DRaNN_PSO: A deep random neural network with particle swarm optimization for intrusion detection in the industrial internet of things

The Industrial Internet of Things (IIoT) is a rapidly emerging technology that increases the efficiency and productivity industrial environments by integrating smart sensors devices with internet. advancements in communication technologies have introduced stable connectivity higher data transfer rate IIoT. IIoT generate massive amount information requires intelligent processing techniques for development cybersecurity mechanisms. In this regard, deep learning (DL) can be an appropriate choice. This paper proposes Deep Random Neural Network (DRaNN) based fast reliable attack detection scheme environments. RaNN advanced variant traditional Artificial (ANN) highly distributed nature better generalization capabilities. To attain accuracy, proposed optimally trained incorporating hybrid particle swarm optimization (PSO) sequential quadratic programming (SQP). SQP-enabled PSO facilitates neural network to select optimal hyperparameters. efficacy suggested analyzed both binary multiclass configurations conducting extensive experiments on three new datasets. experimental outcomes demonstrates promising performance design all