An Artificial Intelligence (AI) Framework for Detection of Distributed Reflection Denial of Service Attacks

Abstract

In the contemporary digital world, cyber space is growing continuously witnessing amalgamation of different technologies associated with telecommunications, networking and sensing to mention few. This has enabled Service Oriented Architecture (SOA) to realize distributed applications that cater to the needs of enterprises in the real world. With the advantages of such environments, there has been increased number of instances of cyber-attacks. Distributed Denial of Service (DDoS) is the large-scale attack targeting critical digital infrastructure to make it useless for certain amount of time. Such attacks have several implications and lead to collapse of businesses unless there are countermeasures to detect it and handle it properly. Distributed Reflection Denial of Service (DRDoS) is a variant of such attacks which is more destructive in nature. It is more so in the presence of Internet of Things (IoT) devices deployed in cyber space in large scale. The existing DDoS countermeasures do not work to solve the problem of DRDoS directly. We propose an Artificial Intelligence (AI) framework for detection of DRDoS attacks. We propose an algorithm known as Machine Learning based DRDoS Attack Detection (ML-DAD) for effective detection of attacks. The prototype service built in Python monitors such attacks and take necessary steps to defeat it. The empirical results revealed that the proposed framework has superior performance improvement over the stat of the art. The research in this paper leads to new ideas in the area of detection and prevention of DRDoS attacks.