Cyber Attack Detection and Defense for the IoT Devices

As more and more Internet of Things (IoT) devices are included into Cyber-Physical Systems (CPS), it is critical to maintain the security of these networked systems.This work focuses on using Support Vector Machine (SVM) models to create and deploy a cyber attack detection system for IoT-based CPS.The SVM model is trained on historical data to discern between normal and harmful behavior patterns displayed by IoT devices within the CPS.The SVM model is well-known for its efficacy in binary classification tasks.Through training to understand the system's typical behavior, the SVM model gains the ability to recognize variations that could be signs of a cyber-attack.Simulations and real-world tests show how well the SVM-based detection system works to identify different kinds of cyber-threats.This study does, however, admit several shortcomings.The dynamic and changing nature of cyber threats may have an effect on the SVM model's performance because it mostly depends on previous data for training and detection accuracy problems.In this study work, a novel CNN model based on deep learning was proposed in order to overcome the shortcomings of the current cyber threat detection model.The suggested model enhances performance metrics and the detection of cyber-attacks.When compared to the existing model, the new model performs better.The f1-score, accuracy, precision, and recall metrics are used to gauge performance.