摘 要
越來越多現代人擁有行動裝置，並且經由網際網路觀看視訊(Video)服務的需求快速成長，思科行動數據視覺網路指標(Cisco VNI Mobile) 預估在2018年，影音資料量將會占全部行動網路資料量的69.1%。
為了有效利用網路頻寬並且讓使用者可以觀看一個穩定和具有高品質的多媒體影音資訊，動態影像專業組織(MPEG)和3GPP制定了在HTTP具有動態與適應的串流(Dynamic Adaptive Streaming over HTTP, DASH)，DASH可以使得內容傳遞網路(Content Distribution Networks ,CDNs)能夠更有效路的傳遞影音串流。DASH將不同解析度的影像進行編碼，並且儲存CDNs 的伺服器內一次，使用者只須根據當前預估的可用頻寬，對距離自己最近的CDNs伺服器要求相對應的影音串流，如此一來可以減少對遠距離傳輸的需要。
不同於過往的必須先在載到一定程度才能撥放影片，DASH可以根據當前所預估的可用頻寬來動態的調整影像的解析度，如此一來也可以避免使用者下載完卻決定不觀看所造成的網路頻寬浪費。
在設計DASH 影片速率(video rate)的動態調整，主要有三個評估依據分別為1)網路頻寬的使用效率(Utilization) 2)影片品質的穩定性(Instability) 3) 頻寬分配的公平性(Unfairness)。在傳統的方法，使用者在下載時，會將所下載的資料先存入緩衝器(buffer)，當緩衝器溢出(overflow)所進來的資料就會直接丟棄，此時使用者就會先停止要求下載，就會出現off-period，而off-period卻會造成其他使用者高估了可用頻寬。
我們的貢獻是避免緩衝器的溢出，並且利用機率來動態的調整所下載的影片速率，如此一來使用者可以預估到具有公平性的可用頻寬，如此一來可以有效的下載可撥放的影片，並且實作在Network Simulator 2上。
關鍵字: 公平性、HTTP串流調適

Abstract
Dynamic Adaptation Streaming over HTTP (DASH) is deployed for getting appropriate video adaptively to the available bandwidth and improving the bandwidth utilization. HTTP server divide the different versions of the same video into smaller unit, segment, which is the specific piece of video. Clients can switch video version dynamically by downloading the video segments under the fluctuant network bandwidth.
We use the client’s buffer length to avoid the buffer overflow (or underflow) which is the key factor of the bandwidth oscillation, so clients can estimate the fair-share bandwidth and watch the video with stable video rate.
Because the buffer length can smooth video rate, we don’t have to switch video rate with the variation of TCP throughput. We determine to switch video rate by probability until the buffer is larger than the buffer reference instead of switching the video rate directly by the estimated bandwidth. We use the random probability as the threshold to decide whether to switch the video rate or not, it is fairly to client to get a better viewing experience.

Keywords: Fairness, Dynamic Adaptive Streaming over HTTP

摘 要…………………………………………………………………………………………………………………………………… i
Abstract………………………………………………………………………………………………………………………… ii
Content ……………………………………………………………………………………………………………………… iii
Chapter 1 Introduction …………………………………………………………………………………… 1
Motivation ……………………………………………………………………………………………… 2
Chapter 2 Background ………………………………………………………………………………………… 4
2.1 DASH …………………………………………………………………………………………………… 4
2.1.1 Adaptive HTTP streaming ……………………………………… 5
2.1.2 The component of DASH …………………………………………… 5
2.2 The buffer length ………………………………………………………………… 7
2.3 ON-OFF period …………………………………………………………………………… 8
Chapter 3 Related work ………………………………………………………………………………… 11
Chapter 4 Proposed method ………………………………………………………………………… 13
4.1 Competing criteria …………………………………………………………… 15
4.2 Average buffered video time …………………………………… 16
4.2.1 The boundary for ω_q …………………………………………… 16
4.3 The bandwidth estimation …………………………………………… 22
4.4 The probability of updating video rate ……… 22
4.5 The size of the average buffer length ………… 23
4.5.1 The avg is smaller than the minimum threshold …………………………………………………………………………………………………………………… 23
4.5.2 The avg is between the minimum threshold and the buffer reference……………………………………………………………………………… 24
4.5.3 The avg is larger than the buffer reference …………………………………………………………………………………………………………………… 24
Chapter 5 Experiment ……………………………………………………………………………………… 26
5.1 Two clients ……………………………………………………………………………… 27
5.2 Multiple clients ………………………………………………………………… 34
Chapter 6 Conclusion ……………………………………………………………………………………… 37
References ………………………………………………………………………………………………………………… 38

References
[1] F. Dobrian, V. Sekar, A. Awan, I. Stoica, D. Joseph, A. Ganjam, J. Zhan, and H. Zhang, “Understanding the impact of video quality on user engagement,” In Proc. of ACM SIGCOMM, pp. 362–373, 2011.
[2] P. Ni, R. Eg, A. Eichhorn, C. Griwodz, and P. Halvorsen, “Flicker effects in adaptive video streaming to handheld devices,” in Proc. of ACM Int. Conf. Multimedia, pp. 463–472, 2011.
[3] G. Tian, and Y. Liu, “Towards agile and smooth video adaptation in dynamic HTTP streaming,” In Proc. of Conf. Emerging networking experiments and technologies (CoNEXT), pp. 109-120, 2012.
[4] C. Zhou, C. W. Lin, X. Zhang, and Z. Guo, “A control-theoretic approach to rate adaption for DASH over multiple content distribution servers,” IEEE Trans. Circuits Syst. Video Technol., pp. 681-694, 2014.
[5] J. Jiang, V. Sekar, and H. Zhang, “Improving fairness, efficiency, and stability in http-based adaptive video streaming with festive,” In Proc. of Conf. Emerging networking experiments and technologies (CoNEXT), pp. 97-108, 2012.
[6] R.K.P. Mok, E.W.W. Chan, X. Luo, and R.K.C. Chang, “Inferring the QoE of HTTP video streaming from user-viewing activities,” In Proc. of ACM SIGCOMM workshop on Measurements up the stack (W-MUST), pp.31-36, 2011.
[7] S. Floyd, and V. Jacobson, “Random Early Detection Gateways for Congestion Avoidance,” IEEE Trans. Networking, pp. 397-413, 1993.
[8] Z. Li, X. Zhu, J. Gahm, R. Pan, H. Hu, Ali C. Begen, and D. Oran, “Probe and adapt: Rate adaptation for HTTP video streaming at scale,” IEEE J. Sel. Areas Commun., pp. 719-733, 2014.
[9] Harmonic mean. http://en.wikipedia.org/wiki/Harmonic_mean
[10] S. Akhshabi, S. Narayanaswamy, Ali C. Begen, C. Dovrolis, “An experimental evaluation of rate-adaptive video players over HTTP,” Signal Process.: Image Commun., pp. 271–287, 2012.
[11] N. Cranley, P. Perry, and L. Murphy, “User perception of adapting video quality,” International Journal of Human-Computer Studies, pp. 637-647, 2006.
[12] R. Jain, D. Chiu, and W. Hawe, “A quantitative measure of fairness and discrimination for resource allocation in shared computer system,” Digital Equipment Corporation, 1984.
[13] Cisco White Paper: Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2009-2014, http://bit.ly/bwGY7L
[14] T. Stockhammer, “Dynamic Adaptive Streaming over HTTP –Standards and Design Principles,” in Proc. of ACM Int. Conf. Multimedia, pp. 133-144, 2011.
[15] S. Akhshabi, L. Anantakrishnan, C. Dovrolis, and Ali C. Begen, “What happens when HTTP adaptive streaming players compete for bandwidth,” In Proc. of ACM Workshop on Network and Operating System Support for Digital Audio and Video, pp. 9-14, 2012.
[16] V. Jacobson, “Congestion avoidance and control,” In Proc. of ACM SIGCOMM, pp. 314–329, 1988.
[17] X. Zhu, Z. Zhi, R. Pan, J. Gahm, and H. Hu, “Fixing Multi-Client Oscillations In HTTP-based Adaptive Streaming: A Control Theoretic Approach,” IEEE International Workshop on Multimedia Signal Processing, pp. 230-235, 2013.
[18] L. De Cicco, V. Palmisano, and S. Mascolo, “ELASTIC: a Client-side Controller for Dynamic Adaptive Streaming over HTTP,” Packet Video Workshop, 2013.
[19] Villa, Bjørn J., and Poul E. Heegaard, “Improving perceived fairness and QoE for adaptive video streams,” The Eighth International Conference on Networking and Services. 2012