ABSTRACT
Distributed sensor networks (DSNs) are composed of numerous small, low-cost, randomly located nodes. Smart Dust is a term coined at the University of California, Berkeley, to describe massively DSNs consisting of cubic-millimeter-sized motes. The small size and anticipated low cost of the motes will help to collect information cost-effectively and less intrusively. Each mote depends on low-capacity batteries as energy sources. The mote's sensors vary from application to application: temperature, light, magnetic field, vibration, and acoustic. Sensors convert physical variables into electrical signals. Motes incorporate a processor to carry out computations locally. The synchronous processor's clock distribution network is characterized by significant power consumption. Smart Dust motes may have analog subsystems that are influenced by the electromagnetic radiation. Asynchronous designs promise to overcome the clock-related problems. Using variable frequency processors, power consumption can be gradually controlled by scaling the clock frequency.
