Cost-Based Placement of Virtualized Deep Packet Inspection Functions in SDN

摘要: In today's IT systems, cyber security requires fine-grained, flexible, adaptable and cost optimized monitoring mechanisms. The emergence of new networking technologies, like Network Function Virtualization (NFV) Software Defined Networking (SDN), opens up venues for large scale adoption these tools. particular, Deep Packet Inspection (DPI) engines can be virtualized dynamically deployed as pieces software on commodity hardware. Deploying such DPI is costly in terms license fees power consumption. Designing effective engine deployment strategies that meet the cybersecurity operational constraints thus mandatory this approach. For purpose, we propose a method, based genetic algorithms, optimizes deployment, minimizing their number, global network load number unanalyzed flows. We conduct several experiments with different types traffic structures. results show method able to reach trade-off between load. Furthermore, reduced 58% when relaxing constraint used link capacity, provisioning rate.