Enhancing Cybersecurity Through Machine Learning-Based Classification of IoT Network Traffic

An exponential increase in smart devices connected to the internet leads to progress in the growth of Internet of Things (IoT) technology that has become integral to part of our daily lives. IoT plays a pivotal role in manufacturing, enabling companies to monitor machine statuses in real-time, assess product quality, and track environmental variables within factories. This not only empowers managers to mitigate risks and minimize losses but also facilitates decision-making from a broader operational perspective. However, the proliferation of IoT also introduces security and safety concerns, making anomaly detection crucial for IoT networks. Detecting anomalies promptly and alerting users is essential to prevent potential damage or losses. In this work, utilizing the Machine Learning and Deep Learning techniques for detecting anomalies in the IoT network and the dataset used to conduct the experimentation on IoT-23 dataset. Finally, identifying the best model for detecting the anomalies in the IoT network by considering the performance of the model and its training time of different models. In the proposed machine and deep learning models, decision tree outperformed with other models in detecting the anomalies with an accuracy of 73% and model time of 7 seconds.