Abstract—Dealing with a high churn rate is very challenging in live peer-to-peer streaming. State-of-the-art studies try to mitigate the problem by exploiting peer dynamic models, analyzing traces from real world systems, or using enhanced coding techniques, e.g., network coding. Applications of social networking in peer-to-peer systems, especially on file sharing, have recently received research attention. In such systems, the establishment of connections among peers is based on social relationships among users, which are, however, not formed in the context of a peer-to-peer session but, e.g., imported from other social networks. Since friends in such a separate social network do not always have similar interests, they may not necessarily join or stay long in the same peer-to-peer session. We believe that a tight integration between the high level social network of users and the low level overlay of peers would bring significant benefits in dealing with high churn rates and providing personalized streaming services. This paper presents Stir, the first attempt towards an integrated social peer-to-peer streaming system. The key feature of Stir is that social relationships among users are spontaneously formed in a streaming session, and can be exploited directly by the underlying streaming protocol. Stir users, who join the same session, can make friends by means of spontaneous communication, e.g., instant messaging. Such social network formation provides a reliable indication to deal with high churn rate. Our simulations with real social data and peer dynamic traces have demonstrated the benefits of Stir and shed light on building such a system in practice. I.

We have recently witnessed an explosion of user-generated content (UGC) sharing, particularly video clips, as the new killer Internet application. Given the sheer amount of resource demands, the peer-to-peer (or peer-assisted) model has been suggested for this new service scenario. There are however a series of unique challenges from the UGC videos to be addressed, in particular, their significantly shorter lengths. As such, any delay, even being minor as compared to those for conventional movie-like videos, will be perceptually amplified. Given the much more stringent delay requirement, the UGC video sharing thus calls for sophisticated scheduling to provide quality playback. In this paper, we propose a novel collaborative delay-aware scheduling (CODAS) that is customized for the short UGC videos. CODAS improves playback quality and reduces server workload, through adaptive prioritization of data requests and tighter collaboration between peer suppliers and the server. We present detailed design and optimization of CODAS, particularly the synergy policies in different zones of a shrinking window. We evaluate it through extensive tracedriven simulations and PlanetLab prototype experiments, and the results show the great improvement over the stateof-the-art solutions.

The last five years have witnessed an explosion of networked video sharing, represented by YouTube, as a new killer Internet application. Their sustainable development however is severely hindered by the intrinsic limit of their client/server architecture. A shift to the peer-to-peer paradigm has been widely suggested with success already shown in live video streaming and movieon-demand. Unfortunately, our latest measurement demonstrates that short video clips exhibit drastically different statistics, which would simply render these existing solutions suboptimal, if not entirely inapplicable. Our long-term measurement over five million YouTube videos, on the other hand, reveals interesting social networks with strong correlation among the videos, thus opening new opportunities to explore. In this article, we present NetTube, a novel peerto-peer assisted delivering framework that explores the user interest correlation for short video sharing. We address a series of key design issues to realize the system, including a bi-layer overlay, an efficient indexing scheme, a delay-aware scheduling mechanism, and a prefetching strategy leveraging interest correlation. We evaluate NetTube through both simulations and prototype experiments, which show that it greatly reduces the server workload, improves the playback quality and scales well.

Social networked applications have been more and more popular, and have brought great challenges to the network engineering, particularly the huge demands of bandwidth and storage for social media. The recently emerged content clouds shed light on this dilemma. Towards the migration to clouds, partitioning the social contents has drawn significant interests from the literature. Yet the existing works focus on preserving the social relationship only, while an important factor, user access pattern, is largely overlooked. In this paper, by examining a large collection of YouTube video data, we first demonstrate that partitioning the network entirely based on social relationship would lead to unbalanced partitions in terms of access. We further analyze the role of social relationship in the social media applications, and conclude that user access pattern should be taken into account and social relationship should be dynamically preserved. We formulate the problem as a constrained k-medoids clustering problem, and propose a novel Weighted Partitioning Around Medoids (wPAM) solution. We present a dissimilarity/similarity metric to facilitate the preservation of the social relationship. We compare our solution with other state-of-the-art algorithms, and the preliminary results show that it significantly decreases the access deviation in each cloud server, and flexibly preserves the social relationship.

Abstract. User-generated content (UGC) site has become a new killer Internet application in the recent four years. Among those popular sites, YouTube is the most representative and successful one providing a new generation of short video sharing service. Today, YouTube is a dominant provider of online video in the Internet, and is still growing fast. Understanding the features of YouTube and similar video sharing sites is thus crucial to their sustainable development and to network traffic engineering. We investigate the YouTube site from two perspectives, internal and external. Using tracescrawledina1.5-year span, we systematic measure the characteristics of YouTube videos. We find that YouTube videos have noticeably different statistics compared to traditional streaming videos, ranging from length, access pattern, to their active life span. The series of datasets also allows us to identify the growth trend of this fast evolving Internet site in various aspects, which has seldom been explored before. We also look closely at the social networking aspect of YouTube, as this is a key driving force toward its success. In particular, we find that the links to related videos generated by uploaders ’ choices form a small-world network. This suggests that the videos have strong correlations with each other, and creates opportunities for developing novel caching or peer-to-peer distribution schemes to efficiently deliver videos to end users. We also provide an in-depth study into the effects of the external links of YouTube. We collected nearly one million videos ’ external link information, and traced different types of videos for more than two months. Our study shows interesting characteristics of external links of YouTube. In particular, we find that views from external links are independent from total views in each category. Also, videos benefit more from external links in the early stage. Our work can serve as a initial step for the study of the external environment. 1

Abstract — The requirement for supporting DVD-like functions raises new challenges to the design of P2P VoD systems. The uncertainty of frequent user DVD-like interactivity makes it difficult to ensure user perceived Quality of Experience (QoE) for real-time streaming services over distributed self-organized P2P overlay networks. Most existing solutions are based on the unreasonable assumption that all the users in P2P VoD systems have the same preference. Few attention has been paid to personalization, which accommodates the differences between users. In this paper, we present a video model which characterizes the personalization information for users ’ contents and preferences. Based on this model, we develop APEX, a practical personalization framework for P2P VoD applications. APEX makes the personalization practical by using a hybrid architecture which leverages the offline pattern mining on the server side and online collaborative filtering on the peer side. Furthermore, APEX helps peers to personalize navigation, prefetching and membership management, aiming at improving QoE for DVD-like functions by reducing response latency and optimizing content sharing. Both theoretical analysis and comprehensive simulations show that APEX outperforms most existing schemes in terms of accumulated hit ratio, response latency, and searching efficiency. I.

Abstract—User Generated Content (UGC) video applications, such as YouTube, are enormously popular. UGC systems can potentially reduce their distribution costs by allowing peers to store and redistribute the videos that they have seen in the past. We study peer-assisted UGC from three perspectives. First, we undertake a measurement study of the peer-assisted distribution system of Tudou (a popular UGC network in China), revealing several fundamental characteristics that models need to take into account. Second, we develop analytical models for peer-assisted distribution of UGC. Our models capture essential aspects of peer-assisted UGC systems, including system size, peer bandwidth heterogeneity, limited peer storage, and video characteristics. We apply these models to numerically study YouTube-like UGC services. And third, we develop analytical models to understand the rate at which users would install P2P client applications to make peer-assisted UGC a success. Our results provide a comprehensive study of peer-assisted UGC distribution, exposing its fundamental characteristics and limitations. I.