Research on Network Security Collaborative Early-Warning Based on Multi-agent

To solve the lack of collaborative control and whole effect of security modules in the traditional early-warning systems, a novel network security collaborative early-warning technology is studied and a collaborative control framework and early-alert info quick-publish mechanism is proposed. This framework was based on multi-agent, which can make secure modules in the framework be associated with each other to accomplish to communication and work together. Furthermore, two corresponding multicast trees of early-alert information publish are established based on two multi-agent network typology structures: the bottom random overlay network and top hypercube structured overlay network. Simulation results show that this network security collaborative early-warning can make all module functions work normally and reduce this info load and publish time, furthermore, improve the network security system´s collaborative defense capability well.