Online Hybrid Classifier System of Internet Traffic
based on Machine Learning Approach and Port
Number
Journal:
GRENZE International Journal of Engineering and Technology
Authors:
Hamza Awad Hamza Ibrahim, Omer Radhi AL Zuobi, Awad M .Abaker, Marwan A.Al-Namari
Volume:
6
Issue:
2
Grenze ID:
01.GIJET.6.2.6_3
Pages:
61-68
Abstract
Internet traffic classification is valuable mechanism in the direction of traffic
detection and monitoring. Even though several classification approaches were proposed by
the research community, there still exist many open problems on Internet traffic
classification. The hybrid classifier is a classifier which combines more than one
classification method to identify the Internet traffic. Using only one method to classify
Internet traffic poses many risks. Therefore, this paper proposed a hybrid classifier (HC)
system to identify internet traffic. HC is based on two common classification methods, i.e.
port-base and ML-base. CH was able to perform an online classification since it able to
identify the live Internet traffic at the same time as when the traffic was generated. HC was
used to classify three common Internet applications classes i.e. web, WhatsApp, and
Twitter. HC is produces more than 90% classification accuracy which is higher when
compared with other individual classifiers.