Gradient Boost-Based Network Intrusion Detection in SDN-IoT Networks

The growth of computer applications has increased the importance of Internet of Things networks in daily life. The Internet of Things (IoT) technology currently under development enables portable sensor networks. Over time, addressing the network's security issues and safeguarding them becomes a more crucial concern. The Internet of Things (IoT) and a Software-Defined Network (SDN) are being coupled to safeguard the environment from possible threats. The SDN-IoT network is a centralized, regulated network. Users benefit from flexible device connections. The benefits of the SDN-IoT network, conversely, make it crucial to safeguard the network's security against problems such as network failure, system lockdown, and thefts, and build its defenses against intrusions. SDN-powered IoT networks are susceptible to Distributed Denial of Service (DDoS) attacks, which, if initiated, may quickly disrupt the centralized system. In the suggested study, using an intrusion detection system, a range of machine learning models are investigated for an SDN-IoT network to have a stable, secure network to support future smart networks. Performance is increased by evaluating the output of various Machine Learning (ML) methods. Network intrusions are detected using ML approaches that combine binary and multiclass classification. A multi-class categorization is used to identify the kind of attack, such as DoS, DDoS, Port Scanning, OS Fingerprinting, and Fuzzing. XGBoost performs well with a high accuracy of 98.81% in identifying intrusion and 98.80% in detecting multiclass intrusions in the proposed model using multiple ML approaches.