DLIDF: DEEP LEARNING BASED INTRUSIONDETECTION FRAMEWORK IN INDUSTRIALINTERNET OF THINGS

The rapid growth of Internet of Things (IOT) equipment in areas such as healthcare, transportation and smart homes has raised important concerns about data security and network integrity. These devices produce vast amounts of sensitive data and are unsafe for cyber threats due to their distributed nature, limited resources and weak security protocols. Traditional infiltration detection systems (IDs), including signature-based and traditional machine learning models, often fail to detect advanced or novel attacks. To remove these boundaries, we propose a 2D Convolutional Neural Network (CNN) Architecture based on a deep learning infiltration system (DL-IDS). This model effectively removes complex features from IOT traffic to identify abnormal behavior with high precision. UNSW-NB15 was evaluated using a dataset, including diverse and realistic attack landscapes, the proposed DL -DS displays strong performance. Our results indicate that this approach provides a more strong and responsive solution to achieve IOT network against developing cyber threats.