Abstract
The security of linked devices and systems has become a top priority due to the Industrial Internet-of-Things' (IIoT) rapid expansion. The identification and prevention of any intrusions that might compromise the availability and integrity of IIoT networks is one of the major difficulties in this field. The exploration of Deep Learning (DL) architectures for Intrusion Detection Systems (IDS) in IIoT contexts has been driven by their promising findings in a variety of cybersecurity applications. This survey explores and evaluates the current deep learning architectures utilized for IIoT intrusion detection in order to provide an overview of them. It also points out possible areas that need improvement. This article evaluates the durability, performance, and adaptability of several deep learning (DL) methodologies, including hybrid architectures, recurrent-neural-networks (RNNs), deep-neural-networks (DNNs) and convolutional-neural-networks (CNNs), in the context of IIoT environments.