Research And Application Of Network Simulation With High Abstraction

Simulation is a key method to do research on network activity and network behavior and it is of great value in research and application. With the rapid development of computer technology and the expansion of networks, there are higher requirements of research and application on simulation. Network simulation research needs to solve a key problem that is in complex networks, how to not only reduce the complexity of network topology, computation overhead and simulation time, but also ensure the simulation fidelity. In addition, there is a key problem in application of network simulation, which is how to apply simulation technology to network security applications such as Intrusion Detection System(IDS).Address to the two problems above, this paper makes a deep research on network simulation methods with high abstraction and optimize the detection model in IDS combined with the simulation platform. So in this paper, the main research work is as follows:1) A network traffic simulation method for topology focusing is proposed. The main idea of this method is partitioning the network topology into Focusing Topology Area(FTA) and Non-Focusing Topology Area(NFTA). Packets are classified into three types according to the area that the routing path belongs to. For each type of packets, the method adopts different simulation strategies. Experiments based on network topology show that this simulation method can reduce the complexity of network and the computation overhead, ensuring the simulation fidelity in FTA, especially the packets simulation fidelity.2) Propose a framework of security application based on high abstract network simulation. Do research on detection model in IDS based on network simulation platform. A novel detection model with Optimized Dempster-Shafer evidence theory(ODS) and Regression Basic Probability Assignment function(RBPA) is presented, that is ODS+RBPA model. In this model, we adding weights for each sensor to optimize DS evidence theory according to its previous predict accuracy.3) ODS+RBPA model is verified in the large-scale complex simulated network with intrusion attack by simulation technology and method. Compared with other models in several experiments, the novel detection model has higher detection rate, lower false rate and stronger stability. Moreover, the model has a better new attack detection performance than others.