论文部分内容阅读
Aimed at the problem that existing net-flow identification models only take capacity into consideration but lack of unified evaluation standards of analysis of robustness and invisibility, net-flow identification model based on optimization theory is proposed.Firstly, by analyzing existing net-flow identification systems and net-flow transformation process, it defines the net-flow identification model and summarizes net-flow transformation problems.Secondly, the definition and relationship of robustness and invisibility are given.Besides, according to existing attacks, robustness and invisibility of models are classified into three levels, and K-S goodness-to-fit test is used to judge the level of net-flow identification model.Whats more, it proves the existence of capacity.By using optimization theory, it analyzes maximum capacity of model under different conditions of robustness,invisibility and net-flow transformation problem constraints.Finally, simulation experiments verify the correctness of net-flow identification model based on optimization theory.