so, the source node will transmit a new control packet called RREP_ACK along the forward path set up by this RREP to the destination node. Every medium node receiving the RREP_ACK will set up a reverse path for the the source node. Finally, the omitted reverse path will be completely set up by the destination node. The simulation results show the superiority of the optimized AOMDV relative to AOMDV in packet loss percentage, average end-to-end delay of data packets, route discovery frequency and routing overhead. So the effectiveness of the optimization scheme is proved.