The main task of the testbed designed is to capture network traffic and perform its classification in real-time (or close to it), by utilizing pre-trained deep neural network models. The testbed, in its essence, is a computer with specifically configured virtual machine acting as a traffic monitor for network traffic analysis and recognition, for example, by detecting the video quality for streaming video. The structural diagram of the testbed is displayed in figure below:
Software Defined Network simulation; Real-time network traffic classification.
High classification accuracy, over 95% in tests.
Machine Learning,Deep Neural Networks,Computer networks traffic,Python