A Scalable Intrusion Detection Framework for Industrial IoT Using Distributed DNNs and Extreme Learning Machines

The development of a system for intrusion detection (IDS) that can automatically identify and categorize cyber-attacks at the host and network levels is heavily reliant on machine learning methods. Nevertheless, no prior research has shown a comprehensive evaluation of the efficacy of various algorithms for machine learning on a variety of openly accessible datasets. In order to create adaptable and efficient intrusion detection systems (IDS) that can identify and categorize unexpected and unanticipated cyber-attacks, this article investigates deep neural networks (DNNs), a kind of deep learning model. It is essential to assess different datasets produced over the years using static and dynamic methods due to the ever-changing nature of network behavior and the quick development of attacks. Research like this helps find the most effective algorithm for spotting cyberattacks in the future. We provide a thorough analysis of experiments conducted on many publicly accessible benchmark malware datasets using DNNs and other traditional machine learning classifiers. By feeding the IDS data into its many hidden layers, our DNN model is able to learn its abstract and highdimensional feature representation. It has been shown via extensive experimental testing that DNNs outperform standard machine learning classifiers. In conclusion, we provide Scale-Hybrid-IDS-AlertNet (SHIA), a framework for hybrid DNNs that can be utilized in real-time to successfully monitor host-level events and network traffic in order to proactively notify potential cyber-attacks. In order to achieve higher accuracy with less execution time, we are also using parallel processing and the Extreme Machine Learning method.