Run-time and Design-time Exploration of Field-level Energy, Space, Time and Fidelity Trade-offs in Distributed Sensor Networks

由于微电子学的进步，廉价而微小的处理器、无线电、传感器和致动器正在导致一种新的嵌入式系统，通常被称为无线传感器网络，它由大量物理耦合、能量受限、空间分布(通常以自组织方式)和无线联网的独立节点组成。这类系统以前所未有的详细程度提供有关物理环境的信息，并基于这些信息在各种应用中操纵物理环境，如安全和监视、野生动物栖息地监测、智能传感器仪表环境以及复杂系统的基于条件的维护。这项研究涉及系统地设计、优化和管理无线传感器网络的技术，以满足应用需求，例如系统应该持续多长时间，它应该覆盖多大空间，以及它应该以多准确和多快的速度感知环境。无线传感器网络是一种自主、自配置和自适应的分布式系统，它在能量受限的节点之间执行协作计算，以产生关于物理世界的所需信息。该研究正在为这类系统开发设计时资源分配和运行时资源管理方法，同时利用这些系统中作为服务质量概念基础的能量、空间、时间和精度维度之间的相互作用。这项研究对于理解无线传感器网络的基本性能极限，以及开发针对特定应用系统地部署和运行传感器网络的实用方法都是重要的。

Cheap and tiny processors, radios, sensors, and actuators resulting from progress in microelectronics are leading to a new class of embedded systems, often called Wireless Sensor Networks, that consist of a large number of individual nodes that are physically-coupled, energy-constrained, spatially-distributed (often in an ad hoc fashion), and wirelessly-networked. Such systems provide information about the physical environment at an unprecedented level of detail, and to manipulate the physical environment based on this information, in diverse applications such as security and surveillance, monitoring of wildlife habitats, smart sensor-instrumented environments, and condition-based maintenance of complex systems. This research involves the study of techniques to systematically design, optimize, and manage Wireless Sensor Networks so as to meet applications requirements such as how long the system should last, what space should it cover, and how accurately and how rapidly should it sense the environment. Wireless Sensor Networks are autonomous, self-configuring, and adaptive distributed systems that perform collaborative computation among energy-constrained nodes to produce the desired information about the physical world. The study is developing design-time resource allocation and run-time resource management methods for such systems while exploiting the inter-play of energy, space, time, and accuracy dimensions that underlies the notion of quality of service in these systems. The focus is on trade-offs that manifest themselves at the level of the entire "sensor field' as opposed to the individual sensor nodes. This study is important both for understanding the fundamental performance limits of wireless sensor networks, as well as for developing practical methods to systematically deploy and operate sensor networks for specific applications.