Classification of Network Traffic Based on Improved Convolution Neural Network

In recent decades, the development of network dimensions and the evolution of communications technology have rendered precise and efficient traffic classification an essential priority for network administrators and operators in order to ensure the quality of service (QoS) and network security. The dynamic character of contemporary networks renders traditional methods for classifying traffic networks ineffective. The paper employed network deep learning methods, which are represented in an enhanced convolutional neural network (CNN) known as LCNN, to classify traffic networks. LCNN is the name of the algorithm that combines two technical CNNs with principal component analysis (PCA) to reduce the dimension of data. In the initial phase of the problem, the paper relied on the CICIDS 2017 dataset for preprocessing; feature extraction using PCA, and classification using LCNN in the final step. Subsequently, traffic categorization was investigated and examined during the online implementation. Such as the expenses associated with online deployment, dataset, and zero-day application verification, the performance of LCNN overcame other technical factors, with accuracy reaching 96%. Encountered by contemporary researchers in the classification of network traffic is ultimately addressed in the performance metrics of network classification