Intrusion Detection in Smart Cities Using Machine Learning

Frothy Disturbance Intrusion Detection Systems (FIDSs) hold significant importance in safeguarding against security breaches in the ever-expanding, where the integration of the internet spans diverse domains such as smart homes, healthcare, smart grids, manufacturing, supply chains, and environmental monitoring. Conventional intrusion detection techniques, although useful in traditional settings, need to be modified and improved to address the particular difficulties, such as devices with limited resources, memory that is restricted, batteries that have limited capacity, and certain protocol stacks. In this paper, we propose a lightweight attack detection approach that detects and prevents attackers from trying to introduce unnecessary data into the network by using a supervised machine learning-based FIDS. Our simulation findings show that the suggested FIDS-based classifier performs satisfactorily in terms of classification accuracy and detection time when enhanced with a combination of two or three complex features.