Evaluating UDP Traffic and Machine Learning Based Intrusion Detection in Software-Defined Network

摘要

In the modern era, the growth of the Internet is getting wider with enormous impact on traffic and security issues. To ensure high availability of storage, data centers are evolving. In order to maintain the complex operations over the highly interconnected network, data center administratorsrun multiple traffic management tasks. Software Defined Networking is the new technology paradigm shift which reduces the network management complexity. OpenFlow is the communication protocol which is the enabler of software-defined network. The system allows the network administrator to classifya variety of traffics in a network to enhance the security. The projected algorithmic program reduces the number of packets out of order by increasing the hard timeout time of flow entries once the CPU’s usage will increase and by gathering statistics from the switches. The system allowsthe network administrators to classify varied traffics in a network to enhance the security. The svm classifier will classify the normal and attack packets. On classifying attack packets the system will produce the alert to ensure the network security. This work enhances the security in softwaredefined networks.