Deep Convolutional Neural Network for Active Intrusion Detection and Protect data from Passive Intrusion by Pascal Triangle

Active and passive intrusion are the two types of intrusion. The active intrusion attempts to modify the data and the passive intrusion observes the data and replica. This intrusion creates big damage and congestion in the network. Deep learning techniques have been extensively used to advance intrusion detection systems, which can rapidly and effectively identify and classify intrusions at various levels of networks. These technologies are capable of properly and swiftly identifying threats. Networks, however, require a sophisticated security solution due to the frequent emergence and evolution of hostile threats. Publicly accessible intrusion databases must be updated often due to the intricacy of attacks and constantly varying detection. To solve these issues, Deep Convolutional Neural Network for Active intrusion detection and Protect data from Passive Intrusion by Pascal triangle is introduced. This work proposes a convolutional neural network (CNN) based network intrusion detection model with five convolutional layers. Furthermore, this approach uses the Pascal Triangle method, making the dummy route for protecting the data from passive intrusion. This model is evaluated for binary and multiclass classification using the CICIDS2018 dataset, a publicly available dataset comprising 80 statistical features. The dataset is preprocessed by data transformation and numerical standardization techniques. To evaluate the performance of the suggested system, experiments are carried out. The research findings show that the proposed CNN performs better at detecting multiclass categorization than existing intrusion detection methods, with average accuracy, precision, recall, and F1-score values of 99.16, 99.20, 99.63, and 98.76%, respectively.