A network self-protection mechanism based on multivariate abnormality analysis

A network self-protection mechanism against network attacks was proposed based on the network self-protection theory and multivariate abnormality analysis. According to PDRR theory model, the main function modules of network self-protection system were designed. By applying multivariate abnormality analysis theory, a flow-based multivariate abnormality analysis network attack detection algorithm was proposed. The algorithm uses a metric of abnormal distance to classify network flow into different types and prioritize the routing of different network flow packets, thus reducing the impact of network attacks against the normal traffic flow. Experimental results demonstrate that the proposed mechanism can significantly protect the network against attacks.