CAREER: Priority Users and Applications on the Internet

这项工作旨在为有特色的学术生涯奠定基础，影响学生的生活，并为社会做出持久而有价值的贡献。它寻求使用以下方法来实现这些目标。-本项目建议制定和评估在公共宽带TCP/IP网络(即“因特网”)上支持高优先级用户的机制。对这类用户的支持将增加互联网服务提供商(ISP)的收入机会，并支持对社会具有重要意义的活动，如军事、国家安全和应急行动。-通过为本科生和研究生开发创新的实习和创业培训机会，对学生的生活产生影响。还将开发新的课程和课程，教授从该项目的研究工作中获得的原则和新知识。学生将积极参与这项研究，包括本科生和来自代表性不足群体的学生。-这一学术生涯只有热情地从事超越技术发展的研究，才能取得成功，才能改变人们的生活。这项工作为PI提供了一个工具和机会，以建立一个追求这种激情的学术生涯。值得注意的是，该项目的研究重点将是优先用户，而不是他们产生的流量或他们寻求使用的服务。优先用户可能期望或可能不期望高服务质量(例如，低延迟、高数据速率)或对网络提出高要求的应用(例如，交互式视频)。然而，无论网络的状态或其运行的环境如何(例如，在自然灾害之后)，所有优先用户都将期望网络服务对他们可用并且可靠地执行，而不会失败。国家安全/应急准备用户特别感兴趣，因为他们经常需要在网络因故障和流量过载而最难支持他们时执行其活动。优先用户的主要要求如下。-可用性-在需要时立即访问网络资源。-可信赖性-无论网络或社交环境中发生什么情况，都能够执行和完成通信会话。-稳定性优先的用户必须能够信任他们使用的技术，即使这些技术正在发展。此外，分组级别的服务质量和优惠待遇对一些优先用户可能很重要，但所有优先用户都希望满足上述三个主要要求。这项工作提出了四项项目任务，分五年进行。前三项任务将制定在尽力、面向连接和无连接(即差异化服务)环境中支持高优先级用户的机制。任务4将评估完整端到端路径上多种类型网络的性能预期。这四项工作如下。-任务1-尽力而为环境的终端流量控制-将为最终用户开发机制，以控制他们在尽力而为网络中收到的性能。这是使用转发主机来实现的，转发主机充当流量的中间目的地。多个转发主机的存在为最终用户提供了测试和发送流量的路径选择。-任务2-集成路由和连接准入控制-连接准入控制(CAC)在面向连接的环境中非常有用，可以限制来自低优先级用户的流量。新的CAC功能将通过与动态路由有效集成的模拟实验来开发。-任务3-区分服务器域中的优先级处理-在区分服务域中，流量以聚合、无连接的方式处理。将开发新的流量识别技术、每个域的行为、流量工程指南和服务级别协议模板，以专门解决此类环境中高优先级用户的要求。-任务4--端到端优先级处理--该项目的最终研究任务最终在任务4中完成，在该任务中，优先级用户将能够通过新机制获得端到端的网络活动支持，以了解并影响数据路径上不同类型网络的性能。

This work seeks to build a foundation for an academic career that is distinctive, has an impact on the lives of students, and makes a lasting and valuable contribution to society. It seeks to accomplish these objectives using the following approach. -This project proposes the development and performance evaluation of mechanisms for the support of highpriority users over public broadband TCP/IP networks (i.e., "the Internet"). Support for these types of users will enhance Internet service provider (ISP) revenue opportunities as well as support activities of importance to society, such as military, national security, and emergency response operations. -Impacts will be made on the lives of students through the development of innovative internship andentrepreneurial training opportunities for undergraduate and graduate students. New courses and curricula will also be developed to teach principles and new knowledge gained from this project's research work. Students will actively participate in this research, including undergraduates and those from underrepresented groups. - This academic career will only be successful if it passionately pursues research which goes beyond technology development to make a difference in the lives of people. This work provides the PI a vehicle and opportunity to build an academic career that pursues such a passion. It is important to note that the research focus of this project will be on priority users, more than on the traffic they generate or the services they are seeking to use. Priority users may or may not expect high quality of service (e.g.,low delay, high data rates) or applications that place high demands on the network (e.g., interactive video). Without fail, however, all priority users will expect network services to be available to them and to perform reliably, regardless of the state of the network or the context in which it is operating (e.g., after a natural disaster). National security/emergency preparedness users are of particular interest because they frequently need to perform their activities when it is most difficult for networks to support them due to failures and overloaded traffic.The primary requirements for priority users are as follows. - Availability - Immediate access to network resources whenever needed. -Dependability - Ability to carry out and complete a communication session, no matter what is happening in the network or social context. -Stability - Priority users must be able to trust the technologies they use, even as those technologies evolve. In addition, packet-level quality of service and preferential treatment may be important to some priority users, but all priority users expect the above three primary requirements to be met. This work proposes four project tasks that are staged over five years. The first three tasks will develop mechanisms for supporting high priority users in best effort, connection-oriented, and connectionless (i.e., Differentiated Services) environments. Task 4 will assess performance expectations across multiple types of networks on a full end-to-end path. These four tasks are as follows. -Task 1 - Endpoint Traffic Control for Best Effort Environments - Mechanisms will be developed for end users to control the performance they receive in best effort networks. This is accomplished using forwarding hosts which serve as intermediate destinations for traffic. The presence of multiple forwarding hosts gives end users choices of paths over which to test and send traffic. -Task 2 - Integrated Routing and Connection Admission Control - Connection admission control (CAC) is very useful in connection-oriented environments to limit the amount of traffic from low priority users. New CAC functions will be developed through simulation experiments that integrate efficiently with dynamic routing. -Task 3 - Priority Treatment in Diff-Serv Domains - In Differentiated Services domains, traffic is handled in an aggregated, connectionless manner. New traffic identification techniques, per domain behaviors, traffic engineering guidelines, and service level agreement templates will be developed to specifically address the requirements of high priority users in such environments. -Task 4 - End-to-end Priority Treatment - The ultimate research mission of this project is finally accomplished in Task 4, where priority users will be able to have their network activities supported end-to-end through new mechanisms to understand and influence performance in the different types of networks along a data path.