Video accelerator is developed for better user experience in video sharing websites such as YouTube. PPLive video accelerator （PPVA）, which has the largest number of users in China, is based on peer-to-peer （P2P） system. The number of videos and peers in PPVA is by orders of magnitude many times larger than which in traditional P2P video on demand （VoD） system. As a result, even though the resource is sufficient, due to unfairness assignment, the quality of service can hardly satisfy all users. In this paper, we concentrate on the assignment of the fundamental resources in PPVA： storage and bandwidth. The problem of storage assignment is formulated as a nonlinear program （NLP） regarding the number of request as a random variable. The results show that the influence of the variance of requests is not negligible and proportional approach is appropriate only when the mean is much larger than the variance. The criteria about how to locate videos to appropriate peers are also presented, taking into account constrains such as the utilization of total bandwidth, the probability of bandwidth competition and the fairness between videos. Furthermore, the heuristic algorithms of allocating upload bandwidth in centralized and distributed fashion are proposed and evaluated against a widely used strategy （equal allocation） with respect to the balance among videos. Simulation results demonstrate that both algorithms can lead to significant performance improvement.
Conventionally, P2P video is regarded as CBR traffic. However, our measurements have shown that the rate reset cannot be neglected in a practical IPTV system because each rate reset often leads to performance degradation. Thus, addressing the problem of inferring playback rate and rate reset in a P2P video system is significant. In this article, an algorithm termed piecewise linear envelope approximation （PLEA） is proposed, in which a follow-up time is introduced to smooth rate fluctuations in a small time scale and to adapt rate jumps in a large time scale automatically. With the PLEA algorithm, discontinuity introduced by blind segmentations adopted by current methods is avoided. Furthermore, unlike existing algorithms in which both segmentation and combinations are performed in multiple runs, only a single computation path is involved in the PLEA algorithm. This makes PLEA algorithm amenable to implementation of low complexity by either software or hardware. Both theoretical analysis and experiment based on measured data show that the PLEA outperforms existing algorithms based on segmentation.
LI Chun-xi CHEN Chang-jia School of Electronics and Information Engineering,Beijing Jiaotong University,Beijing 100044,China
In peer-to-peer （P2P） video on demand （VoD） system, once the P2P downloading rate cannot satisfy the need of playback, server is quickly referred to for help in providing enough bandwidth. Thus the switch of downloading from other peers （P2P） and server （HTTP） exists. This paper uses the proportion of P2P downloading amount （PPDA） during the video watching process to measure server load. This article is interested in finding a better strategy or switch rule between P2P and HTTP downloading for saving server bandwidth. The authors suggest and model a kind of switch rule based on local buffer amount, using mathematical theory of Brownian motion. It can effectively alleviate the impact of P2P rate fluctuation, reduce the switch times and improve the PPDA by at least 3%-5% on the basis of the former switch rule, which means substantial cost can be saved. Particularly the PPDA is related with the ratio of playback bit rate to the HTTP downloading rate which means the PPDA can be restricted by controlling the ratio in the real-world system. Though the result comes from constant bit rate （CBR） video supposition, it provides perspective and method for variable bit rate （VBR） application, and valuable insights for the future development of P2P VoD system.
The design of piece scheduling algorithm is fundamental in mesh-based peer-to-peer （P2P） live streaming systems. This article regards the piece propagation process in P2P live streaming system as a conflict process, where peers compete to get the piece as soon as possible to meet the tight timing requirement of the live system. In this vision, the authors reduce the piece scheduling problem as a conflict resolution problem, and propose a blind random scheduling method to resolve it. The algorithm can be described as： 1） pieces are scheduled independently; 2） conflicts are resolved by the random back off algorithm. The theoretical analysis and simulation results prove that the algorithm scales well with the network size, and can obtain near-optimal in-time delivery of pieces. The algorithm is also easy to implement in the real-world system.
CHEN Yi-shuai CHEN Chang-jia ZHAO Yong-xiang LI Chun-xi
In peer-to-peer （P2P） video-on-demand （VoD） streaming systems, each peer contributes a fixed amount of hard disk storage （usually 2 GB） to store viewed videos and then uploads them to other requesting peers. However, the daily hits （namely popularity） of different segments of a video is highly diverse, which means that taking the whole video as the basic storage unit may lead to redundancy of unpopular segment replicas and scarcity of popular segment replicas in the P2P storage network. To address this issue, we propose a video slicing mechanism （VSM） in which the whole video is sliced into small blocks （20 MB, for instance）. Under VSM, peers can moderately remove unpopular blocks from and accordingly add popular ones into their contributed hard disk storage, which increases the usage of peers＇ contributed resource （storage and bandwidth）. To reasonably assign bandwidth among peers with different download capacity, we propose a moderate prefetching strategy （MPS） based on VSM. Under MPS, when the amount of prefetched content reaches the predefined threshold, peers immediately stop prefetching video content and then release occupied bandwidth for others. A stochastic model is established to analyze the performance of the MPS and it is found that perfect playback continuity can be got under MPS. Then the MPS is applied to PPLive VoD system （one of the largest P2P VoD systems in China） and measurement results demonstrate that low server load and perfect user satisfaction can be achieved. Also, the server bandwidth contribution of PPLive VoD system under MPS （namely 5%） is much lower than that of UUSee VoD system （namely 30%）.
