An Ecient Framework for DDoS Attack Detection in IoT Networks Leveraging Software-Defined IoT(SD-IoT)

Abstract

The Internet of things (IoT) introduces emerging applications (i.e., smart homes, smart cities, smart health, and smart gird) that assist the traditional infrastructure environ ments to be connected with smart objects. Things are connected with the Internet, and numerous new IoT devices are developing at a rapid pace. As these smart objects are connected and able to communicate with each other in unprotected environments; there fore, the whole communication ecosystem requires security solutions at different levels. There is a dangerous hazard to Internet networks and even to individual cyber-physical systems that are also connected to the Internet, with billions of such devices already on the market that have major vulnerabilities. IoT technology possesses unique characteris tics with various resource constraints and heterogeneous network protocol requirements, unlike traditional networks. The attacker exploits numerous security vulnerabilities of an IoT infrastructure, to generate a Distributed Denial of Services (DDoS) attack. The increase in DDoS attacks has made it essential to address the consequences which imply in the IoT industry. This research proposes an efficient, Software-Defined Internet of Things (SD-IoT) based framework that provides security services to the IoT network. We developed a novel framework for DDoS attack detection in SD-IoT networks leverag ing SD-IoT. The proposed framework is based on three different modules Counter-based DDoS Attack Detection (C-DAD), Log-based DDoS Attack Detection (L-DAD), and Machine-learning Based DDoS Attack Detection (M-DAD) to detect DDoS attack suc cessfully. These three modules are based on counter values, log file, and supervised learning classifier using different network parameters. The framework is a dynamic and programmable solution and is deeply tested with different network parameters. The algorithms demonstrate good performance with better results through Software-Defined Networking (SDN).The C-DAD and L-DAD detection approaches have 97% accuracy, and M-DAD has 99% accuracy; meanwhile, overall 98% accuracy performance of the proposed is noted. Moreover, the proposed framework detects the attack efficiently in a minimum amount of time and with lesser CPU and memory resources consumption. Keywords: SD-IoT, SDN, Attack Detection, DDoS, Counter-Based DDoS Detection, Machine Learning, Hybrid approach