Collaborative Research: CSR-EHS: Modeling and Exploiting Cross-Layer Timing in Distributed Embedded Systems

合作研究：CSR-EHS：分布式嵌入式系统中的跨层时序建模和利用

• 批准号: 0615438
• 负责人: Nalini Venkatasubramanian
• 金额: $ 12.5万
• 依托单位: University of California-Irvine
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2006
• 资助国家: 美国
• 起止时间: 2006-07-01 至 2008-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0615438&HistoricalAwards=false
• 关键词: Collaborative; Research; CSR; EHS; Modeling

Embedded and hybrid system (EHS) applications are highly networked systems with end-to-end requirements that involve interactions among multiple layers (application, middleware, network, OS, architecture) in a distributed environment. A holistic approach to understanding timing in these distributed cross layer systems is critical since:EHS applications often face end-to-end hard or soft real-time needs; timing granularities for cross-layer systems can vary by orders of magnitude posing a challenge for timing analysis that can account for these variations; and knowledge of timing parameters at the different levels can dramatically improve the utility and performance of EHS applications that often execute in constrained environments where CPU, memory, network bandwidth and device energy is limited.This project explores the notion of cross-layer timing in highly distributed embedded systems using a blend of formal methods and experimental systems, bringing together researchers with expertise in embedded computing, distributed systems and formal methods. The project will (1) develop novel formal methodologies for modeling and reasoning about cross-layer timing properties in distributed embedded systems, and (2) design mechanisms/policies that will cost-effectively address the QoS/performance tradeoffs based on the cross-layer timing analyses. The results will be validated and tested in the context of mobile multimedia applications that execute in highly dynamic environments and present interesting opportunities for tradeoff analysis and enforcement. A comprehensive solution to the timing issue will enable the wider applicability and adoption of distributed embedded applications and lay the groundwork for the unified treatment of other non-functional constraints across multiple abstraction levels.

嵌入式和混合系统(EHS)应用是具有端到端需求的高度网络化的系统，涉及分布式环境中多层(应用、中间件、网络、操作系统、体系结构)之间的交互。理解这些分布式跨层系统中的时序的整体方法是至关重要的，因为：EHS应用程序通常面临端到端的硬实时或软实时需求；跨层系统的时序粒度可能因数量级而异，这给能够解释这些变化的时序分析带来了挑战；了解不同级别的计时参数可以极大地提高EHS应用程序的实用性和性能，这些应用程序通常在CPU、内存、网络带宽和设备能量有限的受限环境中执行。该项目使用正式方法和实验系统的混合，探索高度分布式嵌入式系统中的跨层计时概念，将具有嵌入式计算、分布式系统和正式方法专业知识的研究人员聚集在一起。该项目将(1)开发新的形式化方法，用于对分布式嵌入式系统中的跨层时序属性进行建模和推理，以及(2)设计基于跨层时序分析的机制/策略，这些机制/策略将有效地解决服务质量/性能之间的权衡。结果将在移动多媒体应用程序的背景下进行验证和测试，这些应用程序在高度动态的环境中执行，并为权衡分析和实施提供了有趣的机会。时间问题的全面解决方案将使分布式嵌入式应用程序得到更广泛的应用和采用，并为跨多个抽象级别统一处理其他非功能约束奠定基础。