Why CDN-level video watermarking is not sufficient to track piracy and needs an A/B variant approach

The OTT-VoD content industry has seen a boom in consumption in the last few years which is fueled by an increase in internet penetration and smartphone usage. This has also necessitated the use of adaptive bit rate (ABR) streaming technologies. In this method, the content is first encoded at different bit rates and then segmented based on the bit rates. The bit rate segments thus created are then time aligned. A separate file known as the manifest file is used to provide information about the content resolutions, bit rates, and the rules for accessing each bit-rate segment.

Session-based and user-specific video watermarking enables the service operator to embed a unique identifier for each user. The ID is then used to trace any unauthorized redistribution of premium content. This technique is a critical anti-piracy tool to protect UHD premium content and live events broadcasting since it is possible to identify the actual device that is “leaking” or restreaming the content. This anti-piracy strategy allows for real-time detection of the source of pirated content and real-time shutdown of the illegal redistribution, and the ability to take other actions subject to the service provider’s policy.

Digital rights management (DRM) solutions can only protect content during transmission. However, once the DRM protected content reaches the consumer device, it is free to be redistributed by legitimate users and becomes susceptible to piracy attacks. Hence, forensic watermarking is needed for the identification of the content owner/broadcaster as well as the source of leakage at the user level in case of  content infringement. 

This is of even more importance in the current times when OTT video content is distributed using a large network of servers deployed globally across several data centers, also known as a content delivery network (CDN). While CDN servers ensure high network availability and performance by storing a cached copy of the content on each server and streaming them locally, all viewers receiving content from a particular server get the same watermarked content. This makes it difficult to track individual users and identify the exact point of leakage.A workaround to handle content security at the CDN edge is therefore necessary. Typically, manifest level video watermarking solutions are used for ABR streaming. In this method, two different uniquely watermarked versions (called A and B) of a content file are created. These streams are then broken up into segments which are ultimately combined to form a single stream with a unique combination of A and B segments. This ensures that no two users receive the same sequence. A/B watermarking can also be performed on the client side, where the unique watermarked version for each user is created by the client video player instead of the CDN edge server.

Share your love
Christophe Rude

Christophe Rude

Articles: 15885

Leave a Reply

Your email address will not be published. Required fields are marked *