Real-Time Encrypted Traffic Classification with Deep Learning

Confidentiality requirements of individuals and companies led to the dominance of encrypted payloads in the overall Internet traffic. Hence, traffic classification on a network became increasingly difficult as it must rely on only the packet headers. Many vital tasks such as differential pricing, pr...

Full description

Saved in:
Bibliographic Details
Main Authors: Onur Demir, Deniz Tuana Ergönül
Format: Article
Language:English
Published: Sakarya University 2022-04-01
Series:Sakarya Üniversitesi Fen Bilimleri Enstitüsü Dergisi
Subjects:
Online Access:https://dergipark.org.tr/tr/download/article-file/2092192
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850133738408640512
author Onur Demir
Deniz Tuana Ergönül
author_facet Onur Demir
Deniz Tuana Ergönül
author_sort Onur Demir
collection DOAJ
description Confidentiality requirements of individuals and companies led to the dominance of encrypted payloads in the overall Internet traffic. Hence, traffic classification on a network became increasingly difficult as it must rely on only the packet headers. Many vital tasks such as differential pricing, providing a safe Internet for children, and eliminating malicious connections require traffic classification, even if the payload contents are encrypted. Encrypted traffic is harder to classify as packet content becomes unreadable. In this work, we aim to provide an insight into traffic classification using encrypted packets in terms of both accuracy and packet processing time. LSTM (Long Short-Term Memory) architecture is a good candidate for this problem as it can handle sequences. Each flow can be modeled as a sequence and patterns of the sequences can provide valuable information. We compare the performance of LSTM with other methods in both real-time and offline experiments. Compared to a machine learning method both online and offline LSTM excelled with precision and recall differences up to 50%. Average accuracy with LSTM was measured as 97.77% offline and 91.7% in real-time. Average packet processing time in real-time was recorded as 0.593 msec which is 5 times faster than a recent work that uses LSTM method.
format Article
id doaj-art-51fc734a291342b8b91c10a66982aab6
institution OA Journals
issn 2147-835X
language English
publishDate 2022-04-01
publisher Sakarya University
record_format Article
series Sakarya Üniversitesi Fen Bilimleri Enstitüsü Dergisi
spelling doaj-art-51fc734a291342b8b91c10a66982aab62025-08-20T02:31:52ZengSakarya UniversitySakarya Üniversitesi Fen Bilimleri Enstitüsü Dergisi2147-835X2022-04-0126231333210.16984/saufenbilder.102650228Real-Time Encrypted Traffic Classification with Deep LearningOnur Demir0https://orcid.org/0000-0002-1088-6461Deniz Tuana Ergönül1https://orcid.org/0000-0003-2945-0833YEDİTEPE ÜNİVERSİTESİ, FEN BİLİMLERİ ENSTİTÜSÜYEDİTEPE ÜNİVERSİTESİ, FEN BİLİMLERİ ENSTİTÜSÜConfidentiality requirements of individuals and companies led to the dominance of encrypted payloads in the overall Internet traffic. Hence, traffic classification on a network became increasingly difficult as it must rely on only the packet headers. Many vital tasks such as differential pricing, providing a safe Internet for children, and eliminating malicious connections require traffic classification, even if the payload contents are encrypted. Encrypted traffic is harder to classify as packet content becomes unreadable. In this work, we aim to provide an insight into traffic classification using encrypted packets in terms of both accuracy and packet processing time. LSTM (Long Short-Term Memory) architecture is a good candidate for this problem as it can handle sequences. Each flow can be modeled as a sequence and patterns of the sequences can provide valuable information. We compare the performance of LSTM with other methods in both real-time and offline experiments. Compared to a machine learning method both online and offline LSTM excelled with precision and recall differences up to 50%. Average accuracy with LSTM was measured as 97.77% offline and 91.7% in real-time. Average packet processing time in real-time was recorded as 0.593 msec which is 5 times faster than a recent work that uses LSTM method.https://dergipark.org.tr/tr/download/article-file/2092192deep learningcomputer communication networksclassificationneural networks (computer)
spellingShingle Onur Demir
Deniz Tuana Ergönül
Real-Time Encrypted Traffic Classification with Deep Learning
Sakarya Üniversitesi Fen Bilimleri Enstitüsü Dergisi
deep learning
computer communication networks
classification
neural networks (computer)
title Real-Time Encrypted Traffic Classification with Deep Learning
title_full Real-Time Encrypted Traffic Classification with Deep Learning
title_fullStr Real-Time Encrypted Traffic Classification with Deep Learning
title_full_unstemmed Real-Time Encrypted Traffic Classification with Deep Learning
title_short Real-Time Encrypted Traffic Classification with Deep Learning
title_sort real time encrypted traffic classification with deep learning
topic deep learning
computer communication networks
classification
neural networks (computer)
url https://dergipark.org.tr/tr/download/article-file/2092192
work_keys_str_mv AT onurdemir realtimeencryptedtrafficclassificationwithdeeplearning
AT deniztuanaergonul realtimeencryptedtrafficclassificationwithdeeplearning