CPS: Breakthrough: Collaborative Research: A Framework for Extensibility-Driven Design of Cyber-Physical Systems

CPS：突破：协作研究：网络物理系统可扩展性驱动设计的框架

• 批准号: 1646497
• 负责人: Wenchao Li
• 金额: $ 22.5万
• 依托单位: Trustees of Boston University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-09-01 至 2020-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1646497&HistoricalAwards=false
• 关键词: CPS; Breakthrough; Collaborative; Research; Framework

A longstanding problem in the design of cyber-physical systems is the inability and ineffectiveness in coping with software and hardware evolutions over the lifetime of a design or across multiple versions in the same product family. The objective of this project is to develop a systematic framework for designing extensible cyber-physical systems that can enable efficient and correct updates with minimal redesign and re-verification efforts. The intellectual merits are (1) a new and unified framework that optimizes system extensibility by addressing both functional correctness and platform feasibility, and (2) new algorithms for functional verification with platform consideration, software architecture synthesis driven by extensibility metrics, and integration of verification and synthesis for joint design space exploration. The project?s broader significance and importance are (1) enabling engineers to cope with continual changes in cyber-physical design components or operating conditions, thereby significantly reducing redesign and re-verification cost, (2) providing a general framework for designing extensible systems that is applicable to a wide range of systems including robotic, automotive, and avionic systems, and (3) providing new methodologies and techniques that facilitate the training of undergraduate and graduate students to meet the design challenges of cyber-physical systems.Many cyber-physical systems today are one-off designs -- systems designed without future changes in mind. The proposed extensibility-driven design (EDD) framework treats extensibility as a first-class design objective and addresses it with a holistic consideration of functional properties and platform implementation. An EDD design flow provides the following capabilities. At the initial design stage, EDD identifies certain constraints (e.g., timing) that are critical for functional correctness, and explores the design space to maximize the amount of future software and hardware changes that can be made without violating these constraints. During design updates, EDD first determines whether it is possible to accommodate the updates through software architecture re-synthesis, so as to avoid costly re-verification. In the cases where the updates violate existing platform constraints and requirements, EDD selectively modifies some of them to explore feasible changes while minimizing re-verification efforts.

网络物理系统设计中的一个长期存在的问题是，在设计生命周期内或同一产品系列中的多个版本之间，无法应对软件和硬件的演变，而且效率低下。该项目的目标是为设计可扩展的网络物理系统开发一个系统框架，使其能够以最少的重新设计和重新核查努力实现有效和正确的更新。其智能优点是(1)新的统一框架，通过同时解决功能正确性和平台可行性来优化系统的可扩展性；(2)考虑平台的功能验证的新算法，可扩展度量驱动的软件体系结构综合，以及用于联合设计空间探索的验证和综合的集成。S项目更广泛的意义和重要性是：(1)使工程师能够应对网络物理设计部件或运行条件的不断变化，从而显著降低重新设计和重新验证的成本；(2)为设计可扩展的系统提供一个通用框架，适用于包括机器人、汽车和航空电子系统在内的各种系统；(3)提供新的方法和技术，有助于培养本科生和研究生，以应对网络物理系统的设计挑战。目前许多网络物理系统是一次性设计--系统设计时没有考虑未来的变化。拟议的可扩展性驱动设计(EDD)框架将可扩展性视为首要的设计目标，并从整体上考虑功能属性和平台实现。EDD设计流程提供以下功能。在初始设计阶段，EDD确定对功能正确性至关重要的某些约束(例如，时间)，并探索设计空间，以最大化在不违反这些约束的情况下可以进行的未来软件和硬件更改的数量。在设计更新过程中，EDD首先确定是否有可能通过软件体系结构重新综合来适应更新，以避免昂贵的重新验证。在更新违反现有平台约束和要求的情况下，EDD会有选择地修改其中一些更新，以探索可行的更改，同时最大限度地减少重新验证的努力。

项目成果

ReachNN: Reachability Analysis of Neural-Network Controlled Systems

• DOI: 10.1145/3358228
• 发表时间: 2019-10-01
• 期刊: ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS
• 影响因子: 2
• 作者: Huang, Chao;Fan, Jiameng;Zhu, Qi
• 通讯作者: Zhu, Qi

Application-Aware Scheduling of Networked Applications over the Low-Power Wireless Bus

通过低功耗无线总线进行网络应用的应用感知调度

• DOI: 10.23919/date48585.2020.9116206
• 发表时间: 2020
• 期刊: Automation & Test in Europe Conference & Exhibition (DATE
• 作者: Wardega, Kacper;Li, Wenchao
• 通讯作者: Li, Wenchao

ReachNN*: A Tool for Reachability Analysis of Neural-Network Controlled Systems

• DOI: 10.1007/978-3-030-59152-6_30
• 发表时间: 2020-10
• 作者: Jiameng Fan;Chao Huang;Xin Chen;Wenchao Li;Qi Zhu

Divide and Slide: Layer-Wise Refinement for Output Range Analysis of Deep Neural Networks

• DOI: 10.1109/tcad.2020.3013071
• 发表时间: 2020-11
• 期刊: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
• 影响因子: 2.9
• 作者: Chao Huang;Jiameng Fan;Xin Chen;Wenchao Li;Qi Zhu

Design Automation of Cyber-Physical Systems: Challenges, Advances, and Opportunities

• DOI: 10.1109/tcad.2016.2633961
• 发表时间: 2017-09
• 期刊: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
• 影响因子: 2.9
• 作者: S. Seshia;Shiyan Hu;Wenchao Li;Qi Zhu