secure media streaming & secure adaptation for non-scalable video
Two important capabilities in media streaming are (1) adapting the media for the time-varying available network bandwidth and diverse client capabilities, and (2) protecting the security of the media. Providing both end-to-end security and adapting at a (potentially untrusted) sender or mid-network...
Gespeichert in:
1. Verfasser: | |
---|---|
Format: | Tagungsbericht |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Two important capabilities in media streaming are (1) adapting the media for the time-varying available network bandwidth and diverse client capabilities, and (2) protecting the security of the media. Providing both end-to-end security and adapting at a (potentially untrusted) sender or mid-network node or proxy can be solved via a framework called secure scalable streaming (SSS) which provides the ability to transcode the content without requiring decryption. In addition, this enables secure transcoding to be performed in a R-D optimized manner. The original SSS work was performed for scalably coded media. This paper examines its potential application to non-scalable media. Specifically, we examine the problems of how to scale non-scalable H.264/MPEG-4 AVC video and how to do it securely. We first show, perhaps surprisingly, (hat the importance of different P-frames in a sequence can vary by two orders of magnitude. Then we propose two approaches for securely streaming and adapting encrypted H.264 video streams in an R-D optimized manner using (1) secure-media R-D hint tracks, and (2) secure scalable packets. While we can not scale the bit rate of encrypted non-scalable H.264 to the same extent possible for scalably coded media, our method does provide some scaling capability and more importantly provides 4-8 dB gain compared to conventional approaches. |
---|---|
ISSN: | 1522-4880 2381-8549 |
DOI: | 10.1109/ICIP.2004.1421415 |