1 / 20

Caching Strategies in Transcoding-Enabled Proxy System for Streaming Media Distribution Networks

Caching Strategies in Transcoding-Enabled Proxy System for Streaming Media Distribution Networks. Bo Shen Sung-Ju Lee Sujoy Basu IEEE Transactions On Multimedia, April 2004. Outline. Introduction Transcoding-Enabled Caching System Architecture Transcoding-Enabled Caching Algorithms

fauna
Download Presentation

Caching Strategies in Transcoding-Enabled Proxy System for Streaming Media Distribution Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Caching Strategies in Transcoding-Enabled Proxy System for Streaming Media Distribution Networks Bo Shen Sung-Ju Lee Sujoy Basu IEEE Transactions On Multimedia, April 2004

  2. Outline • Introduction • Transcoding-Enabled Caching System Architecture • Transcoding-Enabled Caching Algorithms • Other Proxy Caching Algorithms • Performance Analysis • Conclusion

  3. IntroductionTraditional Proxy Caching Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2 (1) 64 kbps (2) 128 kbps ………… ………… Video 1 256 kbps Server Desktop Client Video 1 64 kbps Video 1 (1) 64 kbps(2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Proxy Video 2 128 kbps Cell Phone Video 1 128 kbps PDA PDA

  4. IntroductionTEC Purpose • Transcoding-enabled Catching (TEC) proxy system is designed for efficient delivery of rich media web content to heterogeneous network and client capabilities. • Save cache size of proxy.

  5. IntroductionOther method for heterogeneous client • Scalable coding and layered coding • Majority if not all of the current multimedia content on the Internet is coded in nonscalable, single-layered format.

  6. TEC System Architecture

  7. TEC System ArchitectureTranscoding Techniques • A higher bit-rate version can be transcoded to a lower bit-rate version. • The transcoded version may have fidelity degradation compared with the original version. • Compressed domain transcoding • Bit-rate reduction transcoding • Spatial resolution reduction transcoding

  8. TEC System ArchitectureTranscoding Techniques Bit-rate reduction transcoding • Faster than a decode-and-re-encode approach

  9. TEC System ArchitectureTranscoding Techniques • Spatial resolution reduction transcoding • Transcoder obtains motion information for the down-sampled video directly from the orginal video. • TEC System uses bit-rate reduction model

  10. TEC Algorithm • Exact hit : the requested version of the video object exists in the cache. • Transcode hit : the requested version does not exist in the cache, but a transcodable version of the video exists. • Miss : the requested or transcodable version of the video does not exist in the cache.

  11. TEC AlgorithmTEC 11 If cache size is not enough, use LRU replacement V1 256 kbps is trancoded to V1 128 kbps Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2 (1) 64 kbps (2) 128 kbps ………… ………… Miss Exact hit V1 256 kbps Server V1 128 kbps V1 256 kbps Video 1(1) 64 kbps (2) 128 kbps(3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Video 1(1) 64 kbps(2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Request V1 128 kbps Request V1 256 kbps Proxy PDA Cache Single Version of Video

  12. TEC AlgorithmTEC 12 V1 256 kbps is trancoded to V1 128 kbps Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2 (1) 64 kbps (2) 128 kbps ………… ………… Server V1 128 kbps Video 1(1) 64 kbps(2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Video 1(1) 64 kbps (2) 128 kbps(3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Request V1 128 kbps Proxy PDA Cache Single Version of Video

  13. TEC AlgorithmTEC 2 (1) V1 128 kbps is transcoded to 64 kbps (2)cache 64 kbps, if cache size is not enough, use LRU Cache V1 256 kbps, if cache is not enough, use LRU replacement Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2 (1) 64 kbps (2) 128 kbps ………… ………… Server Video 1(1) 64 kbs (2) 128 kbps(3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Video 1 (1) 64 kbps (2) 128 kbps(3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Request V1 64 kbps Request V1 256 kbps Proxy PDA Cache Multiple Version of Video

  14. Other Proxy Cache Algorithm FVO (Full Version Only) Transcoding from V1 256 kbps Video 1 (1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2 (1) 64 kbps (2) 128 kbps ………… ………… Server Video 1(1) 64 kbps (2) 128 kbps (3) 256 kbpsVideo 2(1) 64 kbps (2) 128 kbps ………… ………… Request V1 64 or 128 kbps Proxy PDA

  15. Other Proxy Cache Algorithm • TVO (Transcoded Version Only) : Always caching the transcoded objects, and if a request does not result in an exact hit, the full version is fetched from the origin to produce a transcoded version. • LRU : Caching all video across proxy without any transcoding, if cache size is not enough, using LRU replacement.

  16. Performance AnalysisHighly Heterogeneous Environment • Video version : Origin server has 512 kbps, 256 kbps, 128 kbps, 64 kbps • The full version is not requested by the clients, but could be accessed by the proxy when FVO or TVO are used. • Byte hit ratio : The number of bytes served from the cache to the clients over the number of bytes requested from the client • Relative cache size : Dividing the cache size by the size summation of total unique objects.

  17. Performance AnalysisHighly Heterogeneous Environment Indicates the transcoding load each algorithm requires

  18. Performance AnalysisHighly Heterogeneous Environment Quality degradation less than 1 dB

  19. Performance Analysis

  20. Conclusion • TEC improves the cache effectiveness, decreases the user-perceived latency, and reduces the traffic between the proxy and the content origin server. • TEC-2 gave good performances when the caches size is large. • TEC-11 generally showed the highest byte-hit ratio but also had the highest transcoding load.

More Related