Traffic Pattern Plot: Video Identification in Encrypted Network Traffic

Most of the Internet traffic is encrypted, and the challenge is its ability to recognize the streaming videos from the Internet traffic. In this paper, we present a methodology named traffic pattern plot (TPP) to identify video streams in encrypted network traffic. The proposed methodology plots the video traffic flows and uses a convolutional neural network (CNN) to detect the videos. The results show that the traffic pattern plot generated from 120 s of sniffing network traffic is enough to identify the video even in the encrypted network traffic with 94% accuracy.