SENSORS: Hourglass: An Infrastructure for Sensor Network

该提案将解决以下研究问题：如何构建能够有效运行数十年或数百年的传感器；如何在有限的功率预算下构建和管理这些传感器，以及如何构建能够从部署的传感器网络收集数据并将数据有效地路由到需要它们的应用程序的基础设施；如何构建一个能够支持我们今天甚至无法想象的应用程序的系统，这些系统的业务价值是什么，以及它们如何从根本上改变可以提供的服务。在这个项目下开发的拟议系统有三个整体组成部分：传感器网络本身，我们称之为数据收集网络的基础设施，以及开发实际应用所需的工具。在硬件层面，将开发低功耗、长寿命设备的结构，并通过利用电路、架构和软件等设计层次的各个层面来实现节能目标。在基础设施层面。沙漏是一种在传感器和应用程序之间动态指导数据和控制的架构，其设计足够灵活，能够适应不可预见的应用，并能够应对嵌入式基础设施的传感器网络，这些传感器网络可以适应技术的快速发展，同时仍然保留与传统传感器网络交互的能力，而传统传感器网络由于后勤原因无法改变。在应用层，能够确保要构建的系统能够解决业务、医学和环境监测中的实际问题的实际问题。这三个层通过有线和无线通信链路连接，我们建议解决与以下相关的基本问题：1)开发一种具有嵌入式优先MAC的节能自适应通信协议，用于节点内通信和控制；2)开发超视界（OTH），这是一种传感器网络与外界之间的远程通信协议，用于在诸如紧急情况下支持网络基础设施故障的情况下。

This proposal will address research issues such as: how to construct sensors that can usefully function for tens or hundreds of years; how to construct and manage such sensors on a limited power budget, and how to construct an infrastructure that can collect data from deployed networks of sensors and route that data efficiently to the applications that need them; how to build a system capable of supporting applications that we cannot even imagine today, and what is the business value of these systems and how can they fundamentally transform the services that can be provided.The proposed system to be developed under this project has three overall components: the sensor networks themselves, the infrastructure that we call a data collection network, and the tools necessary to enable development of real applications. At the hardware level, construction of low-power, long-lived devices will be developed and energy-efficiency goals will be met by leveraging all levels of the design hierarchy: from the circuits, architecture, and software. At the infrastructure level. Hourglass is an architecture for dynamically directing data and control between sensors and applications, and is designed to be flexible enough to adapt to unforeseen applications and to cope with sensor networks that are embedded infrastructures designed to last many years, by enabling adaptivity to rapid advances in technology while still retaining the ability to interact with legacy sensor networks, which cannot be changed for logistical reasons. At the application layer, real problems who can ensure that the systems to be build solve real problems in business, medicine, and environmental monitoring. These three layers are connected via both wired and wireless communication links and we propose to address fundamental questions related to: 1) the development of a power efficient adaptive communication protocol with embedded prioritized MAC for intra-node communications and control and 2) the development of Over The Horizon (OTH), a long distance communication protocol between the sensor network and the outside world in case of support network infrastructure failure in cases such as in the event of an emergency.