Supporting non linear and non continuous media access in peer to peer multimedia systems

SUPPORTING NON-LINEAR AND NON-CONTINUOUS MEDIA ACCESS IN PEER-TO-PEER MULTIMEDIA SYSTEMS ZHAO ZHENWEI B.Comp.(Hons.), NUS A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY NUS GRADUATE SCHOOL FOR INTEGRATIVE SCIENCES AND ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2013 Declaration I hereby declare that this thesis is my original work and it has been written by me in its entirety. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. Zhao Zhenwei February 10, 2014 i Acknowledgements First and foremost, I would like to express my sincere gratitude to my advisor Prof. Wei Tsang Ooi, for his continuous guidance and support during my course of study. In the past four years, he had trained me not only on how to conduct research, but much more than that, including technical writing, communication, and social interaction skills. I will carry forward the spirit of self-motivation and independence, which he taught me during my study. Without his help, I would not be able to finish my Ph.D study and this thesis would have been nowhere. I also would like to express my great thanks to Prof. Roger Zimmermann, Prof. Ben Leong, and Prof. Mehul Motani. They kindly agreed to serve in my thesis advisory committee. They have been giving me valuable advice and help me move in the right directions during my study. I also want to thank Prof. Yong Chiang Tay for his guidance on analytical modeling. The analytical skills that I learned from him benefits me a lot, be it now or in the future. I would like to thank M.Sc. Sameer Samarth for his collaboration on one of my works. I also want to express my gratitude to Mr. Ngo Quang Minh Khiem, Dr. Guntur Ravindra, Mr. Manoranjan Mohanty, and Mr. Wang Hui. They frequently discuss and exchange their views with me. Moreover, they have kindly help me on several paper proof reading. Great thanks are given to M.Sc. Chanaka Aruna Munasinge for providing me the Second Life traces. Special thanks are given to my friends Mr. Guo Xiangfa and Mr. Wang Wei. We have discussed a number of research problems, and I benefit a lot from the discussion. Finally, I want to dedicate this thesis to my parents. They give me not only the unconditional love and support, but also the freedom to pursue my dreams. Their encouragement and understanding had bailed me out while I was under great pressure. I would not have gone so far without their support and encouragement. I’m in their debt. ii Contents Summary viii List of Tables x List of Figures xi List of Acronyms xiv Introduction 1.1 Representative Applications . . . . . . . . . . . . . . . . . . . . 1.2 Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 Prefetching . . . . . . . . . . . . . . . . . . . . . . . . 1.2.2 Understanding the Effect of User Interactions . . . . . . 1.2.3 Content Discovery . . . . . . . . . . . . . . . . . . . . 1.2.4 Request and Service Scheduling . . . . . . . . . . . . . 10 1.3 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.3.1 Understanding the Effect of VCR Operations on the Server Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.3.2 Access Pattern-Driven Content Discovery Middleware . 12 1.3.3 Joint Request and Service Scheduling . . . . . . . . . . 13 iii CONTENTS Background and Related Works 15 2.1 P2P Streaming System Design . . . . . . . . . . . . . . . . . . 15 2.2 Analytical Models of P2P Systems . . . . . . . . . . . . . . . . 17 2.3 2.4 2.2.1 BitTorrent File Sharing Systems . . . . . . . . . . . . . 18 2.2.2 P2P VoD Streaming Systems . . . . . . . . . . . . . . . 20 Content Discovery in P2P Media Streaming Systems . . . . . . 21 2.3.1 Centralized Approach . . . . . . . . . . . . . . . . . . . 21 2.3.2 Gossip-based Approach . . . . . . . . . . . . . . . . . . 21 2.3.3 Indexing Tree-based Approach . . . . . . . . . . . . . . 23 2.3.4 DHT-based Approach . . . . . . . . . . . . . . . . . . . 25 2.3.5 Cell-based Approach . . . . . . . . . . . . . . . . . . . 26 2.3.6 Social-based Approach . . . . . . . . . . . . . . . . . . 27 Request and Service Scheduling . . . . . . . . . . . . . . . . . 28 2.4.1 P2P Live Streaming . . . . . . . . . . . . . . . . . . . . 28 2.4.2 P2P VoD Streaming . . . . . . . . . . . . . . . . . . . . 30 2.4.3 P2P NVE Streaming . . . . . . . . . . . . . . . . . . . 32 2.5 Prefetching Algorithm . . . . . . . . . . . . . . . . . . . . . . . 33 2.6 User Behavior Study . . . . . . . . . . . . . . . . . . . . . . . 34 2.6.1 VoD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.6.2 Networked Virtual Environment . . . . . . . . . . . . . 36 2.6.3 Others . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 P2PVCR: Modeling VCR Operations 3.1 38 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 iv CONTENTS 3.2 Systems Model . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.3 Analytical Model . . . . . . . . . . . . . . . . . . . . . . . . . 41 3.4 3.5 3.6 3.3.1 Characterizing Seek and Pause . . . . . . . . . . . . . . 41 3.3.2 Estimating the Gap Size . . . . . . . . . . . . . . . . . 45 3.3.3 Estimating the Server Load . . . . . . . . . . . . . . . . 48 3.3.4 Random Departure . . . . . . . . . . . . . . . . . . . . 52 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.4.1 Multiple Video Approach . . . . . . . . . . . . . . . . . 56 3.4.2 Data Availability . . . . . . . . . . . . . . . . . . . . . 57 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.5.1 Simulation Setup . . . . . . . . . . . . . . . . . . . . . 60 3.5.2 User Interaction Parameter Details . . . . . . . . . . . . 60 3.5.3 Model Validation . . . . . . . . . . . . . . . . . . . . . 62 3.5.4 Comparing the Effect of Different Distribution Types . . 68 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 APRICOD: Access Pattern-Driven Content Discovery 72 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.2 General System Model . . . . . . . . . . . . . . . . . . . . . . 74 4.3 System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 4.3.1 Peer Navigation Model . . . . . . . . . . . . . . . . . . 77 4.3.2 Query Resolution . . . . . . . . . . . . . . . . . . . . . 78 4.3.3 Peer Failure and Flash Crowd . . . . . . . . . . . . . . 79 4.3.4 Registration and Deregistration . . . . . . . . . . . . . . 80 v CONTENTS 4.3.5 4.4 Link and Peer Prefetching . . . . . . . . . . . . . . . . 83 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.4.1 Query Hit Rate . . . . . . . . . . . . . . . . . . . . . . 85 4.4.2 Relation to Prefetching . . . . . . . . . . . . . . . . . . 88 4.5 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.6 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 4.7 4.6.1 Trace Collection and Simulation Setup . . . . . . . . . . 93 4.6.2 Examining Correlations in the Traces . . . . . . . . . . 94 4.6.3 Illustration of Correlations Using the VoD Trace . . . . . 95 4.6.4 Different APRICOD Variants . . . . . . . . . . . . . . . 97 4.6.5 Evaluation of Lookup Hops and Latency . . . . . . . . . 98 4.6.6 Effect of Various System Parameters . . . . . . . . . . . 105 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Joserlin: Joint Request and Service Scheduling 111 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 5.2 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 5.3 On-demand Requests . . . . . . . . . . . . . . . . . . . . . . . 116 5.4 5.5 5.3.1 Request Binning Algorithm . . . . . . . . . . . . . . . . 117 5.3.2 Service Policy and Rejection Policy . . . . . . . . . . . 118 Prefetch Requests . . . . . . . . . . . . . . . . . . . . . . . . . 121 5.4.1 Prefetch Gain Function . . . . . . . . . . . . . . . . . . 122 5.4.2 Prefetch Request Issuing Algorithm . . . . . . . . . . . 125 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 vi CONTENTS 5.6 5.5.1 Trace Collection and Parameter Settings . . . . . . . . . 127 5.5.2 Performance Comparison . . . . . . . . . . . . . . . . . 128 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Conclusion and Future Work 139 6.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 6.2.1 Quantifying the Amount of Non-linear and Non-continuous Media Accesses . . . . . . . . . . . . . . . . . . . . . . 141 6.2.2 Automatic Access Path Recommendation . . . . . . . . 142 6.2.3 Neighborhood Maintenance . . . . . . . . . . . . . . . 143 6.2.4 Layered Coding . . . . . . . . . . . . . . . . . . . . . . 144 Bibliography 145 Appendix A Related Publications 157 vii Summary Interactive media have become a trend. Examples of interactive media include, but are not limited to, Video-on-Demand (VoD), Networked Virtual Environment (NVE), Massively Multiplayer Online Game (MMOG), Google Earth, zoomable video, and free-viewpoint video. The ubiquity of user interactions causes user access patterns to become non-linear and non-continuous in interactive media. Peer-to-Peer (P2P) streaming systems are widely adopted to deliver media content due to their proven scalability and low operating cost. Non-linear and non-continuous access patterns, however, pose non-trivial challenges on P2P streaming systems, thanks to the uncertainties in user interactions. In this thesis, we work towards addressing three major challenges: understanding the effect of user interactions, fast content discovery, and request and service scheduling, in order to provide good system support for streaming interactive media over P2P systems. First, we try to understand how user access patterns affect P2P streaming systems’ performance. We pick the P2P VoD scenario and analytically study how VCR (Video Cassette Recording) operations, such as forward seeks and pauses, affect the streaming system performance, in particular, the server cost. The resulting analytical model can help us understand the relationship between user interactions and system performance. With this model, we find that forward seeks and pauses may potentially increase the server load when coupled with an imperfect prefetching algorithm (e.g. sequential prefetching). Further, either small or large seek distance and pause time are beneficial in terms of server load, as opposed to medium ones. More interestingly, interaction patterns with larger variations tend to incur less server load. Second, we propose APRICOD, an access-pattern-driven content discovery caching middleware, to meet the short content discovery latency requirement during non-continuous accesses. APRICOD exploits correlations among media objects accessed by users and actively adapts its overlay structure to optimize the viii SUMMARY performance as user access patterns change. APRICOD can effectively resolve all continuous access queries with a single hop deterministically (with node failure as exception) and can resolve a significant portion of non-continuous access queries with a single hop. Third, we devise a joint request and service scheduling scheme named Joserlin to efficiently schedule requests in non-linear access scenarios. With nonlinear accesses, data availability in neighborhood changes fast and prefetch misses become the norm, causing many on-demand requests that have to be served within a stringent time limit. Joserlin helps avoid request contention both within the same type and between different types of requests. More importantly, we systematically study the interplay between on-demand and prefetch requests, and jointly schedule them based on a derived gain function. Our evaluation shows that Joserlin reduces the server load by 20% ∼ 60% compared to existing state-of-the-art solutions. Supporting non-linear and non-continuous access patterns in P2P systems is a relatively new research area, where not much prior work exists. This thesis formalizes non-linear and non-continuous access patterns and addresses the aforementioned three major challenges. Work in this thesis can help scalably stream interactive media to a large pool of users and retain good user experience during user interactions. ix CHAPTER 6. CONCLUSION instance, given the current accessed media object di , we can let the next accessed media object be a random variable H that draws its value from the space of accessible media objects. H may follow different distributions for different user access patterns. A potential metric may be derived from the distribution function of H (e.g., the variance value). This guess, however, is just our preliminary idea towards addressing this problem. Deriving a well-accepted metric is still worth a careful study and requires non-trivial work. 6.2.2 Automatic Access Path Recommendation Figure 6.2: Auto access path recommendation. As discussed in Section 1.2 and confirmed by user studies in Section 2.6.1, the user interface (UI) design can subconsciously affect user access patterns. Therefore, another research direction that we can potentially work in is to study how to properly design the UI to guide user access patterns, not from an aesthetic point of view, but from the system perspective. Uncertainties in user accesses can be reduced significantly by guiding the user access patterns, leading to improvement in system performance (e.g., higher prefetch hit rate). A specific UI feature that we have in mind is automatic access path recommendation. That is we try to recommend navigation paths for a particular user as shown in Figure 6.2. In general, users would be more likely to follow one of the recommended paths, reducing the amount of uncertainties and increasing the predictability accordingly. Further, we can snap users’ positions to a recommended path if they are very close to it. Path snapping does not degrade user experience much, but may benefit the system a lot. Automatic access path recommendation faces non-trivial challenges. First, users’ navigation paths have to be collected and clustered. Next, we need to classify users into different classes based on their navigation patterns. Finally, 142 CHAPTER 6. CONCLUSION we guess the user class that a new user may belong to and recommend the paths that users in that class tend to take. Even though there have been a few works targeting at some specific components such as path clustering [57], we are still far from a fully fledged automatic path recommendation system. 6.2.3 Neighborhood Maintenance Another problem worth a careful study is neighborhood maintenance. In P2P streaming systems, establishing a neighborhood is not the end of the story. Peers’ neighborhood needs to be maintained and updated over time, mainly due to peer departure, changes in data availability, etc. Maintenance of neighborhood in face of peer departure has been well studied in existing literature. Updating neighborhood according to changes in data availability, however, has not been well addressed. In VoD, where user access patterns are largely linear, when to replace a neighbor is easy to decide due to the linear playback property. For instance, if a neighbor does not have the requested content at the moment, it is likely not to have the requested content in the near future. Therefore, we can go ahead to replace this neighbor. In scenarios where non-linear access patterns are intrinsic (such as networked virtual environment), when to replace a neighbor is not that obvious. As even if a neighbor does not possess the requested content now, it may possess the requested content later due to uncertainties in user accesses. Besides, we cannot change peers’ neighborhood too frequently. Otherwise, not only large overhead is incurred, normal streaming service can be disrupted as well. A balance point needs to be found on how frequently we should update a peer’s neighborhood. The neighborhood maintenance problem can be formulated as follows: Given a set of peers, their cached contents, and their respective current positions in the resource space, we may find a good (or optimal if possible) arrangement R to construct the overlay through a peer selection process. As peers fetch and prefetch content, their cached contents change over time. As peers navigate in the resource space, their positions change as well. With the changes of all these system conditions, the original arrangement R may no longer be good (or optimal). Therefore, we need a neighborhood maintenance algorithm to adapt to the changing system conditions and update the arrangement R over time to retain a high streaming efficiency. 143 CHAPTER 6. CONCLUSION 6.2.4 Layered Coding Layered coding such as scalable video coding (SVC) has been widely exploited for adaptive media streaming [35, 50]. With layered coding, media content is encoded into one base layer and a few refinement layers. All layers form a dependency chain, with the base layer at the bottom. To successfully render the content, the base layer has to be present. Further, the more refinement layers are received, the better quality we get. Note that the dependencies among different layers have to be respected. Layered coding may potentially bring lots of benefits. First, we can adapt the content quality according to the current system conditions. If there is not enough upload bandwidth, instead of retaining a high playback quality but making user accesses very jitter, we can reduce the quality by fetching less refinement layers and ensure smooth user accesses. Second, after non-continuous accesses, peers often need to perform content discovery, during which user accesses are stalled, awaiting for the new content to arrive. Khiem et al.’s user behavior study with zoomable video shows that the presence of a low quality version is better than having nothing [58]. The low quality version significantly increases the user tolerance to content retrieval latency. Therefore, layered coding can potentially be used to buy some time after user interactions. Third, layered coding can be exploited to support heterogeneous devices such as wired PCs and mobile phones over 3G/WiFi. Therefore, applying layered coding to support non-linear and non-continuous accesses is an interesting direction to explore. In general, its adoption can alleviate the demand on short content discovery latency after non-continuous accesses. Layered coding, however, may increase the request and service scheduling complexity as the scheduling units become more fine-grained and dependencies exist among different layers. Further, it is worth noticing that frequent adaption of playback quality may potentially harm user experience. Hence, we suggest that the adoption of layered coding should go side by side with appropriate user behavior studies. In conclusion, supporting non-linear and non-continuous media accesses in P2P streaming systems is a challenging topic. The essence of this topic is to deal with uncertainties in a highly distributed and dynamic environment. We spot and address three major challenges in this thesis. Continuous research input is needed to address others. 144 Bibliography [1] Broadband Reports and Speed Test Statistics. dslreports.com/archive. http://www. [2] Freebies. http://maps.secondlife.com/secondlife/ Freebies/127/114/65. Accessed on 06-June-2012. [3] FreePastry. http://www.freepastry.org/. Accessed on 30Nov-2013. [4] Korea2. http://maps.secondlife.com/secondlife/ Korea2/128/128/26. Accessed on 06-June-2012. [5] Oversim. http://www.oversim.org/. Accessed on 06-February2013. [6] Sunland. http://slurl.com/secondlife/SUNLAND/55/ 210/22. Accessed on 15-August-2011. [7] S. Aalto, P. E. Lassila, N. Raatikainen, P. Savolainen, and S. Tarkoma. P2P Video-on-Demand: Steady State and Scalability. In Proc. of GLOBECOM’10, pages 1–5, Miami, Florida. [8] U. Abbasi and T. Ahmed. COOCHING: Cooperative Prefetching Strategy for P2P Video-on-Demand System. In Wired-Wireless Multimedia Networks and Services Management, volume 5842 of Lecture Notes in Computer Science, pages 195–200. Springer Berlin Heidelberg, 2009. [9] U. Abbasi, G. Simo, and T. Ahmed. Differentiated chunk scheduling for P2P Video-on-Demand system. In Proc. of CCNC’11, pages 622–626, Las Vegas, NV. [10] J. M. Almeida, J. Krueger, D. L. Eager, and M. K. Vernon. Analysis of Educational Media Server Workloads. In Proc. of NOSSDAV’01, pages 21–30, Port Jefferson, NY. 145 BIBLIOGRAPHY [11] S. Annapureddy, S. Guha, C. Gkantsidis, D. Gunawardena, and P. R. Rodriguez. Is high-quality vod feasible using P2P swarming? In Proc. of WWW’07, pages 903–912, Banff, Alberta, Canada. [12] M. E. Bertinat, D. Padula, F. R. Amoza, P. Rodr´ıguez-Bocca, and P. Romero. Optimal bandwidth allocation in mesh-based peer-to-peer streaming networks. In Proc. of INOC’11, pages 529–534, Hamburg, Germany. [13] A. R. Bharambe, C. Herley, and V. N. Padmanabhan. Some Observations on BitTorrent Performance. In Proc. of SIGMETRICS’05, pages 398– 399, Banff, Alberta. [14] T. Bonald, L. Massouli´e, F. Mathieu, D. Perino, and A. Twigg. Epidemic live streaming: optimal performance trade-offs. SIGMETRICS Perform. Eval. Rev., 36(1):325–336, June 2008. [15] A. Brampton, A. MacQuire, I. A. Rai, N. J. P. Race, L. Mathy, and M. Fry. Characterising User Interactivity for Sports Video-on-Demand. In Proc. of NOSSDAV’07, pages 99–104, Urbana-Champaign, IL. [16] A. Carlier, R. Guntur, and W. T. Ooi. Towards characterizing users’ interaction with zoomable video. In Proc. of SAPMIA’10, pages 21–24, Firenze, Italy. [17] Z. Chen, K. Xue, and P. Hong. A Study on Reducing Chunk Scheduling Delay for Mesh-Based P2P Live Streaming. In Proc. of GCC’08, pages 356–361, Shenzhen, China. [18] B. Cheng, H. Jin, and X. Liao. Supporting VCR Functions in P2P VoD Services Using Ring-Assisted Overlays. In Proc. of ICC’07, pages 1698– 1703, Glasgow, Scotland, UK. [19] B. Cheng, X. Liu, Z. Zhang, and H. Jin. A Measurement Study of a Peer-to-Peer Video-on-Demand System. In Proc. of IPTPS’07, Bellevue, WA. [20] B. Cheng, L. Stein, H. Jin, X. Liao, and Z. Zhang. GridCast: Improving Peer Sharing for P2P VoD. ACM Trans. Multimedia Comput. Commun. Appl., 4(4):1–31, 2008. 146 BIBLIOGRAPHY [21] W. Cheng, D. Liu, and W. T. Ooi. Peer-assisted view-dependent progressive mesh streaming. In Proc. of MM’09, pages 441–450, Beijing, China. [22] X. Cheng and J. Liu. NetTube: Exploring Social Networks for Peer-toPeer Short Video Sharing. In Proc. of INFOCOM’09, pages 1152–1160, Rio de Janeiro, Brazil. [23] M. Chesire, A. Wolman, G. M. Voelker, and H. M. Levy. Measurement and Analysis of a Streaming-Media Workload. In Proc. of USITS’01, San Francisco, CA. [24] H. Chi, Q. Zhang, J. Jia, and X. (sherman) Shen. Efficient search and scheduling in P2P-based media-on-demand streaming service. IEEE Journal on Selected Areas in Communications, 25:119–130, 2007. [25] C.-H. Chien, S.-Y. Hu, J.-R. Jiang, and C.-W. Cheng. Bandwidth-Aware Peer-to-Peer 3D streaming. IEEE Journal on Selected Areas in Communications, 4(4):324–342, Nov. 2010. [26] J. H. P. Chim, R. W. H. Lau, H. V. Leong, and A. Si. CyberWalk: a webbased distributed virtual walkthrough environment. IEEE Transactions on Multimedia, 5(4):503–515, 2003. [27] Y. R. Choe, D. L. Schuff, J. M. Dyaberi, and V. S. Pai. Improving VoD server efficiency with bittorrent. In Proc. of MM’07, pages 117–126, Augsburg, Germany. [28] M. Claypool and K. Claypool. Latency and player actions in online games. Communications of the ACM, 49:40–45, November 2006. [29] Y. Coady, R. Cox, J. DeTreville, P. Druschel, J. Hellerstein, A. Hume, K. Keeton, T. Nguyen, C. Small, L. Stein, and A. Warfield. Falling Off the Cliff: When Systems Go Nonlinear. In Proc. of HotOS’05, Santa Fe, New Mexico, 2005. [30] C. P. Costa, I. S. Cunha, A. Borges, C. V. Ramos, M. M. Rocha, J. M. Almeida, and B. Ribeiro-Neto. Analyzing Client Interactivity in Streaming Media. In Proc. of WWW’04, pages 534–543, NY. 147 BIBLIOGRAPHY [31] C. Crockford and H. Agius. An empirical investigation into user navigation of digital video using the VCR-like control set. International Journal of Human-Computer Studies, 64(4):340–355, April 2006. [32] Y. Cui, B. Li, and K. Nahrstedt. oStream: Asynchronous Streaming Multicast in Application-Layer Overlay Networks. IEEE Journal on Selected Areas in Communications, 22:91–106, 2003. [33] F. Dabek, M. F. Kaashoek, D. Karger, R. Morris, and I. Stoica. Wide-area cooperative storage with CFS. SIGOPS Oper. Syst. Rev., 35:202–215, October 2001. [34] R. N. De Silva, W. Cheng, W. T. Ooi, and S. Zhao. Towards understanding user tolerance to network latency and data rate in remote viewing of progressive meshes. In Proc. of NOSSDAV’10, pages 123–128, Amsterdam, Netherlands. [35] Y. Ding, J. Liu, D. Wang, and H. Jiang. Peer-to-peer video-on-demand with scalable video coding. Comput. Commun., 33(14):1589–1597, Sept. 2010. [36] T. Do, K. A. Hua, and M. Tantaoui. P2VoD: Providing Fault Tolerant Video-on-Demand Streaming in Peer-to-Peer Environment. In Proc. of ICC’04, pages 1467–1472, Paris, France. [37] P. Garbacki, D. H. J. Epema, J. Pouwelse, and M. van Steen. Offloading servers with collaborative video on demand. In Proc. of IPTPS’08, Tampa Bay, Florida. [38] F. R. Giordano, M. D. Weir, and W. P. Fox. A First Course in Mathematical Modeling. Brooks Cole, edition, 2002. [39] R. Grigoras, V. Charvillat, and M. Douze. Optimizing hypervideo navigation using a Markov decision process approach. In Proc. of MM’02, pages 39–48, Juan-les-Pins, France. [40] L. Guo, S. Chen, Z. Xiao, E. Tan, X. Ding, and X. Zhang. A Performance Study of BitTorrent-like Peer-to-Peer Systems. IEEE Journal on Selected Areas in Communications, 25(1):155–169, 2007. 148 BIBLIOGRAPHY [41] Y. Guo, C. Liang, and Y. Liu. AQCS: adaptive queue-based chunk scheduling for P2P live streaming. In Proc. of NETWORKING’08, pages 433–444, Singapore. [42] Y. Guo, S. Mathur, K. Ramaswamy, S. Yu, and B. Patel. PONDER: Performance Aware P2P Video-on-Demand Service. In Proc. of GLOBECOM’07, pages 225–230, WA. [43] Y. Guo, S. Mathur, K. Ramaswamy, S. Yu, and B. Patel. PONDER: Performance Aware P2P Video-on-Demand Service. In Proc. of GLOBECOM’07, pages 225–230, WA. [44] Y. Guo, K. Suh, J. Kurose, and D. Towsley. P2Cast: Peer-to-peer Patching Scheme for VoD Service. In Proc. of WWW’03, pages 301–309, Budapest, Hungary. [45] Y. Guo, S. Yu, H. Liu, S. Mathur, and K. Ramaswamy. Supporting VCR Operation in a Mesh-Based P2P VoD System. In Proc. of CCNC’08, pages 452–457, Las Vegas, Nevada. [46] Y. He and Y. Liu. VOVO: VCR-Oriented Video-on-Demand in LargeScale Peer-to-Peer Networks. IEEE Transactions on Parallel and Distributed Systems, 20(4):528–539, April 2008. [47] Y. He, G. Shen, Y. Xiong, and L. Guan. Optimal prefetching scheme in P2P VoD applications with guided seeks. IEEE Transactions on Multimedia, 11(1):138–151, Jan. 2009. [48] M. Hefeeda, A. Habib, B. Botev, D. Xu, and B. Bhargava. PROMISE: peer-to-peer media streaming using CollectCast. In Proc. of MM’03, pages 45–54, Berkeley, CA. [49] S.-Y. Hu, J.-F. Chen, and T.-H. Chen. VON: a scalable peer-to-peer network for virtual environments. IEEE Network, 20(4):22–31, July 2006. [50] S.-Y. Hu, J.-R. Jiang, and B.-Y. Chen. Peer-to-Peer 3D Streaming. IEEE Internet Computing, 14:54–61, March 2010. [51] C. Huang, J. Li, and K. W. Ross. Can Internet Video-on-Demand be Profitable? In Proc. of SIGCOMM’07, pages 133–144, Kyoto, Japan. 149 BIBLIOGRAPHY [52] C. Huang, J. Li, and K. W. Ross. Peer-Assisted VoD: Making Internet Video Distribution Cheap. In Proc. of IPTPS’07, Bellevue, WA. [53] C.-M. Huang and T.-H. Hsu. A user-aware prefetching mechanism for video streaming. World Wide Web, 6(4):353–374, Dec. 2003. [54] Y. Huang, T. Z. J. Fu, D.-M. Chiu, J. C. S. Lui, and C. Huang. Challenges, Design and Analysis of a Large-scale P2P VoD System. In Proc. of SIGCOMM’08, pages 375–388, Seattle, WA. [55] K.-W. Hwang, V. Misra, and D. S. Rubenstein. Stored media streaming in bittorrent-like P2P networks. Technical Report CUCS-024-08, Columbia University, New York, April 2008. [56] Q.-j. Ji, H.-j. Zhang, and Z. Yan-qin. One more server for supporting random seeking in a P2P video-on-demand system. In Proc. of CSIP’12, pages 171–174, Xi’an, Shaanxi, China. [57] J.-R. Jiang, C.-C. Huang, and C.-H. Tsai. Avatar Path Clustering in Networked Virtual Environments. In Proc. of ICPADS’10, pages 845–850, Shanghai, China. [58] N. Q. M. Khiem, G. Ravindra, and W. T. Ooi. Towards understanding user tolerance to network latency in zoomable video streaming. In Proc. of MM’11, pages 977–980, Scottsdale, AZ. [59] C. Kir´aly, R. L. Cigno, and L. Abeni. Deadline-Based Differentiation in P2P Streaming. In Proc. of GLOBECOM’10, pages 1–6, Miami, FL. [60] R. Kumar, Y. Liu, and K. Ross. Stochastic Fluid Theory for P2P Streaming Systems. In Proc. of INFOCOM’07, pages 919–927, Anchorage, AK. [61] C. Liang, Y. Guo, and Y. Liu. Is Random Scheduling Sufficient in P2P Video Streaming? In Proc. of ICDCS’08, pages 53–60, Beijing, China. [62] H. Liang, R. De Silva, W. Ooi, and M. Motani. Avatar mobility in usercreated networked virtual worlds: measurements, analysis, and implications. Multimedia Tools and Applications, 45:163–190, 2009. [63] K. Liang, R. Zimmermann, and W. T. Ooi. Peer-assisted texture streaming in metaverses. In Proc. of MM’11, pages 203–212, Scottsdale, AZ. 150 BIBLIOGRAPHY [64] C. L. Liu and J. W. Layland. Scheduling Algorithms for Multiprogramming in a Hard-Real-Time Environment. J. ACM, 20(1):46–61, Jan. 1973. [65] J. Liu and M. Zhou. Tree-Assisted Gossiping for Overlay Video Distribution. Multimedia Tools Appl., 29(3):211–232, June 2006. [66] Y. Liu and M. Hefeeda. Video streaming over cooperative wireless networks. In Proc. of MMSys’10, pages 99–110, Phoenix, AZ. [67] Y. Lu, J. D. Mol, F. Kuipers, and P. V. Mieghem. Analytical Model for Mesh-Based P2PVoD. In Proc. of ISM’08, pages 364–371, Berkeley, CA. [68] N. Magharei and R. Rejaie. Prime: peer-to-peer receiver-driven meshbased streaming. IEEE/ACM Trans. Netw., 17(4):1052–1065, Aug. 2009. [69] D. Marpe, T. Wiegand, and G. J. Sullivan. The H.264/MPEG4 advanced video coding standard and its applications. IEEE Communications Magazine, 44(8):134–143, Aug. 2006. [70] A. Mavlankar and B. Girod. Pre-fetching based on video analysis for interactive region-of-interest streaming of soccer sequences. In Proc. of ICIP’09, pages 3061–3064, Cairo, Egypt. [71] A. Mavlankar and B. Girod. Video Streaming with Interactive Pan/Tilt/Zoom. In High-Quality Visual Experience: Creation, Processing and Interactivity of High-Resolution and High-Dimensional Video Signals. Springer, 2010. [72] N. Megiddo and D. Modha. ARC: A Self-Tuning, Low Overhead Replacement Cache. In Proc. of FAST’03, pages 115–130, San Francisco, CA, 2003. [73] N. Megiddo and D. S. Modha. Outperforming LRU with an Adaptive Replacement Cache Algorithm. Computer, 37(4):58–65, Apr. 2004. [74] T. S. E. Ng and H. Zhang. A network positioning system for the internet. In Proc. of ATEC’04, pages 141–154, Boston, MA. [75] Q. M. K. Ngo, R. Guntur, A. Carlier, and W. T. Ooi. Supporting zoomable video streams with dynamic region-of-interest cropping. In Proc. of MMSys’10, pages 259–270, Phoenix, AZ. 151 BIBLIOGRAPHY [76] J. Padhye and J. Kurose. An empirical study of client interactions with a continuous-media courseware server. In Proc. of NOSSDAV’98, Cambridge, UK. [77] V. Pai, K. Kumar, K. Tamilmani, V. Sambamurthy, and A. E. Mohr. Chainsaw: Eliminating Trees from Overlay Multicast. In Proc. of IPTPS’05, pages 127–140, Ithaca, NY. [78] J. A. Pouwelse, P. Garbacki, J. Wang, A. Bakker, J. Yang, A. Iosup, D. H. J. Epema, M. Reinders, M. R. van Steen, and H. J. Sips. TRIBLER: a social-based peer-to-peer system: Research articles. Concurr. Comput. : Pract. Exper., 20:127–138, February 2008. [79] D. Qiu and R. Srikant. Modeling and Performance Analysis of BitTorrent-Like Peer-to-Peer Networks. In Proc. of SIGCOMM’04, pages 367–378, Portland, Oregon. [80] X. Qiu, C. Wu, X. Lin, and F. C. Lau. InstantLeap: fast neighbor discovery in P2P VoD streaming. In Proc. of NOSSDAV’09, pages 19–24, Williamsburg, VA. [81] V. Ramasubramanian and E. G. Sirer. Beehive: O(1) lookup performance for power-law query distributions in peer-to-peer overlays. In Proc. of NSDI’04, pages 99–112, San Francisco, CA. [82] W. Rao, L. Chen, A. W.-C. Fu, and Y. Bu. Optimal proactive caching in peer-to-peer network: analysis and application. In Proc. of CIKM’07, pages 663–672, Lisbon, Portugal. [83] A. I. T. Rowstron and P. Druschel. Pastry: Scalable, Decentralized Object Location, and Routing for Large-Scale Peer-to-Peer Systems. In Proc. of Middleware’01, pages 329–350, Heidelberg, Germany. [84] P. Shah and J.-F. Pris. Peer-to-Peer Multimedia Streaming Using BitTorrent. In Proc. of IPCCC’07, pages 340–347, New Orleans, LA. [85] Y. Shen, Z. Liu, S. Panwar, K. W. Ross, and Y. Wang. On the Design of Prefetching Strategies in a Peer-Driven Video-on-Demand System. In Proc. of ICME’06, pages 817–820, Toronto, Ontario, Canada. 152 BIBLIOGRAPHY [86] Z. Shen, J. Luo, R. Zimmermann, and A. Vasilakos. Peer-to-Peer Media Streaming: Insights andNew Developments. Proceedings of the IEEE, 12(99):20892109, 2011. [87] W.-L. Sung, S.-Y. Hu, and J.-R. Jiang. Selection strategies for peerto-peer 3D streaming. In Proc. of NOSSDAV’08, pages 15–20, Braunschweig, Germany. [88] S. Tewari and L. Kleinrock. Proportional Replication in Peer-to-Peer Networks. In Proc. of INFOCOM’06, pages 1–12, Barcelona, Catalunya, Spain. [89] M. Varvello, S. Ferrari, E. Biersack, and C. Diot. Exploring second life. IEEE/ACM Trans. Netw., 19(1):80–91, 2011. [90] M. Vilas, X. G. Paneda, R. Garcia, D. Melendi, and V. G. Garcia. User Behaviour Analysis of a Video-On-Demand Service with a Wide Variety of Subjects and Lengths. In Proc. of EUROMICRO’05, pages 330–337, Porto, Portugal. [91] A. Vlavianos, M. Iliofotou, and M. Faloutsos. BiToS: Enhancing BitTorrent for Supporting Streaming Applications. In Proc. of INFOCOM’06, pages 1–6, Barcelona, Catalunya, Spain. [92] N. Vratonji´c, P. Gupta, N. Kneˇzevi´c, D. Kosti´c, and A. Rowstron. Enabling DVD-like features in P2P video-on-demand systems. In Proc. of P2P-TV’07, pages 329–334, Kyoto, Japan. [93] D. Wang and J. Liu. Peer-to-Peer Asynchronous Video Streaming using Skip List. In Proc. of ICME’06, pages 1397–1400, Toronto, Canada. [94] J. Wang, C. Huang, and J. Li. On ISP-friendly rate allocation for peerassisted VoD. In Proc. of MM’08, pages 279–288, Vancouver, British Columbia, Canada. [95] M. Wang and B. Li. R2: Random Push with Random Network Coding in Live Peer-to-Peer Streaming. IEEE Journal on Selected Areas in Communications, 25(9):1655–1666, 2007. [96] X. Wang, C. Zheng, Z. Zhang, H. Lu, and X. Xue. The Design of Video Segmentation-aided VCR Support for P2P VoD Systems. IEEE Transactions on Consumer Electronics, 54(2):531 –537, May 2008. 153 BIBLIOGRAPHY [97] C. Wu, Z. Li, X. Qiu, and F. C. M. Lau. Auction-based P2P VoD streaming: Incentives and optimal scheduling. ACM Trans. Multimedia Comput. Commun. Appl., 8s(1):14:1–14:22, Feb. 2012. [98] T. Xu, W. Wang, B. Ye, W. Li, S. Lu, and Y. Gao. Prediction-Based Prefetching to Support VCR-like Operations in Gossip-Based P2P VoD Systems. In Proc. of ICPADS’09, pages 1–8, Shenzhen, China. [99] X. Yang, M. Gjoka, P. Chhabra, A. Markopoulou, and P. Rodriguez. Kangaroo: video seeking in P2P systems. In Proc. of IPTPS’09, Boston, MA. [100] Y. Yang, A. L. H.Chow, L. Golubchik, and D. Bragg. Improving QoS in bittorrent-like VoD systems. In Proc. of INFOCOM’10, pages 2061– 2069, San Diego, CA. [101] H. Yin, C. Lin, F. Qiu, X. Liu, and D. Wu. TrustStream: a novel secure and scalable media streaming architecture. In Proc. of MM’05, pages 295–298, Hilton, Singapore. [102] H. Yin, X. Liu, T. Zhan, V. Sekar, F. Qiu, C. Lin, H. Zhang, and B. Li. LiveSky: Enhancing CDN with P2P. ACM Trans. Multimedia Comput. Commun. Appl., 6:16:1–16:19, August 2010. [103] L. Ying and A. Basu. pcVOD: Internet peer-to-peer video-on-demand with storage caching on peers. In Proc. of DMS’07, pages 218–223, Hotel Sofitel, San Francisco Bay, CA. [104] W. P. K. Yiu, X. Jin, and S. H. G. Chan. VMesh: Distributed Segment Storage for Peer-to-Peer Interactive Video Streaming. IEEE Journal on Selected Areas in Communications, 25(9):1717–1731, Dec. 2007. [105] A. P. Yu and S. T. Vuong. MOPAR: a mobile peer-to-peer overlay architecture for interest management of massively multiplayer online games. In Proc. of NOSSDAV’05, pages 99–104, Stevenson, WA. [106] H. Yu, D. Zheng, B. Y. Zhao, and W. Zheng. Understanding User Behavior in Large-Scale Video-on-Demand Systems. In Proc. of EuroSys’06, pages 333–344, Leuven, Belgium. [107] Q. Yu and D. Chen. Optimal Data Scheduling for P2P VoD Streaming Systems. In Proc. of ICPADS’10, pages 817–822, Shanghai, China. 154 BIBLIOGRAPHY [108] M. Zhang and B. Feng. A P2P VoD system using dynamic priority. In Proc. of MICC’09, pages 518 – 523, Kuala Lumpur, Malaysia. [109] M. Zhang, Y. Xiong, Q. Zhang, and S. Yang. On the Optimal Scheduling for Media Streaming in Data-driven Overlay Networks. In Proc. of GLOBECOM ’06, pages 1–5, San Francisco, CA. [110] M. Zhang, Q. Zhang, L. Sun, and S. Yang. Understanding the Power of Pull-Based Streaming Protocol: Can We Do Better? IEEE Journal on Selected Areas in Communications, 25(9):1678–1694, Dec. 2007. [111] T. Zhang, F. Ren, and X. Cheng. First-Aid Mechanism for Peer-to-Peer Streaming. In Proc. of ICNP’08, Orlando, FL. [112] X. Zhang, J. Liu, B. Li, and T. shing Peter Yum. CoolStreaming/DONet: A Data-driven Overlay Network for Peer-to-Peer Live Media Streaming. In Proc. of INFOCOM’05, pages 2102–2111, Miami, FL. [113] B. Q. Zhao, J. C.-S. Lui, and D.-M. Chiu. Exploring the Optimal Chunk Selection Policy for Data-Driven P2P Streaming Systems. In Proc. of Peer-to-Peer Computing’09, pages 271–280, Seattle, WA. [114] Z. W. Zhao. Challenges in Supporting Non-linear and Non-Continuous Media Access in P2P Systems. In Proc. of MM’12, pages 1397–1400, Nara, Japan. [115] Z. W. Zhao and W. T. Ooi. APRICOD: An Access Pattern-Driven Distributed Caching Middleware for Fast Content Discovery of Noncontinuous Media Access. In Proc. of MM’11, pages 1241–1244, Scottsdale, AZ. [116] Z. W. Zhao and W. T. Ooi. Joserlin: Joint Request and Service Scheduling for Peer-to-Peer Non-linear Media Access. In Proc. of MM’13, Barcelona, Spain. Accepted. [117] Z. W. Zhao and W. T. Ooi. APRICOD: An Access-Pattern-Driven Distributed Caching Middleware for Fast Content Discovery of Noncontinuous Media Access. ACM Trans. Multimedia Comput. Commun. Appl., 9(2):15:1–15:24, May 2013. 155 BIBLIOGRAPHY [118] Z. W. Zhao, S. Samarth, and W. T. Ooi. Modeling the Effect of User Interactions on Mesh-Based P2P VoD Streaming Systems. ACM Trans. Multimedia Comput. Commun. Appl., 9(2):13:1–13:22, May 2013. [119] Y. Zhou, T. Z. J. Fu, and D. M. Chiu. A unifying model and analysis of P2P VoD replication and scheduling. In Proc. of INFOCOM’12, pages 1530–1538, Orlando, FL. [120] Y. Zhou, T. Z. J. Fu, and D. M. Chiu. Statistical modeling and analysis of P2P replication to support VoD service. In Proc. of INFOCOM’11, pages 945–953, Shanghai, China. 156 Appendix A Related Publications Here we list a few of our publications related to this thesis: • Chapter is largely based on the Doctoral Symposium Abstract published in ACM Multimedia 2012 [114]. • Chapter is based on our published TOMCCAP journal paper [118]. • Chapter is based on our published TOMCCAP journal paper [117], which itself is an extension of our paper [115] published in ACM Multimedia 2011. • Chapter is based on our paper published in ACM Multimedia 2013 [116]. 157 [...]... linear to non- linear and from continuous to non- continuous Designing P2P streaming systems that can support non- linear and non- continuous access patterns well remains a problem Current P2P streaming systems mainly target at scenarios where user access patterns are linear or with a limited amount of user interactions (e.g., seeks in VoD) Non- linear and non- continuous access patterns caused by user interactions... inappropriate to say that a single access is linear or non- linear On the contrary, noncontinuous access patterns are individual behaviors and they are used to refer to individual accesses Figure 1.1 illustrates the concept of linear, non- linear, and non- continuous access patterns In Figure 1.1(a), all peers traverse in the resource space following exactly the same path Such access pattern is linear In Figure... peer C (a) A linear access pattern peer A peer B peer C (b) A non- linear access pattern p2 p1 peer A (c) A non- continuous access pattern Figure 1.1: Non- linear and non- continuous access patterns case, the non- linearity is solely caused by the non- continuity, thanks to the single temporal resource space As a result, non- linear accesses in VoD are rare as there may not be many seeks during a video session... dimensions, including the temporal, spatial, and zoom dimension Non- linear and non- continuous accesses may occur in any one of these dimensions, resulting in much more common non- linear and non- continuous access patterns as observed by Carlier et al [16] Further, similar to the NVE scenario, non- linear accesses exist regardless of non- continuous accesses We want to highlight that the existence of non- linear and. .. streaming systems For example, streaming systems need to respond quickly to user interactions Prior to introducing the challenges posed by non- linear and non- continuous access patterns, we formally define linear, non- linear, and non- continuous media access patterns An access is defined as: Ax = {di → dj }, where di and dj are data units in the resource space S, which consists of all the accessible media. .. that non- linear and non- continuous access patterns pose on P2P streaming systems and summarize our work toward addressing these challenges The rest of this chapter is organized as follows: We give several concrete examples of non- linearly and non- continuously accessed media in Section 1.1; The challenges that non- linear and non- continuous access patterns pose on P2P streaming systems are illustrated in. .. coding only within independent data units As for caching, non- continuous accesses lead to noncontinuously cached content and non- linear accesses lead to skewed popularity of data units Issues mentioned in this paragraph, however, either exist in current streaming systems, except that non- linear and non- continuous accesses make them more severe, or they can be resolved by slightly adapting existing solutions... Applications In this section, we list a few representative interactive media applications, where non- linear and non- continuous accesses are present VoD: VoD allows random seeks, which are considered as non- continuous accesses Seeks also lead to non- linear access patterns: some users may playback continuously and others may seek to different playback positions In this 2 CHAPTER 1 INTRODUCTION peer A peer B peer. .. multiple peers traverse in the resource space following different paths, resulting in a typical non- linear access pattern In Figure 1.1(c), the peer s traversal path is non- continuous (there is a sudden jump from p1 to p2), resulting in a typical non- continuous access pattern Non- linear access patterns create uncertainties in user accesses: given the current accessed data unit, say di , it is uncertain which... to all interactive media applications when streamed using the P2P architecture 1.2.1 Prefetching Obviously, non- linear and non- continuous accesses make prefetching harder compared to linear accesses Without non- continuous accesses, even though 7 CHAPTER 1 INTRODUCTION user access patterns may still be non- linear, the next accessed data unit must be in the vicinity of the current accessed one (can be . patterns shift from linear to non- linear and from continuous to non- continuous. Designing P2P streaming systems that can support non- linear and non- continuous access patterns well remains a problem. Current. INTRODUCTION peer A peer B peer C (a) A linear access pattern peer A peer B peer C (b) A non- linear access pattern peer A p1 p2 (c) A non- continuous access pattern Figure 1.1: Non- linear and non- continuous. to d i , we call this particular access a non- continuous access and the corresponding user access pattern is non- continuous. Definitions 1.1 and 1.2 clearly distinguish that linear and non- linear