1 / 17

A Comparison of Quality Scheduling in Commercial Adaptive HTTP Streaming Solutions on a 3G Network

A Comparison of Quality Scheduling in Commercial Adaptive HTTP Streaming Solutions on a 3G Network. H. Riiser , H. Bergsaker , P. Vigmostad , P. Halvorsen , and C. Griwodz. Proceedings of the 4 th ACM Workshop on Mobile Video ( MoVid ) Chapel Hill, NC, USA February2012. Motivation.

jemma
Télécharger la présentation

A Comparison of Quality Scheduling in Commercial Adaptive HTTP Streaming Solutions on a 3G Network

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. A Comparison of Quality Scheduling in Commercial Adaptive HTTP Streaming Solutions on a 3G Network H. Riiser, H. Bergsaker, P. Vigmostad, P. Halvorsen, and C. Griwodz Proceedings of the 4th ACM Workshop on Mobile Video (MoVid) Chapel Hill, NC, USA February2012

  2. Motivation • HTTP streaming increasing • Good performance under many situations • But HD mobile, where available bandwidth fluctuates a challenge • Overall aim is “best” viewing experience • Avoid buffer underruns (they cause interruptions) • Avoid rapid oscillations in quality (users do not like) • Utilize as much of the available bandwidth as possible • Anecdotally, noticed very different performance from different clients

  3. Fluctuating Bandwidth Problem

  4. Fluctuating Bandwidth Problem

  5. Fluctuating Bandwidth Problem Ferry Metro Question: How well do adaptive streaming media solutions adapt? Bus Tram

  6. Experiment Overview • Gather representative traces from above environment (ferry, subway, tram, bus) • Bitrate per second log, average delay • Apache module replay these exactly in closed testbed • That way, can change just client and get exactly the same network conditions • Streaming over HTTP is pull-based (rather than push-based), in that the client requests data • DASH – Dynamic Adaptive Streaming over HTTP • Not standardized at the time of this paper • So, choose representative players • Netflix and others are closed – cannot setup content to stream in testbed • Plus others have studied [6] • Plus, plus generally expected for fixed endpoints

  7. Adaptive Delivery: Tested Systems • Adobe Strobe Media Playback (v1.6.328 for Flash 10.1)using HTTP Dynamic Streaming Format • Apple’s native iPad player (iOS v4.3.3)using native HLS format • Microsoft Silverlight/IIS Smooth (v4.0.60531.0 on Win7)using native Smooth format and default desktop scheduler • Netview Media Client (v2011-10-10)using Apple HLS format (worst case) and Netview 3G scheduler Server • Apache v2.2.14 • CodeShop Unified Streaming Platform v1.4.25, handles all encodings • Throttling module for trace-based bandwidth limitation, 1sec resolution • Dedicated Linux box, 2GB RAM, AMD 3299_ CPU HTTP streaming HTTP live streaming

  8. Adaptive Delivery: Test Content • Video: Norwegian football game (i.e., soccer) • H.264/AVC • 2 second segments • 6 quality levels: 250, 500, 750, 1000, 1500, 3000 kbps Question: What is “good” adaptation? Akamai recommendations

  9. Traces with Observed Bandwidth

  10. Comparison of Existing Quality Schedulers Bus: Configured for desktops Configured for mobiles Ferry:

  11. Comparison of Existing Quality Schedulers Metro: Tram:

  12. Overall Quality

  13. Conclusion • Large differences in performance, even though all have same information • Apple and Adobe opposites • Apple sacrifices high average quality for stability • Adobe chooses highest average quality, not stable • Microsoft Silverlight in between • Netview similar to Silverlight, but better against buffer underruns

  14. Prediction? • Many people commute using the same route • Many mobile devices have GPS receivers • What about crowdsourcing the throughput on the commute paths at various times of day? From: “Video Streaming Using a Location-based Bandwidth-Lookup Service for Bitrate Planning”, Haakon Riiser, Tore Endestad, Paul Vigmostad, Carsten Griwodz, PålHalvorsen, TOMCCAP

  15. Prediction Metro: Location-based bandwidth-lookup service for bitrate (video quality) planning: Algorithm: • Calculate the predicted amount of data that historically can be downloaded • Calculate maximum steady quality without getting buffer-underruns • Safety: reactive algorithm fallback

  16. Prediction: Metro Metro:

  17. Future Work • Lots of room for adjusting, deciding “best”

More Related