NeTS: Small: Network Function Virtualization Using Dynamic Reconfiguration

在过去的十年中，为了满足网络用户的需求，已经使用了广泛的网络功能集合。各种各样的网络功能，包括防火墙、安全过滤器和服务质量增强器，已经使用越来越复杂的技术实现。这些功能用于在数据和多媒体网络环境中支持可预测的高性能。为了提高灵活性，最近人们对用可编程网络功能取代专用网络硬件很感兴趣。这种方法称为网络功能虚拟化（NFV），它允许在安全性和性能需求发生变化时快速适应网络服务。NFV允许网络管理员根据需要使用适当的可编程网络功能来支持不同的网络策略。在该项目中，将开发一种新的基于硬件的NFV构建方法，提供可扩展性和可编程性，同时支持重要的硬件级并行性。该计算平台将使用现场可编程门阵列（FPGA）来实现许多NFV操作，这些操作可以根据特定的网络流需求进行定制。随着所需功能的数量及其特性的变化，FPGA中的硬件可以重新配置以支持更新的要求。所部署的分类器和流量管理功能的管理将由在附带的微处理器上执行的资源管理器执行。研究人员将创建一系列软件工具和硬件模块来评估该方法，架构资源将通过一种新的资源分配算法进行管理，该算法将允许在NFV性能限制下有效利用可用的FPGA区域。该算法将定期运行，以允许FPGA资源的动态再平衡。更广泛的影响：将制定两个具体方案来扩大这项工作的影响。这些努力将包括一个新的网络功能虚拟化本科课程，侧重于可扩展的现实世界系统，以及一个学生运行的研究研讨会，让学生展示他们的工作。将鼓励FPGA和网络公司在新产品开发中使用该技术。

Over the past decade, a wide-ranging collection of network functions have been used to accommodate the needs of network users. A diverse set of network functions, including firewalls, security filters, and quality of service enhancers have been implemented using increasingly complex techniques. These functions are used to support predictable, high performance in data and multimedia network environments. To promote flexibility, there has been recent interest in replacing dedicated network hardware with programmable network functions. This approach, called network function virtualization (NFV), allows for the rapid adaptation of network services as security and performance requirements change. NFV allows network administrators to support different network policies by using appropriate programmable network functions on demand.In this project, a new hardware-based approach to NFV construction that provides scalability and programmability, while supporting significant hardware-level parallelism, will be developed. The computing platform will use a field-programmable gate array (FPGA) to implement numerous NFV operations that can be customized to specific network flow needs. As the number of required functions and their characteristics change, the hardware in the FPGA can be reconfigured to support the updated requirements. The management of the deployed classifiers and traffic management functions will be performed by a resource manager which is executed on an accompanying microprocessor. The researchers will create a series of software tools and hardware modules to evaluate the approach, Architectural resources will be managed by a new resource allocation algorithm that will allow for the effective use of the available FPGA area under NFV performance constraints. This algorithm will run periodically to allow for dynamic rebalancing of FPGA resources. Broader Impact: Two specific programs to broaden the impact of the work will be developed. These efforts will include a new undergraduate curriculum in network function virtualization focused on scalable real-world systems and a student-run research workshop which will allow students to present their work. FPGA and networking companies will be encouraged to use this technology in new product development.