Collaborative Research (NSF-CNPq): Application Level Adaptation and Control for Retrieval and Delivery of Continuous Media over the Internet

未来的网络化多媒体信息系统将承载各种各样的应用，包括数字图书馆、视频、音频和图像服务、远程学习和协作、网络化虚拟环境和娱乐。导致端到端系统设计困难的多媒体应用程序的主要特征是它们具有非常大的带宽和存储需求，具有非常不同的性能和可靠性需求，通常还伴随着实时限制。伴随着这些特点，对高互可靠性的要求，往往伴随着实时性的限制。除了这些特点之外，各种多媒体应用程序的高度交互性导致了相当不可预测的工作负载，这使得网络化多媒体信息系统的设计和评估成为一个非常具有挑战性的问题。在本提案中，我们概述了该问题的三个方面的研究：(1)设计和评估CM服务器的服务器资源分配算法，以便根据应用程序所需的QoS高效地检索信息；(2)开发用于评估新型服务器设计的性能评估技术。(3)该项目的参与者带来了来自数据库、远程教育、多媒体、网络和绩效评估等广泛领域的专业知识，以解决这些领域的问题。该提案的另一个重要特点是，参与者可以获得四个最先进的多媒体系统的原型：加州大学洛杉矶分校开发的虚拟世界数据服务器（VWDS）；在巴西开发的视频点播服务器的原型，作为巴西机构之间合作研究项目的一部分；多媒体异步网络个性化课件（MANIC）和互联网多媒体代理（IMP）都是在马萨诸塞大学开发的。这些应用程序将用于推动新算法的发展，以便在通过广域网发送数据后检索信息和保持所需的服务质量。它们也将构成许多实验和分析研究的基础，这些研究将用于评估这些新算法。为了帮助进行评估，我们的团队还开发了三种性能评估和建模工具：一种用于构建性能和可靠性模型的最先进工具（Tangram-II），是与加州大学洛杉矶分校在巴西联合开发的；符号模型检查工具（VERUS）：以及促进多媒体存储层次（ViPEr-HiSS）设计的设计、开发和后续性能评估的工具。因此，我们实验室的环境以及它们之间的长距离将为这种类型的评估提供一个独特的测试平台，因为我们的应用程序可用的连接方式截然不同，从千兆低利用率链路到洲际拥塞链路。提出的研究代表了下一代信息服务器和将在其上运行的网络应用程序的设计和性能评估的一个根本性的重要步骤。作为我们研究的结果，我们期望更好地理解存储服务器资源管理策略、通道分配策略和网络适应策略如何相互作用以满足CM应用程序所需的QoS，尽管存在相当不可预测的网络延迟。

Future networked multimedia information systems will carry a wide variety of applications including digital libraries, video, audio and image services, distance learning and collaboration, networked virtual environments, and entertainment. The main characteristics of multimedia applications that lead to difficulties in end-to-end systems design are that they have very large bandwidth and storage requirements, with vastly different performance and reliability requirements, often coupled with real-time constraints. Along with these characteristics, the highly interreliability requirements, often coupled with real-time constraints. Along with these characteristics, the highly interactive nature of a variety of multimedia applications, resulting in fairly unpredictable workloads, makes the design and evaluation of networked multimedia information systems an exceptionally challenging problem. In this proposal, we outline research on three aspects of this problem: (1) the design and evaluation of server resource allocation algorithms for CM servers in order to retrieve information efficiently and according to the QoS demanded by the application; (2) the development of performance evaluation techniques for evaluating new server designs. (3) the participants of this project bring expertise from a wide variety of areas: databases, distance learning, multi-media, networking, and performance evaluation to bear on problems in these areas. Another important feature of this proposal is that the participants have available four prototypes of state of the art multimedia systems: the Virtual World Data Server (VWDS) developed at UCLA: the prototype of a Video-on-Demand server developed in Brazil as part cooperative research project among Brazilian institutions; and the Multimedia Asynchronous Networked Individualized Courseware (MANIC) and the Internet Multimedia Proxy (IMP) both developed at UMass. These applications will be used to motivate the development of new algorithms for retrieving information and for maintaining the desired quality of service after sending the data over a wide area network. They will also form the basis of the many experimental and analytical studies that will be performed to evaluate these new algorithms. To aid in this evaluation, our group also developed three performance evaluation and modeling tools: a state-of-the-art tool for constructing performance and reliability models (Tangram-II) developed in Brazil jointly with UCLA; a symbolic model checking tool (VERUS): and a tool which facilitates design, development, and subsequent performance evaluation of designs of multimedia storage hierarchies (ViPEr-HiSS) under development at UMD. Thus, the environment of our labs as well the long distance among them will provide a unique testbed for this type of an evaluation due to the drastically different connectivities available to our applications, from gigabit low utilized links to intercontinental congested links. The proposed research represents a fundamentally important step in the design and performance evaluation of next generation information servers and the networked applications that will operate on top of them. As a result of our research we expect to have a better understanding of how storage server resource management policies, channel allocation policies, and network adaptation policies interact to satisfy the required QoS of CM applications, despite the fairly unpredictable network delays.