On Providing Quality-of-Service Control for Core-Based Multicast Routing

对于嵌入式实时应用或多媒体集成的分布式控制，对具有端到端服务质量保证的高度可预测、及时和可靠的通信服务的需求日益增长。具体地说，IETF的RSVP工作组已经开发了资源预留协议(RSVP)协议，以提供用于单播/多播数据流的接收器发起的资源预留设置。RSVP支持接收方发起的、固定/共享的预留类型，前提是底层路由协议可用来为单播路由/多播树提供足够的资源以维持足够的服务质量。遗憾的是，目前运行的大多数多播路由协议并不明确地支持服务质量(以端到端延迟界限、端到端目的地间延迟抖动界限和最小保证带宽的形式)。因此，所提出的研究的目标是开发和评估Qos路由框架及其相关的接纳控制、成员加入/离开过程和信息更新过程，以允许在基于核心的多播路由协议中部署Qos路由能力。所提出的研究是两个协同组成部分的结合：在一个定义良好的分析框架中为所有组件方案建立和开发理论基础，并通过软件系统建设和实验对其进行验证。我们将通过(I)设计有效的接纳测试来验证组播组成员是否可以加入组播树，同时不违反对其他树上成员的现有的服务质量保证；(Ii)研究在RSVP中提出的具有各种预留类型的受控负载服务和保证的服务的上下文中的接纳测试所需的最小信息/状态集；(Iii)开发(考虑到可伸缩性问题)可读的有效的信息更新和状态刷新过程，其可与大多数利用软状态概念的组播路由协议中存在的软状态刷新机制相结合；以及(Iv)使用Merit Network，Inc.发布的门控多播源代码构建实验软件，定义并实现与RSVP的丰富的、定义良好的接口(这可能需要对RSVP的路由接口RSRR进行微小的改变)，并开发应用层库以利用所提出的服务来提供具有不同级别的Qos的一套多播服务。有关更多信息，请参阅项目网页(http://eewww.eng.ohio-state.edu/drcl/grants/qosr98/).关于为基于核心的多播路由提供服务质量控制。转换的附件：C：\Eudora\Attach\form1239.ps转换的附件：C：\Eudora\Attach\form1239.p

There has been an increasing need of highly predictable, timely, and dependable communication services with QoS guarantees on an end-to-end basis either for embedded real-time applications or for multimedia-integrated distributed control. In particular, the RSVP Working Group of the IETF has developed a resource ReSerVation Protocol (RSVP) protocol to provide receiver-initiated setup of resource reservations for unicast/multicast data flows. RSVP supports receiver-initiated, fixed/shared reservation styles under the assumption that a underlying routing protocol is available to provide unicast routes/multicast trees with sufficient resources to maintain adequate QoS. Unfortunately, most of the currently operational multicast routing protocols do not explicitly support QoS (in the form of end-to-end delay bound, end-to-end inter-destination delay jitter bound, and minimum guaranteed bandwidth). The goal of the proposed research is thus to develop and evaluate a QoS routing framework and its associated admission control, member join/leave procedures, and information update procedures, to allow deployment of QoS routing capabilities in core-based multicast routing protocols. The proposed research is a combination of two synergistic components: establishment of theoretical base for, and development of, all the component schemes in a well-defined analytic framework and their validation with software system building and experiments. We will pursue this research by (i) devising effective admission tests to verify whether or not a multicast group member may join the multicast tree, while not violating existing QoS guarantees to other on-tree members; (ii) investigating the minimum set of information/state required for admission tests in the context of the controlled-load service and the guaranteed QoS service with various reservation styles as proposed in RSVP; (iii) developing (with consideration of the scalability issue) effective information update and state refresh procedures that can be read ily integrated with the soft state refresh echanism that exist in most multicast routing protocols that exploit the soft state concept; and (iv) building an experimental software using the GateD multicast source codes distributed by Merit Network, Inc., defining and implementing a rich, well-defined interface with RSVP (which may require minor changes in RSRR, the routing interface for RSVP), and developing application level libraries to exploit the proposed services to provide a suite of multicast services with different levels of QoS. For more information, please refer to the project web page (http://eewww.eng.ohio-state.edu/drcl/grants/qosr98/). On providing quality-of-service control for core-based multicast routing. Attachment Converted: C:\EUDORA\ATTACH\form1239.ps Attachment Converted: C:\EUDORA\ATTACH\form1239.p