Optimizing Flow Rule Installations on SDN based Switches

Traditional network monitoring involving packet capturing or flow sampling has many challenges such as scalability, accuracy and availability of processing resource when networks become large-scale, high-speed and heterogeneous. SDN is a promising approach to address these challenges, in which highly granular flow rule installations can provide us with fine-grained flow based statistic. But each SDN switch has its own capacity limitation, such as its cache memory called TCAM, which can get exhausted with a large number of highly granular flow rule installations. Thus, network nodes need coordination of resources with other network nodes to monitor the network in a scalable manner. This thesis introduces an intelligent framework, called lite-flow, which divides flow rule installations into two parts, monitoring and forwarding flow rules. The proposed system distributes the load of monitoring flows among SDN switches, and makes the scalability and accuracy of network\ud monitoring manageable. Also, we introduce a forwarding mechanism which uses a more abundant L2 cache in SDN switches based on MAC labels