IntroductionProblemRate-Adaptive Positioning SystemOverviewMovementSpace-Time HistoryCelltower-RSS BlacklistingBluetooth-based Position SynchronizationDiscussionEvaluationQuantifying the Benefits of RAPS and its ComponentsComparing RAPS to Periodic GPS ActivationIntegration with a WiFi Positioning SystemAre GPS errors Pervasive?Summary and Future WorkRelated WorkConclusionsReferencesEnergy-Efficient Rate-Adaptive GPS-based Positioningfor Smartphones∗Jeongyeup Paek Joongheon Kim Ramesh GovindanEmbedded Networks LaboratoryComputer Science DepartmentUniversity of Southern California{jpaek,joonghek,ramesh}@usc.eduABSTRACTMany emerging smartphone applications require position informa-tion to provide location-based or context-aware services. In theseapplications, GPS is often preferred over its alternatives such asGSM/WiFi based positioning systems because it is known to bemore accurate. However, GPS is extremely power hungry. Hencea common approach is to periodically duty-cycle GPS. However,GPS duty-cycling trades-off positioning accuracy for lower energy.A key requirement for such applications, then, is a positioningsystem that provides accurate position information while spendingminimal energy.In this paper, we present RAPS, rate-adaptive positioning sys-tem for smartphone applications. It is based on the observationthat GPS is generally less accurate in urban areas, so it sufficesto turn on GPS only as often as necessary to achieve this accu-racy. RAPS uses a collection of techniques to cleverly determinewhen to turn on GPS. It uses the location-time history of the userto estimate user velocity and adaptively turn on GPS only if theestimated uncertainty in position exceeds the accuracy threshold.It also efficiently estimates user movement using a duty-cycled ac-celerometer, and utilizes Bluetooth communication to reduce po-sition uncertainty among neighboring devices. Finally, it employscelltower-RSS blacklisting to detect GPS unavailability (e.g., in-doors) and avoid turning on GPS in these cases. We evaluate RAPSthrough real-world experiments using a prototype implementationon a modern smartphone and show that it can increase phone life-times by more than a factor of 3.8 over an approach where GPS isalways on.Categories and Subject DescriptorsC.3 [Special-purpose and Application-based Systems]: Real-timeand embedded systems∗This material is based upon work supported by the National Sci-ence Foundation under Grant Nos. 0121778 and 0905596. Anyopinions, findings and conclusions or recomendations expressed inthis material are those of the author(s) and do not necessarily reflectthe views of the National Science Foundation (NSF).Permission to make digital or hard copies of all or part of this work forpersonal or classroom use is granted without fee provided that copies arenot made or distributed for profit or commercial advantage and that copiesbear this notice and the full citation on the first page. To copy otherwise, torepublish, to post on servers or to redistribute to lists, requires prior specificpermission and/or a fee.MobiSys’10, June 15–18, 2010, San Francisco, California, USA.Copyright 2010 ACM 978-1-60558-985-5/10/06 ...$10.00.Figure 1: Accuracy of GPS, WPS, and GSM-based PositioningGeneral TermsAlgorithms, Design, Experimentation, Performance, MeasurementKeywordsEnergy-Efficient, Adaptive Positioning, Smartphone, GPS, Sensor1. INTRODUCTIONMany emerging smartphone applications require position infor-mation to provide location-based or context-aware services. OurUrban Tomography [13] system is a good example. It allows auser to capture video clips, tags each with the most recent posi-tion information, and then automatically uploads each video to aserver. Many participatory sensing applications, where the phoneis autonomously recording ambient conditions or user activity, alsocontinuously record position information. Other applications thatmake continuous use of location or context information are Micro-Blog [11], TrafficSense [16], Pothole Patrol [8], MetroSense [7],PlaceIts [23], PeopleNet [17], MyExperience [10].In these applications, GPS is often preferred over its alternativessuch as GSM/WiFi based positioning systems because it is knownto be more accurate. Figure 1 plots two example locations fromwhich we have collected positioning data using all three position-ing systems: GPS, WPS (WiFi-based positioning system from Sky-HookWireless [22]), and GSM-based positioning on a N95 smart-phone over the AT&T cellular network. Both locations had a clearview of the sky and usable WiFi access points so that GPS andWPS could work. The figure clearly shows that WPS is less accu-rate than GPS, and GSM-based positioning has an error as high as300 meters. For these reasons, the use of GPS for location-based299Figure 2: Power consumption of GPS on N95 smartphone (Re-peated 30 seconds on and 90 seconds off)applications is unlikely to diminish; it may, however, be augmentedwith these other positioning methods.However, it is also well-known that GPS is extremely power-hungry. Our measurements of the power consumption1(shown inFigure 2) agrees with the results published by others [26, 3] andconfirms that the internal GPS on Nokia N95 smartphones usesaround 0.37 Watt of power on top of ∼0.06 W idle power. Keep-ing GPS activated continuously would drain the 1200 mAh batteryon an N95 smartphone in less than 11 hours, even in the absence ofany other activity. This is clearly a significant roadblock on the wayto all-day smartphone usage, and a more intelligent and energy-efficient activation of GPS is the subject of our paper.The key insight that motivates our work is the observation that,when used by pedestrians in urban areas, GPS can exhibit errorsin the range of 100m. GPS inaccuracy in urban “canyons” is well-known, but we have found that, even in relatively benign environ-ments such as college campuses or residential neighborhoods, GPScan exhibit this kind of inaccuracy, especially for pedestrian smart-phone usage. Location-based applications will have to deal withthis level of error using application-specific methods, such as map-matching or map-snapping. So we ask: if applications can toleratethis position error, why not trade off some position accuracy for re-duced GPS energy usage? A simple way to do this is to periodicallyduty-cycle GPS. This trades-off positioning accuracy for lower en-ergy. However, the key challenge in this periodic GPS duty-cyclingis to decide on
View Full Document
Unlocking...