SmartDefend - IoT Security Using Machine Learning

With the advent of the fourth industrial revolution in recent years, technological advancements have led to massive exponential growth in the Internet of Things (loTs), fog computing, computer security, and cyberattacks. One of the most promising techniques for mitigating cybersecurity risks and ensuring security is machine learning, or ML. In order to address the current security and privacy issues, a number of research have suggested smart intrusion detection systems (IDS) using intelligent architectural frameworks that use AI. This paper presents a systematic literature review (SLR) that maps, surveys, and categorises the body of research on machine learning techniques for identifying cybersecurity threats in Internet of Things environments. Notably, random forests (RF) and support vector machines (SVM) are two of the most popular techniques because of their effective memory usage and excellent detection accuracy. The system aims not only to detect anomalies but to delve deeper, providing detailed insights through the precise classification of detected anomalies into specific attack types. Among the seven algorithms examined, the Random Forest Classifier yields the best accuracy of 92.72% in binary classification and 92.40% in multiclass classification.