Power Aware Remote Information System (PARIS)

摘要

A sensor network is characterized as a set of locally powered sensors linked together via wireless communication with the objective to accurately assess activity within the sensor network field of regard and forward that assessment to command and control elements for appropriate action. To minimize power utilization within that network, each power consuming element must be evaluated for effectiveness to the overall network objective. An assessment of the overall sensor network power optimization was performed in a realistic framework based upon the exploitation of orthogonal sensing elements, environmental situation characterization, and invocation of adaptive processing algorithms in a power aware computing and communicating structure. Issues such as the collective fixed and variable power costs associated with localized versus centralized processing, or the issue of variable costs associated with memory access and storage costs were explicitly evaluated in order to assess the system power efficiency tradeoffs. Robust processing algorithms were included in an adaptive processing suite that selects the optimum algorithm based on available power, target type, image quality, and track history. The architecture developed supports voltage and frequency scaling, sleep modes and power switching that allows the Reduced Instruction Set Computer (RISC) processor to be turned off when not needed.