Lucid Dreaming: Reliable Analog Event Detection forEnergy-Constrained Applications
Sasha Jevtic† Mat Kotowsky‡ Robert P. Dick† Peter A. Dinda† Charles Dowding‡†(sjevtic, dickrp, pdinda)@eecs.northwestern.edu ‡(kotowsky, c-dowding)@northwestern.eduDept. of Electrical Engg. and Computer Science Dept. of Civil and Environmental EngineeringNorthwestern University Northwestern University3 August 2006
Existing sensor network architectures are based on the as-sumption that data will be polled. Therefore, they arenot adequate for long-term battery-powered use in applica-tions that must sense or react to events that occur at un-predictable times. In response, and motivated by a struc-tural autonomous crack monitoring (ACM) application fromcivil engineering that requires bursts of high resolution sam-pling in response to aperiodic vibrations in buildings andbridges, we have designed, implemented, and evaluated
, a hardware/software technique to dramati-cally decrease sensor node power consumption in this andother related event-driven sensing applications. Our hard-ware is an add-on board for standard Crossbow Motes thatmakes use of an ultra-low-power analog comparator and anin-system programmable precision voltage reference. Thesensor, e.g., geophone,output voltageis comparedto the ref-erence. When it exceeds the reference, an interrupt is deliv-ered to the Mote, activating it and triggering high-resolutionsampling. In the structural integrity monitoring application,this is achieved with 1/245 the average power consumptionrequired by existing sensor network architectures, therebydramatically increasing battery lifespan. We believe that theproposedtechniquewillyieldsimilarbeneﬁtsinawiderangeof applications.
Wireless sensor networks have the potential to serve as plat-formsforawiderangeofenvironmentalmonitoringandcon-trol applications. Applications can be considered at manylevels, from the individual sensors, to the individual nodehardware and software, to the local wireless network formedby nodes, and ﬁnally to that network’s interaction with thebroader world. Our work focuses on interaction among sen-sors, microcontrollers, and software within individual wire-less sensor network nodes.In this context, two universal research problems come tothe fore: the maintenance problem and the unpredictableevent problem. How can we arrange for nodes to operatewithout frequent intervention? Low maintenance is neces-sary to allow large-scale deployments in remote environ-ments. It is prevented by short battery life, hence we focuson increasing battery life. How can we arrange for nodesto react to environmental events that occur at unpredictabletimes? We cannot assume that interesting data will be pre-sented, on a silver platter, whenever requested. Jointly ad-dressing the maintenance and unpredictable event problemsrequires changes to the sensor network node architecture, al-lowing it to respond to events at any time while maintain-ing ultra-low power consumption. We claim that addressingthe problem requires a combined hardware and software ap-proach. As described in Sections 2 and 5, attempts to solvethese problems with software, alone, have resulted in highpower consumption or missed events.This work is motivated by applications that have the fol-lowing characteristics:1. They are extremely power-sensitive. The nodes arepowered by batteries that can be replaced only aftermonths or years of operation.2. Low-power sensors and computational elements can beused for detecting, but not necessarily taking detailedmeasurements of, events.3. Eventsarerareandthe computationand/orcommunica-tion they trigger is short relative to the event interarrivaltime.4. Event interarrival times are unpredictable.5. It is preferable not to miss, or ignore, events.1