Programming Language and Middleware Support for Sensor Network Applications

摘要：传感器网络在多个领域得到了应用。它们实现了物理世界与计算系统基础设施的紧密集成，从而实现了更好的信息感知、访问和控制。然而，传感器网络应用的开发一直是极其困难的：几乎没有开发支持。这项提议旨在开发编程语言、编译器和中间件服务，使其更容易实现、部署和管理传感器网络应用程序。编程语言将为定义传感器节点和节点组、节点和组行为以及节点和组之间的交互提供新的抽象。节点行为和群行为完全分离，支持可重用性。这些语言还将为组级聚合、状态管理、组长选择以及一对多和多对多方法调用提供内在支持。其目标是捕获节点和组实例化、数据共享和通信的常见模式，以便轻松开发应用程序。该中间件将为组管理、节点间通信、组节点交互和代码分发实现可扩展的基础设施。研究还将探索语言抽象和中间件服务的性能特征。拟议的研究将对若干领域的应用开发产生重大影响，包括环境科学、灾后恢复、民用基础设施监测、教育、精准农业和普及计算。研究计划还将教育和研究活动结合起来，通过课程、开源软件开发项目、项目让本科生和研究生参与进来

Abstract: Sensor networks are finding applications in several areas. They enable tight integration of the physical world with a computing system infrastructure, thereby permitting better information sensing, access and control. However, development of sensor network applications has been extremely difficult: there is very little or no development support. This proposal is aimed at developing programming languages, compilers and middleware services that will make it easier to implement, deploy and manage sensor network applications. The programming languages will provide novel abstractions for defining sensor nodes and groups of nodes, node and group behaviors, and interaction among nodes and groups. The node and group behaviors are completely separated, which supports reusability. The languages will also provide intrinsic support for group level aggregation, state management, group leader selection, and one-many and many-many method invocations. The objective is to capture common patterns of node and group instantiations, data sharing, and communication so that applications can be developed easily. The middleware will implement a scalable infrastructure for group management, inter-node communication, group-node interactions and code distribution. The research will also explore the performance characteristics of the language abstractions and middleware services. The proposed research will have significant impact on development of applications in several areas, including environmental sciences, disaster recovery, civilian infrastructure monitoring, education, precision agriculture and pervasive computing. The research plan also integrates educational and research activities by involving undergraduate and graduate students through courses, open source software development projects, project