Comparison of mitigating DDoS attacks in software defined networking and IoT platforms

Abstract

The Software-Defined Networking (SDN) paradigm redefines the term "network" by enabling network managers to programmatically initialize, control, alter, and govern network behavior. Network engineers benefit from SDN's ability to rapidly track networks, centrally manage networks, and quickly and effectively detect malicious traffic and connection failure. The attacker will have total control over the system if he is able to access the main controller. The system's resources can be completely exhausted by Distributed Denial of Service (DDoS) assaults, rendering the controller's services entirely unavailable. The low computational and power capabilities of everyday Internet of Things (IoT) devices render the controller highly susceptible to these attacks; the IoT ecosystem prioritizes functionality over security features, making DDoS attacks a significant problem. This paper conducts a comparative study on the use of machine learning (ML) to mitigate DDoS attack traffic, distinguishing it from benign traffic. This is done to prevent several assaults and to provide mitigation security threats in the network, according to specific requirements. So, the study used machine learning-based techniques to make both traditional and SDN-IoT environments less vulnerable to DDoS attacks. Therefore, the primary goals of the comparative study are to determine which SDN and SDN-IoT platform is better at detecting DDoS attacks and to evaluate how well both platforms work when combined with ML techniques.