Security of network is fundamental requirement due to the rapid growth of utilization of network. SDN is nowadays the most preferred evolving networking technology. It provides higher innovation and more integration of services. Including the rapid innovation, also there lies a threat of intrusion in separate planes. Owing to open interfaces present between different planes, risks of intrusion or anonymous traffic inside the network increases. Therefore, on high-traffic networks, monitoring and measurement of traffic is a main area of concern. Several anomaly detection techniques had already been provided for this cause. But still there is a need of efficient anomaly detection methods so that network can work smoothly and intrusion-free with the proper utilization of networking resources. This paper describes a work towards enhancing the efficiency of anomaly detection method while preserving the performance of our network. Also network overhead, response time, and controller workload must be considered while applying monitoring policies. Focus will be on implementing an efficient adaptive flow counting mechanism so that anomaly can be detected dynamically, but the aggregation rules must be modified accordingly.
