| |
Abstract |
Video coding is a process for adapting media content to the constraints of transmission networks delivery and terminal device visualization. Moreover, content protection is also necessary. Nowadays the heterogeneity of client devices is increasing leading to different resolutions, qualities and form factors. Due to this, transcoding and protection are essential processes to be conducted in modern video distribution networks to adapt video to devices and network constraints and to enable pay per quality schemas enforcing content licenses. Unfortunately, transcoding and protection can be no longer considered linear since every single content should be transcoded in several formats and sometimes protected, so it would require a long time to finish. Modern scalable coding techniques, as H264 SVC, can help to save processing power and bandwidth providing in a single stream several video versions. However, if the enhancements of a SVC encoded content are protected separately, it would possible to enable pay-per-quality providing an additional degree of freedom to content delivery industry. Unfortunatelly, transcoding and protection entail huge doses of processing power at provider side and should be distributed. Moreover, processing key streams to decrypt enhancements that were encrypted separately can increase the complexity at receiver side. Cloud computing emerges as a potential solution for coping with large population of users with heterogeneous visualization devices. The elastic nature of cloud computing can be an advantage given the difficulty to predict the computing resources video content would require to be distributed during the entire content life. This article describes a system that distributes and parallelizes the video transcoding process as well as the content encryption, following the SaaS approach in cloud computing. Moreover, the article describes an experimental approach for generating and processing a flexible key stream that would help to simplify key management at receiver side and would allow legacy receivers to consume SVC content with separate enhancement protection. |
|
