Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{184389,
        author = {Rushikesh Dalwe},
        title = {Encrypted Traffic Classification Using Deep Neural Networks and Statistical Flow Features},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {1689-1698},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=184389},
        abstract = {The rapid adoption of encryption protocols such as TLS and VPN has rendered traditional port-based and payload-based traffic classification techniques ineffective. This paper presents a framework for encrypted traffic classification leveraging statistical flow features, byte distributions, and protocol metadata. We propose a preprocessing pipeline that converts raw PCAP network traffic into feature-rich tabular datasets and evaluates classification performance using deep neural networks (DNNs) and a baseline Boost model. Experiments on the ISCXVPN2016 dataset demonstrate the high discriminative power of the proposed feature set. The Boost classifier achieved a high accuracy of 93.12% and a macro F1-score of 90.79% on a 12-class task, outperforming the implemented DNN, which attained 82.59% accuracy and a 79.28% macro F1-score. While the method is effective in distinguishing VPN vs. non-VPN flows and major applications, classification performance is challenged by traffic categories with overlapping characteristics such as P2P and streaming.},
        keywords = {},
        month = {September},
        }