DP Matching Approach for Streaming Contents Detection Using Traffic Pattern

Since the video streaming technology has become widespread, the security of video contents has gained more importance. Our research group previously envisioned a video leakage detection method for streaming contents that makes use of traffic patterns extracted from only traffic volume information obtained from routers. However, as the background traffic increases, packet delay and jitter increase. As a consequence, the detection accuracy decreases. In this paper, we propose a new robustmethod, which is more resilient to increasing packet delay and jitter by improving the generation process of traffic patterns. In addition, we solve another problem induced by packet loss by using Dynamic Programming (DP) matching. Finally, we evaluate the influence of the background traffic on streaming video detection and discuss the results.