Open AccessUse of SPAN for Identified Network Traffic via Machine Learning
Author(s) -
Akifa Abbas,
Danish Ali,
Sidra Noureen,
Rehmatullah
Publication year - 2019
Publication title -
lahore garrison university research journal of computer science and information technology
Language(s) - English
Resource type - Journals
eISSN - 2521-0122
pISSN - 2519-7991
DOI - 10.54692/lgurjcsit.2019.030276
Subject(s) - computer science , identification (biology) , computer network , encryption , wireless network , computer security , mac address , web traffic , traffic generation model , private network , wireless , the internet , telecommunications , world wide web , botany , biology
Few years back the number of wireless devices and their use in our daily life has been increased a lot. All devices cell phones, laptops, tablets, camera, TVs, home appliances have become a part of network now. As the network devices are growing and getting connected to each other the security risks are getting higher. All the companies and organizations are now establishing and implanting the public and private wireless networks. Organization have to pay heavy cost to implement and integrate all devices together on a network. As wireless networks are more vulnerable to threats and in security's a huge network all the devices should be identified whenever they enter or leave a network traffic pool the experimental work in this paper will elaborate the methods to identify the network traffic identification under encryption. This paper emphases on identification of devices based on layer 2 functionality by MAC (Media Access Code). Later on, the identification was improved using labeled or tagged traffic methods by use of SPAN (Switch port analyzer technique) technology or protocol with assistance ofVirtual Local Area Network. Many Supervised learning methods were examined during experiment and were referenced on data collected by real time traffic. The network traffic of multiple deceives gradually passes through network so incremental learning method is implemented as classification for streaming data.