Efficient Multicast Communication in Switch-Based Networks for Quality-of-Service

建议：0073085标题：基于交换机的网络中用于服务质量的高效组播通信PI：源源杨组播涉及从一个源到多个目的地的信息传输，是高性能网络的要求。预计多播将越来越多地用于支持各种交互应用，如视频点播、电话会议、网络服务器和因特网上的电子商务。这些应用中的许多不仅需要组播功能，还需要可预测的性能，例如有保证的组播延迟和带宽，称为服务质量(Qos)。本研究的目的是在组播网络的基础上，设计一种支持组播通信的具有服务质量的组播体系结构，支持组播通信。具体研究内容包括：(1)设计具有更好的服务质量(Qos)功能和较低网络成本的基本积木式组播网络；(2)建立具有统计Qos能力的组播体系结构的性能指标解析模型；(3)研究不同服务质量要求下的高效组播调度方案；(4)构建用于验证理论结果的仿真工具，并将其提供给研究团体和系统设计者。该研究对可扩展的高性能并行/分布式计算具有重要影响，因为对多播通信的有效支持可以显著降低通信延迟，简化大量并行/分布式应用的设计和实现。

Proposal: 0073085Title: Efficient Multicast Communication in Switch-Based Networks for Quality of ServicePI: Yuanyuan YangMulticast involves transmitting information from a source to multiple destinations, and is a requirement in high-performance networks. It is projected that multicast will be increasingly used to support various interactive applications such as video-on-demand, teleconferencing, web servers and E-commerce on the Internet. Many of these applications require not only multicast capability but also predictable performance, such as guaranteed multicast latency and bandwidth, called quality-of-service (QoS). The objective of this research is to design a QoS capable multicast architecture based on multicast networks with built-in QoS functions to support multicast communication. Specially, the research focuses on:(1) design of basic building-block multicast networks with better QoS functions and lower network cost;(2) establishing analytical models on performance metrics of statistical QoS-capable multicast architectures;(3) study of efficient schemes to schedule multicast for different QoS requirements;(4) building simulation tools for validating theoretical results, and making them available to research community and system designers.This research combines algorithmic, probabilistic, combinatorial, and simulation techniques to conduct comprehensive studies. The proposed research has a significant impact on scalable and high-performance parallel/distributed computing as efficient support for multicast communication can significantly reduce the communication latency and simplify the design and implementation of numerous parallel/distributed applications.