Real-Time Distributed Service De-Identification And Internet Problem Mitigation

Distributed denial-of-service (DDoS) attacks are increasing in size, and complexity, and becoming more frequent where volumetric or high-rate attacks are the most prevalent type of DDoS penetrations. Cyber-attack and breaches against the networks of the Internet of Things (IoT) have dire consequences for the data and services that are used in the applications and devices while at the same time increasing the security challenges for its departments due to the increased resources required. A DDoS attack is an endeavor aimed at making a network, online service, or a whole organization unavailable by bombarding to saturate its traffic from multiple sources. The software-defined paradigm has been utilized as an effective mitigation technique for DDoS in IoT. Most researchers and experts have embraced this technique; therefore, in this study, we will present the conventional framework of software-defined detection and mitigation techniques for IoT. Furthermore, we will advance an algorithm for detecting and alleviating DDoS attacks that can be easily interfaced with software-defined frameworks (SD-IOT) for IoT. This algorithm will be based on neural networks for the packet in and out traffic while monitoring for DDoS attacks. The technique was successfully tested and revealed 89.047% accuracy for identifying actual DDoS attack datasets.