On scheduling of peer-to-peer video services
On scheduling of peer-to-peer video services
复制标题
点对点视频业务调度研究
DOI：
10.1109/jsac.2007.070114
复制
发表时间：
2007
影响因子：
16.4
通讯作者：
J. Wong
中科院分区：
文献类型：
--
作者：
Ying Cai;A. Natarajan;J. Wong
点对点（P2P）视频系统为大量主机协作共享视频提供了一种经济有效的方式。这种系统有两个特点：1)一个视频通常在许多参与的主机上可用，2)不同的主机通常有不同的视频集，尽管有些可能部分重叠。从客户的角度来看，它可以由任何拥有其请求的视频的主机提供服务。从服务器的角度来看，它可以用来为任何请求它拥有的视频的客户端提供服务。因此，一个重要的问题是，应该使用哪些服务器来服务系统中的哪些客户机？在本文中，我们将此问题称为服务调度，并表明客户端和服务器之间的不同匹配会导致系统性能的显著差异。为每个客户机找到合适的服务器是一项挑战，这不仅是因为客户机只能选择在其有限搜索范围内的服务器，而且还因为客户机到达的时间不同，这些时间不是先验的。在本文中，我们用一种叫做“摇动”的新技术来解决这些挑战。虽然所建议的技术使客户机可以由超出客户机自身搜索范围的服务器提供服务，但它能够在新请求到达时动态调整服务器与其挂起请求之间的匹配。我们的性能研究表明，我们的新技术可以动态平衡系统工作负载，并显着提高系统的整体性能
Peer-to-peer (P2P) video systems provide a cost-effective way for a large number of hosts to collaborate for video sharing. Two features characterize such a system: 1) a video is usually available on many participating hosts, and 2) different hosts typically have different sets of videos, though some may partially overlap. From a client's perspective, it can be served by any host having the video it requests. From a server's perspective, it be used to serve any client requesting the videos it has. Thus, an important question is, which servers should be used to serve which clients in the system? In this paper, we refer to this problem as service scheduling and show that different matches between clients and servers can result in significantly different system performance. Finding a right server for each client is challenging not only because a client can choose only the servers that are within its limited search scope, but also because clients arrive at different times, which are not known a priori. In this paper, we address these challenges with a novel technique called Shaking. While the proposed technique makes it possible for a client to be served by a server that is beyond the client's own search scope, it is able to dynamically adjust the match between the servers and their pending requests as new requests arrive. Our performance study shows that our new technique can dynamically balance the system workload and significantly improve the overall system performance