MPEG-DASH & Common Encryption: The Promise of Broad Reach and Interoperability
Adaptive streaming technology is converging towards a client-based adaption logic, which requires a manifest description provided by the server and content retrieval via HTTP. An open standard is needed to extend the content silos created by HTTP Live Streaming, Smooth Streaming and HTTP Dynamic Streaming that tend to balkanize the world of Internet content.
The MPEG-DASH standard is filling this need by embracing a superset of these proprietary streaming formats and accumulating many of the proven features into an international standard. Best practices on how to use this standard are provided by the DASH Industry Forum (DASH-IF). Both are essential to foster adoption by playback clients but also by other standards that can build on this open and stable framework for high quality services using adaptive bitrate streaming.
In fact, just last week the DASH IF held a press conference to share implementation guidelines known as DASH-AVC/264, further continuing its move toward true service interoperability. (MPEG-DASH: DASH264 Implementation Guidelines Released Today)
One component in DASH that is an important contribution to commercial adaptive streaming services is the inclusion of Common Encryption (CENC - standardized as ISO ISO/IEC23001-7). It builds a simple, yet effective, system for different digital rights management (DRM) systems to share keys, key identifiers, encryption algorithm, parameters and signaling as well as a location to store proprietary data in a Protection System Specific Header (PSSH), but leaves DRM implementation to individual systems. Similar to DVB Simulcrypt—which stores this information for each DRM in ECMs—this enables each DRM to apply their own robustness rules, security implementation and key exchange system to the encrypted content of the same content file.
The underlying implementation is left to each DRM, which requires individual support for different end devices, meaning the playback of encrypted premium content is still dependent on specific DRM support. On the other hand, it enables DRM renewability, replacement and addition to a growing ecosystem of supported devices. In contrast, a standard that enables every client to decrypt the content for physical media distribution on many unconnected devices from multiple vendors, needs to specify rules on how the key is ultimately obtained.
The ISO Common Encryption standard specifies the encryption algorithm and key length only relies on the security of the underlying cypher and it is a reasonable assumption that this will not be easily compromised. As such, the CENC approach meets the needs to play back a common media resource in devices from multiple vendors, even those that use different systems to determine how the decryption key or keys are ultimately obtained. It’s even feasible to understand how the encryption approach could evolve over time without creating device incompatibilities. This enables a long lasting security standard that can be used by individual DRM providers to implement and improve robustness rules and more specifically match the evolving requirements from content owners for premium content distribution.
The concept of enabling the same content to be used by multiple DRM implementations at the time of content creation also allows redistribution of the content to any device that supports any of those DRMs. While DASH-AVC/264 is designed for adaptive streaming, it is also well suited to be downloaded and re-distributed in the home to additional devices using different DRMs—a model that is not supported by other adaptive streaming approaches in this fashion. In addition, the format also enables a new DRM to be attached to the content even several years later without the need to re-encrypt it, making it extremely flexible.
I recently shared more details on this topic at the Streaming Forum 2013 in London during my presentation on “Leveraging MPEG-DASH for Enhanced Revenue Security.” They will be posting my slides and videotaped session shortly so stay tuned.