Traffic Control for Networks with Aggregate Quality-of-Service

随着Internet不断扩展到新的应用领域，对提供不同程度的服务质量（QoS）的差异化服务的需求不断增长。直到最近，互联网中服务差异化的方法一直集中在为单个流量提供QoS保证上。这种按流模型还没有被广泛接受，很大程度上是因为需要在网络中维护大量的状态信息。因此，网络社区目前正在为Internet重新定义QoS的概念，从而产生一种新的服务模型，其中对适当的流而不是单个流进行保证。例如，由IETF的差异化服务（DiffServ）工作组定义的聚合QoS的概念与传统的网络概念非常不同，需要新颖的算法来进行流量控制和供应。拟议研究的目的是对流量聚合的QoS的性质和交付提供基本见解，因此称为聚合QoS。聚合QoS服务模型的一个核心特征是不需要用户指定流量的组成或目的地。这大大增加了网络资源供应的复杂性，并建议必须开发新的反馈密集型流量控制算法，以减轻网络负载的不确定性。目前，很少或没有理论可以指导这种算法的发展。拟议的研究将通过开发分析工具来纠正这种情况，这些工具将使下一代互联网的聚合QoS成为可能。本方案的三个主要研究部分包括分析技术的发展和计算模拟实验。该项目建议开发一种新的流体流网络优化模型，该模型将促进(a)对聚合QoS网络的快速分析和(b)基于网络流量模式反馈的改进路由。该项目将使用一种所谓有效包络的新结构，开发一种在具有聚合QoS的网络中计算统计多路复用增益的方法。了解统计多路复用增益对于确定网络中可支持的具有聚合QoS的通信量至关重要。该项目将建立一个分组级模拟环境，这将使开发和测试用于聚合QoS网络的各种流量控制算法成为可能。这项研究将有助于确定比较流量控制算法的效果的指标，以支持聚合QoS。我们工作的一个重要部分是确定基于反馈的流量控制能在多大程度上缓解供应问题。通过为聚合QoS提供一个理论参考框架，它提供了容量配置和准入控制的工具和技术，提议的研究将补充目前正在进行的针对互联网上差异化服务的标准化工作。

As the Internet continues to expand into new application domains, there is a growing demand for differentiated services that provide varying degrees of Quality of Service (QoS). Until recently, the approach for service differentiation in the Internet has focused on providing QoS guarantees to individual traffic flows. This per-flow model has not been widely embraced, largely due to the vast amount of state information that needs to be maintained in the network. As a consequence, the network community is currently redefining the notion of QoS for the Internet, leading to a new service model where guarantees are made to appropriate flows, rather than to individual flows.The notions of aggregate QoS which are being defined, e.g., by the Differentiated Services (DiffServ) working group in the IETF, are very different from traditional networking concepts and require novel algorithms for traffic control and provisioning. The objective of the proposed research is to give fundamental insights into the nature and delivery of QoS to traffic aggregates, henceforth called aggregate QoS.A central feature of service models for aggregate QoS is that there is no requirement for users to specify the composition or the destination of traffic. This significantly complicates the provisioning of network resources and suggests that new, feedback-intensive traffic control algorithms must be developed to mitigate uncertainty about network loading. At present there is little or no theory available to guide the development of such algorithms. The proposed research will remedy this situation by developing analytical tools that will make aggregate QoS possible for the next-generation Internet.The three main research components of this proposal involve development of analysis techniques and computational simulation experimentation.1. The projectl proposes to develop a novel fluid-flow, network-optimization model that will facilitate (a) rapid analysis of aggregate QoS networks and (b) improved routing, based on feedback about network traffic patterns.2. Using a new construct of so-called effective envelopes, the project will develop a method for computing the statistical multiplexing gain in a network with aggregate QoS. Knowledge of the statistical multiplexing gain is essential for determining the amount of traffic with aggregate QoS that can be supported in a network.3. The project will build a packet-level simulation environment, which will enable the development and testing of a wide variety of traffic control algorithms for aggregate QoS networks. This research will help to identify metrics for comparing the effect of traffic control algorithms to support aggregate QoS. An important part of our work is to determine the extent to which feedback-base traffic control can alleviate the provisioning problem.By providing a theoretical reference frame for aggregate QoS, which provides tools and techniques for capacity provisioning and admission control, the proposed research will complement the currently ongoing standardization efforts for differentiated services on the Internet.