Mediacoop:层次化寻址提升P2P-VoD系统随机跳跃效率

5 浏览量更新于2024-07-15 收藏 1.78MB PDF 举报

本文主要探讨了在P2P视频点播(Peer-to-Peer Video-on-Demand, P2P-VoD)系统中提供层次化的查找服务。现有的P2P-VoD系统为了支持随机播放，需要一种高效的方式来寻找满足"最佳"供应商，这里的"最佳"不仅指内容匹配，还包括网络质量匹配。然而，大多数研究倾向于使用分布式哈希表(Distributed Hash Table, DHT)进行内容查找，这种方法在满足网络质量要求和适应视频流媒体特性方面存在不足，因为其开销较大。针对这一问题，研究人员提出了Mediacoop，一个结合内容和质量匹配的新型层次化查找方案。Mediacoop利用播放位置信息，能够有效地定位到具有所需数据（即内容匹配）的潜在供应商。接着，它在这些候选人中进行进一步的精细查找，以确保网络质量匹配。论文通过理论分析和模拟实验展示了Mediacoop相较于传统的DHT-based方法，能够显著降低跳跃延迟并有效控制开销。 Mediacoop的优势在于其新颖的设计，它解决了P2P-VoD系统中的查找效率和服务质量之间的平衡问题。通过层次结构和针对性的搜索策略，它提高了系统的整体性能，减少了用户在寻找合适供应商时的时间消耗，从而提升了用户的观看体验。这种解决方案对于提高P2P-VoD系统的稳定性和用户体验具有重要意义，对于未来P2P网络架构的优化和实时流媒体服务的发展提供了有价值的参考。

Providing Hierarchical Lookup Service for P2P-VoD Systems · 5

4. DESIGN OF MEDIACOOP

In this section, we ﬁrst address the importance of the end-to-end delay for P2P-VoD and give the

method of how to obtain the delay in Section 4.1. Then we present the ﬁnger tables utilizing the

distance of playpoint and IP preﬁx in Section 4.2. How to conduct lookup operation using the ﬁnger

table is shown in Section 4.3. Finally, in Section 4.4 we describe the maintenance of Mediacoop.

4.1 Exploit Network Properties

In real-time media systems, lower end-to-end delay leads to less waiting time and hence improves

the interactivity [Liu 2007; Noh et al. 2008; Ren et al. 2006]. Especially in VoD services, the user

interactivity frequently happens [Huang et al. 2008], and the large delay between peers results in

long response time. Therefore, the end-to-end delay is adopted to measure the network property of

supplying peers. One might argue how to get the exact value of the end-to-end delay, because it

varies with time. In practice, we only need to record the average delays for diﬀerent candidates, and

choose the smaller ones as suppliers. So we do not require the exact value of the delay. Then the

question is how to get the end-to-end delay. The Internet is composed of many Autonomous Systems

(ASes). Peers within one AS are usually close to each other and inter-AS routing is speciﬁed by

Border Gateway Protocol (BGP). Thus, we only need to know the AS-AS delays or cluster level

delays and choose suppliers from those clusters whose delay is the least with the initiator. We use

the method proposed in [Ren et al. 2006] to construct the AS topology and obtain AS-AS delays.

A sketch of the method is brieﬂy described as follows (cf. [Ren et al. 2006] for more details).

Firstly, we collect a large number of public BGP routing tables and BGP updates, such as those

from RouteViews [RouteViews 2010] and RIPERIS [RIPERIS 2010]. From these routing tables, we

build an AS graph and group IPs with the same longest preﬁx into one cluster. Then, we randomly

choose one IP out of each cluster as the cluster delegate. A delay measurement between each pair of

cluster delegates can be estimated by the tool King [Gummadi et al. 2002]. Finally, we obtain two

tables: (1) IP to cluster mapping table (ICMT); (2) Cluster to cluster delay table (CCDT). From

ICMT, a p eer can learn about its own cluster, and get the target clusters from CCDT, which has

the least delay with itself. Fig. 3 illustrates the work procedure of the delay detector, and Table

I shows a CCDT example in CoolFish [CoolFish 2011]. In Mediacoop, the delay detector program

runs and updates ICMT and CCDT all the time. When a peer joins the system, it gets ICMT and

CCDT from the probing server. In order to prevent frequent requests in normal lookup operations,

the peers will ask the server to update ICMT and CCDT at intervals. Now, the key problem is how

to organize these clusters to provide eﬃcient lookup.

IP prefix

extraction

Kin

tool

( C C D T )

Cluster-Cluster delay table

( I C M T )

IP-Cluster mapping table

IPs in P2P network

BGP tables & updates

Fig. 3. Cluster (IP preﬁx) level delay detector. Using this detector we can get the global tables: ICMT and CCDT.

Table I. CCDT of CoolFish. Time: 15:00, 2/9/2009.

Delay from X to Y 159.226.40.* 202.127.200.* 210.72.15.* . . .

159.226.40.* 0 31(ms) 8(ms) . . .

202.127.200.* 31(ms) 0 58(ms) . . .

210.72.15.* 8(ms) 58(ms) 0 . . .

. . . . . . . . . . . . . . .

ACM Transactions on Multimedia Computing, Communications and Applications.

剩余22页未读，继续阅读

weixin_38522636

粉丝: 3
资源: 913

Mediacoop:层次化寻址提升P2P-VoD系统随机跳跃效率

challenges in design and implementation of middlewares for real-time systems

关于P2P的一些非常好的论文

Spring cloud

Deep Reinforcement Learning Approach for UAV-Assisted Mobile Edge Computing Networks

ipython: error: argument --user: expected one argument

android App Store Service

Facile Synthesis of Hollow Cu2−xS Microspheres as High Performance Anode Materials for Lithium-Ion Batteries

最新资源