A Comparative Evaluation of Deep Learning Approaches to Online Network Traffic Classification for Community Networks

Community networks have emerged as a promising solution to providing internet connectivity in low-resource rural areas around the world. In order to improve the network experience for the members of community networks, network traffic packet classification can be used to assign different priorities to different applications (e.g. Zoom) through quality of service (QoS) algorithms. In this project we thus investigate the suitability of three different deep learning architectures for real-time network traffic packet classification in the context of the resource constraints of community networks.