Simulation And Research Of AOMDV Protocol Based On Path-State In Ad Hoc Networks

Because of its fast start, independence on infrastructure and other characteristics, Ad Hoc network, as a non-center, temporary and multi-hop dynamic autonomous system, has been widely applied in military field, disaster relief, ad hoc meetings and other occasions,. But its topology structure’s quick change, network bandwidth and limited node energy have resulted in an unreliable route and unstable characteristics, so routing technology has become the key technology of Ad Hoc network.In a route discovery process, multi-path routing protocol establishes a number of paths from the source node to the destination node. Multi-path routing protocol has better performance in terms of network reliability, delay nature power consumption and other aspects than the single-path routing protocol does. And especially in the case of the frequent network topology changes, the advantage of it is more evident. Moreover, in terms of energy balance, load balance and QoS services, multi-path routing also possesses good extending ability.This paper, firstly, researches and analyses deeply on Ad Hoc network routing protocol, comparing and analyzing AODV protocol and AOMDV protocol, to the questions caused by the on-demand distance vector routing protocol’s over-reliance on the short circuit forwarding data to make the node energy and load imbalance, the thesis proposes an improved protocol based on the path AOMDV, and studies on the simulator. And then, through acquiring and dynamically updating the energy of path and load status information, the paper combines with the fitness of path hop count and selects the path forwarding data with more fitness to achieve the purpose of energy balance and load balance. The result of simulator shows, in the aspects of the packet delivery fraction, end-to-end delay, node energy balance and so on, the improved protocol becomes better. This proves the effectiveness of the improvement. Finally, this thesis makes a preliminary analysis of the effect on the network performance by the selfishness of the nodes, and proposes the direction of further research.