Provenance-aware security risk analysis for hosts and network flows

摘要: Detection of high risk network flows and hosts is becoming ever more important challenging. In order to selectively apply deep packet inspection (DPI) one has isolate in real time activities within a huge number monitored flows. To help address this problem, we propose an iterative methodology for simultaneous assessment scores both The proposed approach measures the interdependent manner; thus, score flow influences its source destination hosts, also host evaluated by taking into account initiated or terminated at host. Our experimental results show that such not only effective detecting but, when deployed throughput networks, efficient than PageRank based algorithms.