Published Paper Details:

As network services become more widely utilized, security becomes one of the network's primary and most pressing challenges. Several computers connecting to the network serve critical roles in business and other applications that deliver network services. Consequently, we must look for the most effective measures to protect the system, even though several studies employ deep learning (DL)-based intrusion detection algorithms to identify infiltration. Changes in network traffic may result in reduced accuracy of deep learning-based models regarding network assaults. Deep learning has a plethora of strategies for network attacks. The primary goal of this work is to compare and assess the performance of deep learning algorithms on the 1-D data set. The suggested application uses the gated recurrent units(GRU)1-Dimensional Convolutional Neural Network(1-DCNN) hybrid model to identify network attack rates. The preprocessed dataset is trained and tested with the models to provide notable results that increase prediction accuracy. The experiment was conducted using the KDD99 dataset. The deep learning-based hybrid model could reach an accuracy of about 97 percent. On this standard dataset, the GRU-1DCNN model now outperforms the previous model.