Dependable Distributed Real-Time

满足可靠的分布式实时系统的要求是相当大的挑战。许多要建造的系统既要求是硬实时的，也要求使用现成的商业硬件和软件组件。使用现成组件的需要从根本上与经典的“保证响应”硬实时设计方法相冲突。拟议研究的目的是开发适用于由现成组件构建的硬实时分布式系统的新服务、算法和系统设计方法。第二个目标是将这些新的服务、算法和设计方法体现在一个原型分布式容错应用开发环境中，该环境将允许硬实时服务和软实时服务和谐共存。第三个目标是调查确保此类系统和服务免受恶意攻击的方法。最初，该项目的重点将是继续探索故障感知设计范例，作为“有保证的响应”硬实时设计方法和“尽最大努力”软实时范例之间的中间方法。更广泛地说，预计还将考虑指定、设计、保护和验证可靠的分布式实时系统的其他方面。

Meeting the requirements of dependable distributed real-time systems poses considerable challenges. Many of the systems to be built are required to be both hard real-time and to use off-the- shelf commercial hardware and software components. The need for using off-the-shelf components conflicts fundamentally with the classical ``guaranteed response'' hard real-time design methodology. The aim of the proposed research is to develop novel services, algorithms and system design methodologies that are applicable to the design of hard real-time distributed systems built from off-the-shelf components. A second goal is to embody these novel services, algorithms and design methodologies into a prototype distributed fault-tolerant application development environment that would allow hard-real time services and soft-real time services to co-exist harmoniously. A third goal is to investigate means for making such systems and services secure to adverse attacks. Initially the project will focus on continuing an on-going exploration of the fail-aware design paradigm, as a middle of the road approach between the ``guaranteed response'' hard real-time design methodology and the ``best effort'' soft real time paradigm. More broadly, it is expected that other aspects of specifying, designing, securing and validating dependable distributed real-time systems will be considered.