A self-similar traffic generation method and application in the simulation of mobile ad-hoc network

Based on the model of assembling of multiple ON/OFF sources, a new implementation was designed to create self-similar traffic by using OPNET simulation software. It can generate self-similar network traffic effectively through some ameliorations of the data source module in OPNET. And using ldquovariance-timerdquo technology to verify the self-similar sequence generated by above method. Furthermore, this implementation was applied to the simulation of ad-hoc network. By comparing with the simulation results of variety network performance indexes in, the results has shown that the self-similarity traffic has great influence on ad-hoc network performance, which leads to a sharp contrast with the short-range dependence and weak long-range dependence of the traditional flow model. A conclusion is made that the higher the self-similar degree, the worse the network performance. The method presented is useful for modeling high accuracy of ad-hoc network traffic, improving network performance and optimizing resources configuration.