Retracted: Design of IoT network using Deep learning model for Anomaly Detection

The exponential growth of IoT (Internet of Things) devices increases the attack surface available to criminals, enabling them to launch potentially more damaging cyberattacks. As a result, the security sector has witnessed a rise in cyberattacks. Since hackers use cutting-edge methods to carry out cyberattacks, many of these operations successfully achieved their malevolent objectives. Neural network models are used by an anomaly-based intrusion detection system (IDS) to identify and categorise attacks in IoT networks. Traditional machine learning techniques seem ineffective in the face of erratic network technologies and numerous infiltration strategies. Deep learning techniques have demonstrated their capacity to correctly identify abnormalities in a variety of academic fields. Because of their efficiency in completing speedier calculations and their capacity to automatically identify key properties in incoming data, Soft Actor Critic are a great solution for anomaly classification and identification. In this research, we propose and create a brand-new intrusion detection model for IoT networks based on anomalies. The IoT-23 intrusion detection dataset is used to validate the proposed model. In comparison to existing implementations, our suggested binary and multiclass classification models have obtained good performance in terms of precision, f1-score, recall and finally accuracy.