NeTs: An open architecture for the evolutionary design of routing protocols

NeTs：用于路由协议演进设计的开放架构

• 批准号: 0916873
• 负责人: Yaling Yang
• 金额: $ 15万
• 依托单位: Virginia Polytechnic Institute and State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-10-01 至 2011-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0916873&HistoricalAwards=false
• 关键词: NeTs; open; architecture; evolutionary; design

Traditional routing systems have been designed by network engineers based on complex collections of objectives, policies, principles and past experiences. Due to limited human experience and capability, this manual design process severely impedes the design process of routing. The objective of this project is to bring a revolutionary change to this design process by building a highly flexible architecture, called Orchestra, for the automatic assembling and testing of a great variety of routing designs. Orchestra stores a large set of reusable "genes". Each gene is a small piece of computer code that implements a particular design for a small component of a routing system. The correctness of the "genes" and their mutual compatibility are automatically verified. Orchestra assembles various routing systems from verified "genes" and then tests them in both simulation and real environment. Based on the performance of the assembled protocols, Orchestra uses evolutionary algorithms to switch and tune designs of routing components to eventually identify the best design for a network setting.Orchestra will greatly ease a network engineer's burden of implementing and evaluating an entire routing system. It can efficiently explore a much larger design space for routing systems than any single network engineer can. New areas for routing designs that are not explored by humans can be automatically discovered by Orchestra. The large collection of component designs in Orchestra will also provide a common platform for comparing and evaluating different design choices as well as serving education purposes.

传统的路由系统是由网络工程师根据复杂的目标、策略、原则和过去的经验集合来设计的。由于人的经验和能力有限，这种手工设计过程严重阻碍了布线的设计过程。 该项目的目标是通过构建一个高度灵活的架构（称为 Orchestra）来自动组装和测试各种路由设计，从而给设计流程带来革命性的变化。乐团存储了大量可重复使用的“基因”。每个基因都是一小段计算机代码，为路由系统的一个小组件实现特定的设计。 “基因”的正确性及其相互兼容性会被自动验证。 Orchestra从经过验证的“基因”组装各种路由系统，然后在模拟和真实环境中对其进行测试。 根据组装协议的性能，Orchestra 使用进化算法来切换和调整路由组件的设计，最终确定网络设置的最佳设计。Orchestra 将大大减轻网络工程师实施和评估整个路由系统的负担。它可以有效地探索比任何单个网络工程师更大的路由系统设计空间。 Orchestra 可以自动发现人类未探索过的新路由设计领域。 Orchestra 中的大量组件设计还将提供一个通用平台，用于比较和评估不同的设计选择以及服务于教育目的。