Improving Energy Efficiency of Location Sensingon SmartphonesZhenyun Zhuang1 ∗Kyu-Han Kim2 †Jatinder Pal Singh21Georgia Institute of Technology, Atlanta, GA 30332, U.S.A.2Deutsche Telekom R&D Laboratories USA, Los Altos, CA 94022, [email protected], [email protected], [email protected] applications have become increasingly popular onsmartphones over the past years. The active use of these applica-tions can ho wever cause device battery drain owing to their power-intensive location-sensing operations. This paper presents an adap-tive location-sensing framework that significantly improves the en-ergy efficiency of smartphones running location-based applications.The underlying design principles of the proposed framework in-volve substitution, suppression, piggybacking, and adaptation ofapplications’ location-sensing requests to conserve energy. We im-plement these design principles on Android-based smartphones as amiddleware. Our evaluation results show that the design principlesreduce the usage of the power-intensive GPS (Global PositioningSystem)byupto98%andimprovebatterylifebyupto75%.Categories and Subject DescriptorsC.3.3 [Special-Purpose and Application-Based Systems]: Real-Time and Embedded SystemsGeneral TermsDesign, Experimentation, Measurement, Performance, AlgorithmsKeywordsLocation Sensing, Energy Efficiency, Location-Based Applications,Smartphone.1. INTRODUCTIONWith the increasing perv asiveness of smartphones over the pastyears, se veral Location-Based Applications (LBAs) have been adoptedby mobile users for always-on contact for social-networking, busi-nesses needs, and entertainment. Some instances of currently pop-ular LBAs include mobile social networking [2, 3, 9, 10], health-care [1], local traffic [7, 22, 23, 29, 36], and local restaurants [6].∗Zhenyun Zhuang was an intern at Deutsche Telekom R&D Lab USA whilethis research was conducted.†To whom correspondence should be addressed.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.In spite of the increase in processing po wer, feature-set, andsensing capabilities, the smartphones continue to suffer from bat-tery life limitation, which hinders the active utilization of LBAs.Typical battery capacity of smartphones today is barely above 1000mAh (e.g., the lithium-ion battery of HTC Dream smartphones hasthe capacity of 1150 mAh). Unfortunately, GPS (Global Position-ing System), the core enabler of LBAs, is power-intensive, and itsaggressive usage can cause complete drain of the battery within afew hours [14, 17]. While the aggressiveness of GPS usage is spe-cific to different applications, sev eral LBAs such as local traffic(e.g., [7]) and social networking (e.g., [9]) particularly benefit fromcontinuous location updates. Real Time Traffic [7], for instance,requires continuous GPS location updates. Twidroid [9], a mobileversion of Twitter, features a GPS accuracy booster, which providesan option to enable/disable continuous GPS sensing.Numerous solutions have been proposed to improve the batterylife of mobile devices [11, 32–34], but little rigor and attentionhas been devoted to the battery-efficient use of LBAs. The LBAdevelopers are suggested to reduce the use of GPS by increasinglocation-update intervals (say, to more than a minute), thus allow-ing GPS hardware to sleep between successive location-updates.Such a simple solution can improve battery life by forcing applica-tions to request location information less frequently, but it has fun-damental limitations. For instance, although each LBA can saveenergy by reducing GPS invocation, the effectiveness of this ap-proach could be compromised when multiple LBAs are running, asthe asynchronous use of GPS from different LBAs unnecessarilyleads to an increased number of in vocations.In this paper, we present an energy-efficient location-sensingframework that effectiv ely conserves energy for smartphones run-ning LBAs. In its core, the proposed frame work includes four de-sign principles: Substitution, Suppression, Piggybacking and Adap-tation. Briefly, Substitution makes use of alternative location-sensingmechanisms (e.g., network-based location sensing) that consumeslower power than GPS. Suppression uses less power-intensive sen-sors such as an accelerometer to suppress unnecessary GPS sensingwhen the user is in static state. Piggybacking synchronizes the lo-cation sensing requests from multiple running LBAs. Adaptationaggressively adjusts system-wide sensing parameters such as timeand distance, when battery level is low.We implement the four design principles on a G1 Android De-veloper Phone (ADP) as a middleware and evaluate the implemen-tation extensively via measurements. While the proposed designprinciples are general enough to be applied to any software stack,the middleware implementation allows for better application trans-parency in the sense that applications can be kept as-is. We chooseAndroid-based smartphones for prototyping because of the open-ness of the Android platform [5]. Our evaluation results with the31580 85 90 95 100 0 10 20 30 40 50 60Battery level (%)Time (minutes)GPS disabledGPS enabled(a) Real Time Traffic 0 200 400 600 800 1000 0 20 40 60 80 100 120 140 160 180Power level (mW)Time (Sec)GPS power measure (every 50ms)(b) Power spikes 80 85 90 95 100 0 10 20 30 40 50 60Battery level (%)Time (Minute)One applicationTwo applications(c) Multiple LBAsFigure 1: Energy Consumption of Gpsimplementation sho w that the proposed framework significantlysav es energy in location sensing. For instance, in various scenar-ios, our prototype reduces the number of GPS invocations by up to98%, and thus improves the battery life by up to 75%.To summarize, this work makes the following contributions:• We address and explore energy efficiency of location sensingfor resource-constrained smartphones that often run multiplelocation-based applications (LBAs).• We study four design principles tailored for LBAs to reduce
View Full Document
Unlocking...