You are on page 1of 1

Data classification in sdn useing machine learning

Pedro Amaral
Jo~ao Dinis
Paulo Pinto Include University or Department Names if Needed
Luis Bernardo
Jo~ao Tavares and Henrique S. Mamede

Introduction Results
Software Defined Networks (SDNs) to separate control plane and forward level networks. Implementation of control plane software The trac generated by the host is then sent via port mirroring to the OpenFlow
and data collection mechanisms built into OpenFlow protocol promise to be excellent tools for implementing Machine Learning switch where the SDN application in the controller performs the data collection.
Network (ML) control applications. The first step in this direction is to understand what kind of data can be collected in SDNs and This results in a data-set where the application responsible for the features mea-
how information can be learned from that data in this paper. Explain how many ML traffic classification techniques can be applied. sured in each TCP connection is known.

Methods Conclusions
a traffic classification architecture based in SDN that allows the collection of All authors want to solve The massive growth of the Internet users over the past
trac data both in SDN networks and in legacy networks. It is based solely on years has been accompanied with a rapid evolution of networking systems such
OpenFlow and can be implemented in an SDN app in the controller, paving the as, Software-Defined Networking (SDN) is a network architecture approach that
enables the network to be intelligently and centrally controlled, or programmed
way for the use of data driven learning algorithms for network control.
using software applications. This helps operators manage the entire network
The collected data has very deferent features with values in deferent scales. consistently and holistically.
In order to avoid features with large scale values to overlap smaller scale ones
causing bad performance in the ML algorithms, the collected data is treated to
scale and center the feature values. This is done by calculating the standard
scores for each data feature.

References
Pedro Amaral, João Dinis, Paulo Pinto, Luis Bernardo, João Tavares and Henrique S. Mamede, "Machine Learning in Software Defined
Networks: Data collection and traffic classification", 2016 IEEE 24th International Conference on Network Protocols (ICNP), 2018.

You might also like