Our MLASH can use either the offline classification model or streaming classification model. The streaming model can not only use accumulated transaction logs to improve prediction accuracy, but also adapt to network dynamics. We hence check how many train-ing data are required to obtain a stable and reliable classification model. Figure 4.7 plots
0 2 4 6 8 10 x 105 0
500 1000 1500
Number of training data
Prediction error (kb/s)
bandwidth−based labeling buffer−considered labeling smoothness−based labeling
Figure 4.7: Prediction error of the streaming model over time
the prediction error of the streaming classification model for different labeling algorithms as the number of training data grows over time. The figure shows that the accuracy be-comes quite stable when the classifier is trained by using about 100,000 records of training data.
Chapter 5 Conclusion
This paper presents a machine-learning-based adaptive streaming framework over HTTP (MLASH). Instead of designing a completely new adaptation algorithm, our goal is to combine the machine learning technique with different existing rate adaptation algorithms designed for optimizing different QoE metrics. We train a classification model to de-scribe the explicit relationships between a wide range of network-related features and the label found by any preferable rate adaptation algorithm based on true information. This machine-learning-based approach can hence elastically utilize comprehensive features, and, more importantly, avoids the difficulty of bandwidth estimation faced by many exist-ing adaptation algorithms. We demonstrate via trace-based simulations that, by leveragexist-ing existing adaptation algorithms as the labeling scheme, our MLASH can improve predic-tion accuracy of those algorithms, and hence their target performance metrics.
Bibliography
[1] Athula Balachandran, Vyas Sekar, Aditya Akella, Srinivasan Seshan, Ion Stoica, and Hui Zhang. Developing a predictive model of quality of experience for internet video. In Proceedings of the 2013 ACM Conference on SIGCOMM.
[2] Christopher Müller, Stefan Lederer, and Christian Timmerer. An evaluation of dy-namic adaptive streaming over http in vehicular environments. In Proceedings of the 4th Workshop on Mobile Video, 2012.
[3] Guibin Tian and Yong Liu. Towards agile and smooth video adaptation in dynamic http streaming. In Proceedings of the 8th International Conference on Emerging Networking Experiments and Technologies, 2012.
[4] Ricky K. P. Mok, Xiapu Luo, Edmond W. W. Chan, and Rocky K. C. Chang. Qdash:
A qoe-aware dash system. In Proceedings of the 3rd Multimedia Systems Conference, 2012.
[5] V. Joseph and G. de Veciana. Nova: Qoe-driven optimization of dash-based video delivery in networks. In INFOCOM, 2014 Proceedings IEEE.
[6] Thomas Wirth, Yago Sánchez, Bernd Holfeld, and Thomas Schierl. Advanced down-link lte radio resource management for http-streaming. In Proceedings of the 20th ACM International Conference on Multimedia, 2012.
[7] Te-Yuan Huang, Ramesh Johari, Nick McKeown, Matthew Trunnell, and Mark Wat-son. A buffer-based approach to rate adaptation: Evidence from a large video stream-ing service. In Proceedstream-ings of the 2014 ACM Conference on SIGCOMM.
[8] Te-Yuan Huang, Nikhil Handigol, Brandon Heller, Nick McKeown, and Ramesh Johari. Confused, timid, and unstable: Picking a video streaming rate is hard. In Proceedings of the 2012 ACM Conference on Internet Measurement Conference, 2012.
[9] Te-Yuan Huang, Ramesh Johari, and Nick McKeown. Downton abbey without the hiccups: Buffer-based rate adaptation for http video streaming. In Proceedings of the 2013 ACM SIGCOMM Workshop on Future Human-centric Multimedia Networking, 2013.
[10] Ahmed Hamza and Mohamed Hefeeda. A dash-based free viewpoint video stream-ing system. In Proceedstream-ings of Network and Operatstream-ing System Support on Digital Audio and Video Workshop, 2014.
[11] Zied Aouini, Mamadou Tourad Diallo, Ali Gouta, Anne-Marie Kermarrec, and Yan-nick Lelouedec. Improving caching efficiency and quality of experience with cf-dash. In Proceedings of Network and Operating System Support on Digital Audio and Video Workshop, 2014.
[12] Danny H. Lee, Constantine Dovrolis, and Ali C. Begen. Caching in http adaptive streaming: Friend or foe? In Proceedings of Network and Operating System Support on Digital Audio and Video Workshop, 2014.
[13] Ali C. Begen Saamer Akhshabi, Lakshmi Anantakrishnan and Constantine Dovrolis.
What happens when http adaptive streaming players compete for bandwidth? In Proceedings of Network and Operating System Support on Digital Audio and Video Workshop, 2012.
[14] Sajid Nazir, Ziaul Hossain, Raffaello Secchi, Matthew Broadbent, Andreas Petlund, and Gorry Fairhurst. Performance evaluation of congestion window validation for dash transport. In Proceedings of Network and Operating System Support on Digital Audio and Video Workshop, 2014.
[15] Feng Wang, Jiangchuan Liu, and Minghua Chen. Calms: Cloud-assisted live media streaming for globalized demands with time/region diversities. In INFOCOM, 2012 Proceedings IEEE, 2012.
[16] Cong Wang and Michael Zink. On the feasibility of dash streaming in the cloud. In Proceedings of Network and Operating System Support on Digital Audio and Video Workshop, 2014.
[17] D. De Vleeschauwer, H. Viswanathan, A. Beck, S. Benno, Gang Li, and R. Miller.
Optimization of http adaptive streaming over mobile cellular networks. In INFO-COM, 2013 Proceedings IEEE.
[18] R. Bhatia, T.V. Lakshman, A. Netravali, and K. Sabnani. Improving mobile video streaming with link aware scheduling and client caches. In INFOCOM, 2014 Pro-ceedings IEEE.
[19] S. Cicalo, N. Changuel, R. Miller, B. Sayadi, and V. Tralli. Quality-fair http adaptive streaming over lte network. In Acoustics, Speech and Signal Processing (ICASSP), 2014 IEEE International Conference on.
[20] Sally Floyd, Mark Handley, Jitendra Padhye, and Jörg Widmer. Equation-based congestion control for unicast applications. In Proceedings of the Conference on Applications, Technologies, Architectures, and Protocols for Computer Communi-cation, 2000.
[21] Matthew Mathis, Jeffrey Semke, Jamshid Mahdavi, and Teunis Ott. The macroscopic behavior of the tcp congestion avoidance algorithm. SIGCOMM Comput. Commun.
Rev., 27(3):67–82, 1997.
[22] Jitendra Padhye, Victor Firoiu, Don Towsley, and Jim Kurose. Modeling tcp through-put: A simple model and its empirical validation. In Proceedings of the ACM SIG-COMM ’98 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication, 1998.
[23] Martin Swany and Rich Wolski. Multivariate resource performance forecasting in the network weather service. In Proceedings of the 2002 ACM/IEEE Conference on Supercomputing, 2002.
[24] S. Vazhkudai, J.M. Schopf, and I. Foster. Predicting the performance of wide area data transfers. In Parallel and Distributed Processing Symposium., Proceedings International, IPDPS 2002, Abstracts and CD-ROM, 2002.
[25] Yin Zhang and Nick Duffield. On the constancy of internet path properties. In Proceedings of the 1st ACM SIGCOMM Workshop on Internet Measurement, 2001.
[26] Qi He, Constantine Dovrolis, and Mostafa Ammar. On the predictability of large transfer tcp throughput. In Proceedings of the 2005 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications.
[27] M. Mirza, J. Sommers, P. Barford, and Xiaojin Zhu. A machine learning approach to tcp throughput prediction. Networking, IEEE/ACM Transactions on, 18(4):1026–
1039, 2010.
[28] Vengatanathan Krishnamoorthi, Niklas Carlsson, Derek Eager, Anirban Mahanti, and Nahid Shahmehri. Quality-adaptive prefetching for interactive branched video using http-based adaptive streaming. In Proceedings of the ACM International Con-ference on Multimedia, 2014.
[29] Mohammad Ashraful Hoque, Matti Siekkinen, and Jukka K. Nurminen. Using crowd-sourced viewing statistics to save energy in wireless video streaming. In Pro-ceedings of the 19th Annual International Conference on Mobile Computing &
Networking, 2013.
[30] Hyunwoo Nam, Kyung Hwa Kim, Bong Ho Kim, D. Calin, and H. Schulzrinne.
Towards dynamic qos-aware over-the-top video streaming. In A World of Wireless, Mobile and Multimedia Networks (WoWMoM), 2014 IEEE 15th International Sym-posium on.
[31] Leo Breiman. Random forests. Machine learning, 45(1):5–32, 2001.
[32] Simone Basso, Antonio Servetti, Enrico Masala, and Juan Carlos De Martin. Mea-suring dash streaming performance from the end users perspective using neubot. In Proceedings of the 5th ACM Multimedia Systems Conference, 2014.