Abstract:Online traffic classification has been widely used in quality of service measurements, network management and security monitoring. Currently, more and more research works tend to apply machine learning techniques to online traffic classification, and most of them are based on supervised learning and unsupervised learning techniques. Although supervised learning method has exhibited good classification performance, it needs lots of labeled training samples which are difficult to collect. The co-training method is a semi-supervised learning method, which can use little labeled samples and plenty of unlabeled samples to enhance the performance of supervised learning method. In this paper, we investigate the co-training algorithm for online traffic classification. The co-training algorithm needs two separate features which are sufficient to train a good classifier. We choose packet size and inter-packet time of the first packets of a traffic flow as two features. However, the inter-packet time is dependent to network conditions and will be impacted by network jitter. This paper constructs a robust inter-packet time feature named "Netipt" which is resilient to network jitter, and we integrate Netipt feature to co-training algorithm. We test our co-training algorithm based on two real-world traffic datasets. The results show that the co-training algorithm can enhance the accuracy of traffic classification drastically even when there are very few training samples.

A Heuristic-Based Co-clustering Algorithm for the Internet Traffic Classification

High accurate internet traffic classification based on co-training semisupervised clustering

Online Internet Traffic Identification Algorithm Based on Multistage Classifier

A Statistical-Feature-based Approach to Internet Traffic Classification Using Machine Learning.

Towards a Profiling View for Unsupervised Traffic Classification by Exploring the Statistic Features and Link Patterns

Internet Traffic Clustering with Constraints

Traffic Classification - Towards Accurate Real Time Network Applications

Internet Traffic Classification Using Machine Learning: A Token-based Approach

An Effective Network Traffic Classification Method with Unknown Flow Detection

Service Classification of High-Speed Network Traffic Based on Two-Stage Clustering

Network Traffic Identification Based on Online Clustering

Study on Internet Traffic Classification Using Machine Learning

Online Traffic Classification Based on Co-training Method

Research of IP Flow Classification Based on Heuristic Search

An Internet Traffic Classification Method Based on Semi-Supervised Support Vector Machine

Traffic classification using an improved clustering algorithm

Internet Traffic Identification Using Community Detecting Algorithm

Network Traffic Classification Algorithm Based on Spectral Clustering

A Self-Adaptive Network Traffic Classification System with Unknown Flow Detection

Information Entropy Based Clustering Method for Unsupervised Internet Traffic Classification

Hybrid Internet Traffic Classification Technique