Flow Feature-Based Network Traffic Classification Using Machine Learning

Reliable network traffic classification is essential to management and security tasks. Therefore, it beneficial analyze improve existing techniques. Some of the most traditional methodologies for are based on port number packet payload, each which presents an increasing set problems. Port number-based techniques suffer from misuse numbers tunneling. This primarily due their reliance proper use IANA (Internet Assigned Numbers Authority) assigned numbers. On other hand, payload-based has difficulty dealing with encrypted data legal restrictions accessing user data. Flow feature-based canovercome these challenges by creating profiles patterns applications. Furthermore, machine learning have shown be a good match classification. Thus, goal this paper explore combination fields develop models capable classifying flow features. was achieved using ready dataset train two supervised one unsupervised clustering model.The results classifiers were considered comparable similar studies, while performance model found not satisfactory.