Composing Data-Rich Embedded Systems the Easy Way

组成数据丰富的嵌入式系统的简单方法--除了构建组件来处理这些数据之外，嵌入式系统程序员还必须安排潜在的数百个组件之间的通信，将计算分配给处理元件，以便最小化通信成本并最大化响应能力，并调度处理元件以适应不断变化的优先级和通信模式。这些挑战必须在系统级而不是处理器级来解决，这个项目开发了一个框架，用于组成分布式的、数据丰富的嵌入式系统，它可以自动化许多低级别的进程分配和调度任务。它发生在一个“下一代”人形机器人目前正在耶鲁大学开发的背景下。该机器人包含大量的处理器连接在一个异构的方式。该项目解决了两个基本的研究问题。首先是使用现代编程语言技术来解决关键的嵌入式系统的关注，如可组合性和动态配置更改。第二个是通过在运行时系统中利用高级系统知识来提高系统的整体性能。最终结果将是一种设计方法，能够快速可靠地构建复杂的数据丰富的交互系统。

Composing Data-Rich Embedded Systems the Easy Way-------------------------------------------------Arvind Krishnamurthy and Henrik NilssonDepartment of Computer Science, Yale UniversityEmbedded computing increasingly takes place in a sensor rich environment wherethe acquisition of raw information is much easier than its interpretation. Inaddition to building components to process this data, embedded systemsprogrammers must also arrange the communication among potentially hundreds ofcomponents, distributing computation to the processing elements so as tominimize communication costs and maximize responsiveness, and scheduling ofthe processing elements to adapt to changing priorities and communicationpatterns. These challenges must be addressed at the system level rather thanthe processor level.This project develops a framework for composing distributed, data-richembedded systems that automates many of the low-level process allocation andscheduling tasks. It takes place against a backdrop of a "next generation"humanoid robot currently being developed at Yale. The robot contains asignificant number of processors connected in a heterogeneous fashion. Theproject addresses two fundamental research issues. The first is the use ofmodern programming language techniques to address critical embedded systemconcerns such as composability and dynamic configuration change. The second isimproving overall system performance by exploiting high level system knowledgein the run-time system. The end result will be a design methodology that willenable rapid and reliable construction of complex data-rich interactivesystems.