CRII: CNS: Supporting Mixed-Criticality Real-Time Systems on Heterogeneous Platforms

随着现代计算系统中应用程序的蓬勃发展，为了可伸缩性和可持续性，以统一的方式设计和实现这样的系统是一种趋势。这种趋势导致：首先，在一个共同的共享硬件平台上支持不同关键程度的应用程序或任务；其次，这样的平台由具有不同功能和能力的计算处理器或异构处理器组成。这些可能会严重影响系统的时间可预测性，这对于必须在运行前验证时间保证的实时系统至关重要。本项目将对异构处理器上混合临界任务的实时调度进行探索性研究。实时调度分析将为系统的时序行为提供可证明的保证。本研究将探讨解决任务和处理器两方面整合的整体方法和技术。这项工作将产生以集成方式表示MC任务和异构处理器的新系统模型。在此基础上，开发新的调度算法，在满足MC任务所需时间保证的前提下，有效地分配计算资源。系统原型将被实施，以演示和评估模型和算法的有效性。这些新的模型、算法和原型将促进对实时系统中MC和异构调度的理解，并将实现统一的实时系统设计。这进一步提高了物联网（IoT）和网络物理系统（CPS）中广泛应用的效率、可扩展性和可负担性，这些应用通常存在时间限制。由于是在一个大型的西班牙裔服务机构（HSI）进行的，该项目将促进科学、技术、工程和数学（STEM）教育和研究，面向来自代表性不足群体的学生，并扩大他们在计算机领域的参与。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

With the thriving applications in modern computing systems, there is a trend toward designing and implementing such systems in a consolidated manner for both scalability and sustainability. This trend leads to: first, applications, or tasks, of different criticalities that are supported on a common shared hardware platform; second, such platforms consist of computing processors that have different functionalities and capabilities, or heterogeneous processors. These may significantly impact the temporal predictability of the system, which is crucial for real-time systems where temporal guarantees must be validated prior to runtime. This project will conduct exploratory research on the real-time scheduling of mixed-criticality (MC) tasks on heterogeneous processors. The real-time scheduling analysis will provide provable guarantees on the temporal behaviors of the system. This research will investigate holistic methods and techniques that address the consolidation in both aspects of tasks and processors. This work will result in new system models that represent the MC tasks and the heterogeneous processors in an integrated manner. Based on the models, new scheduling algorithms will be developed to efficiently allocate the computing resources while the required temporal guarantees for the MC tasks are satisfied. System prototypes will be implemented to demonstrate and evaluate the effectiveness of the models and algorithms. These new models, algorithms, and prototypes will advance the understanding of both MC and heterogeneous scheduling in real-time systems, and will enable consolidated real-time system designs. This further benefits the efficiency, scalability, and affordability of a wide range of applications in internet of things (IoT) and cyber-¬physical systems (CPS) where temporal constraints commonly exist. As being conducted in a large Hispanic Serving Institution (HSI), this project will promote science, technology, engineering, and math (STEM) education and research for students from underrepresented groups and broaden their participation in computing.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.