Detection of network security attacks in real time

Abstract

Computer networks have become the target of several attacks. The goal of those attacks is to disrupt the normal behaviour of the services provided from inside the network (DoS attacks) or to gain access to the systems offering those services. For that reason, the detection of those attacks as soon as possible is key to maintain the security of the whole network. This is precisely the objective of this project: the development of a detection module together with a fast packet sniffer which will allow detecting such attacks in real time. The tools that will be used to fulfill the goal are machine learning approaches and the labeled dataset UNSW-NB15 which contains a lot of different types of attacks.