Binary Information Press
In this paper, the authors propose an improved PSO-SVM algorithm by introducing a novel inertia weight coefficient equation of PSO to optimize SVM kernel parameters. This optimization mechanism combined the population entropy which is evolved from information entropy and renyi entropy of packet features with the population diversity degree. The notion of normalized population diversity degree is deduced for identifying potential DDoS attacks under the complex network structure. In order to verify the feasibility of the paper method, simulation experiments are carried out in different Vlans of South-Central University for Nationalities.