LAUSR

Software Defined Network (SDN) is a flexible paradigm that provides support for a variety of data-intensive applications with real-world smart Internet of Things (IoT) devices. This emerging architecture updates with the managing ability and network control. Still, the benefits are challenging to achieve due to the presence of intruder flow into the network. The research topic of intrusion detection and prevention system (IDPS) has grasped the attention to reduce the effect of intruders. Distributed Denial of Service (DDoS) is a targeted attack that develops malicious traffic is flooded into a particular network device. These intruders also involve even with legitimate network devices, the authenticated device will be compromised to inject malicious traffic. In this paper, we investigate the involvement of intruders in three-Tier IDPS with regard to user validation, packet validation and flow validation. Not all the authentication users can be legitimate, since they are compromised, so that the major contribution is to identify all the compromised devices by knee analysis of the packets. Routers are the edge devices employed in first tier which is responsible to validate the IoT user with RFID tag and encrypted signature. Then the authenticated user’s packets are submitted into second tier with switches that validates the packets using type-II fuzzy filtering. Then the key features are extracted from packets and they are classified into normal, suspicious and malicious. The mismatched packets are analyzed in controllers which maintain two queues as suspicious and normal. Then suspicious queue packets are classified and predicted using deep learning method. The proposed work is experimented in OMNeT++ environment and the performances are evaluated in terms of intruder Detection Rate, Failure Rate, Delay, Throughput and Traffic Load.