Seagull Optimization-based Feature Selection with Optimal Extreme Learning Machine for Intrusion Detection in Fog Assisted WSN

• Published in: Tehnički vjesnik Vol. 30; no. 5; pp. 1547 - 1553
• Main Authors: Muthu, Thiruppathi, Kalimuthu, Vinoth Kumar
• Format: Journal Article Paper
• Language: English
• Published: Slavonski Baod University of Osijek 01-10-2023; Josipa Jurja Strossmayer University of Osijek; Strojarski fakultet u Slavonskom Brodu; Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek; Građevinski i arhitektonski fakultet Osijek; Faculty of Mechanical Engineering in Slavonski Brod, Faculty of Electrical Engineering in Osijek, Faculty of Civil Engineering in Osijek
• Subjects: Analysis; Artificial neural networks; Classification; fog computing; Internet of Things; intrusion detection; Intrusion detection systems; Machine learning; metaheuristics; Network latency; Optimization; security issues; Security software; Sensors; Wireless sensor networks
• ISSN: 1330-3651; 1848-6339
• DOI: 10.17559/TV-20230130000295

On the internet, various devices that are connected to the Internet of Things (IoT) and Wireless Sensor Networks (WSNs) share the resources that they have in accordance with their respective needs. The information gathered from these Internet of Things devices was preserved in the cloud. The problem of latency is made significantly worse by the proliferation of Internet of Things devices and the accessing of real-time data. In order to solve this issue, the fog layer, which was previously an adjunct layer between the cloud layer and the user, is now being utilised. As the data could be retrieved from the fog layer even if it was close to the edge of the network, it made the experience more convenient for the user. The lack of security in the fog layer is going to be an issue. The simple access to sources provided by the fog layer architecture makes it vulnerable to a great number of assaults. Consequently, the purpose of this work is to build a seagull optimization-based feature selection approach with optimum extreme learning machine (SGOFS-OELM) for the purpose of intrusion detection in a fog-enabled WSN. The identification of intrusions in the fog-enabled WSN is the primary focus of the SGOFS-OELM approach that has been presented here. The given SGOFS-OELM strategy is designed to accomplish this goal by designing the SGOFS approach to choose the best possible subset of attributes. In this work, the ELM classification model is applied for the purpose of intrusion detection. In conclusion, the political optimizer (PO) is utilised in order to accomplish automatic parameter adjustment of the ELM technique, which ultimately leads to enhanced classification performance. In order to demonstrate the usefulness of the SGOFS-OELM approach, a number of simulations were carried out. As compared to the other benchmark models that were employed for this research, the suggested SGOFS-OELM models give the best accuracy, which is 99.97 percent. The simulation research demonstrates that the SGOFS-OELM approach has the potential to deliver a good performance in the intrusion detection process.