Wavelet-Based Traffic Analysis for Identifying Video Streams over Broadband Networks

Network and service providers are rapidly deploying IPTV networks to deliver a wide variety of video content to subscribers. Some video content may be protected by copyright and/or may be subject to distribution restrictions. Encryption technologies may not always be effective to manage protected video content, particularly when video content is legally decrypted upon receipt by a subscriber. This paper presents a new approach to detect if specific (or protected) downloaded video is being redistributed by a subscriber using the broadband Internet connection. The approach employs a traffic-based signature of the protected video clip. The signature which is shown to be unique is stored in a signature store. We adopt a wavelet-based analysis to match video streams captured from the network to the signatures in the store. The performance of the detection algorithm is evaluated using a large video database populated with a variety of movies and TV shows. The experiment results show that our algorithm achieves high detection rates and low false alarm rates using video clips of only a few seconds.