Design And Implementation Of Self-organizing Qos Routing Mechanisms Based On Antnet And Beehive

The current network model and routing mechanisms are challenged by the expanding scale and increasing dynamic of network. The expanding scale requires the network to be self-organized and self-managed. The increasing dynamic require the routing mechanisms to be distributed and self-adapted, the nodes in the network don't maintain global information. The phenomenon of self-organization is pervasive in many areas of daily life; the individual entities in a self-organized system interact directly with each other in a distributed peer-to-peer fashion. Simple behavior of the individual entities can make the complicated goal of the overall system, and the individual entities continuously adapt to changes in a coordinated manner.In this thesis, a SON (self-organizing network) model is designed and routing mechanisms with QoS supported are designed, including unicast and multicast routing, and also a simple rerouting mechanism. The way about how local information is exchanged is dedigned, and the behavior of the node is designed such as migration, hibernation and clone. In order to adapt the feature of SON, algorithms are designed base on two distributed routing algorithms that are AntNet algorithm and BeeHive algorithm. The two algorithms can both be used to find paths in the network hop by hop, and the next hop selection policies are designed, and the two algorithms are expanded to support QoS-constrainted routing. In order to reduce the time taken by routing, small-world behavior is introduced. Fuzzy mathematics is introduceed to describe the imprecise link parameters. Nash-based and Pareto optimum-based microeconomic theory is introduced to make link Policy selection.In this thesis, the SON network model and the routing mechanisms are simulated on NS2 simulating platform, and the routing mechanisms are also implemented and deployed. The performances are analyzed, and the result show that the designed SON network model has some advantages compared to current network model, and the routing mechanisms have good performance and practicability.