Energy-Delay Tradeoffs in Smartphone Applications∗Moo-Ryong Ra†Jeongyeup Paek†Abhishek B. Sharma†Ramesh Govindan†Martin H. Krieger∗Michael J. Neely⋆Computer Science Dept.†School of Policy, Planning, and Development∗Electrical Engineering Dept.⋆University of Southern California, Los Angeles, CA, USA{mra, jpaek, absharma, ramesh, krieger, mjneely} @ usc.eduABSTRACTMany applications are enabled by the ability to capture videos ona smartphone and to have these videos uploaded to an Internet-connected server. This capability requires the transfer of large vol-umes of data from the phone to the infrastructure. Smartphoneshave multiple wireless interfaces – 3G/EDGE and WiFi – for datatransfer, but there is considerable variability in the availability andachievable data transfer rate for these networks. Moreover, the en-ergy costs for transmitting a given amount of data on these wirelessinterfaces can differ by an order of magnitude. On the other hand,many of these applications are often naturally delay-tolerant, sothat it is possible to delay data transfers until a lower-energy WiFiconnection becomes available. In this paper, we present a prin-cipled approach for designing an optimal online algorithm for thisenergy-delay tradeoff using the Lyapunov optimization framework.Our algorithm, called SALSA, can automatically adapt to channelconditions and requires only local information to decide whetherand when to defer a transmission. We evaluate SALSA using real-world traces as well as experiments using a prototype implementa-tion on a modern smartphone. Our results show that SALSA canbe tuned to achieve a broad spectrum of energy-delay tradeoffs, iscloser to an empirically-determined optimal than any of the alter-natives we compare it to, and, can save 10-40% of battery capacityfor some workloads.Categories and Subject DescriptorsC.4 [Performance of Systems]: Design Studies—Energy Manage-ment on Smartphones∗This research was sponsored by the USC/CSULB METRANS Trans-portation Center and by the Army Research Laboratory under CooperativeAgreement Number W911NF-09-2-0053. The views and conclusions con-tained in this document are those of the authors and should not be inter-preted as representing the official policies, either expressed or implied, ofthe METRANS center, the Army Research Laboratory or the U.S. Gov-ernment. The U.S. Government is authorized to reproduce and distributereprints for Government purposes notwithstanding any copyright notationhereon. In addition, the first author, Moo-Ryong Ra, was supported by An-nenberg Graduate Fellowship.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: Urban Tomography SystemGeneral TermsAlgorithms, Design, Experimentation, Measurement, Theory, Per-formanceKeywordsWiFi, Interface Selection, Smartphone, Lyapunov Optimization1. INTRODUCTIONAs video-enabled smartphones become more prevalent, manynew and interesting applications will be enabled. Our Urban To-mography system [25, 13] is a good example. It allows a userto capture video clips, and then automatically uploads them in thebackground to a server. The system has been operational for overa year and has found several, qualitatively different, uses. A teamof security officials, equipped with smartphones, has been using itfor surveillance at a large transportation hub in Los Angeles. Theteam is able to visually document parts of the facility not coveredby fixed cameras, is able to provide in situ views of developing sit-uations, and, because the videos are automatically uploaded to aserver, the team’s supervisors are able to accurately assess a devel-oping situation. A company that specializes in behavior analysisof developmental disabilities in children has also been piloting thesystem. Their mobile childcare specialists visit area schools, andrecord the behavior of children for analysis by parents and medicalexperts. A professor of public planning and her students have usedour system to document construction in post-Katrina Mississippi,with the goal of evaluating zoning regulations and revising existingordinances.These, and other, users have generated a corpus of over 5000255EDGE3G200 KB/sWiFivideo 2arrives50 KB/svideo 1arrivesarrives40 KB/s50 KB/s10 KB/sMinim m dela algorithmDatao erEDGEMinimum delay algorithm: Total energy= 246 J, Total delay= 246 sData over EDGEData over 3GData over WiFiAlways use WiFi: Total energy = 95 J, Total delay = 305 sEnergy optimal: Total energy=50 J, Total delay=320 sTotal energy 50 J, Total delay 320 sFigure 2: Examplevideos. Figure 1 presents a screenshot of the system’s Web in-terface, showing some user-generated video-clips from our users.Our users report that battery lifetime is a critical usability issue,and video uploads use a significant fraction of the energy in oursystem. This paper explores robust methods for reducing this cost.Recent smartphones have multiple wireless interfaces – 3G/EDGE(Enhanced GPRS) and WiFi – that can be used for data transfer.These two radios have widely different characteristics. First, theirnominal data rates differ significantly (from hundreds of Kbps forEDGE, to a few Mbps for 3G, to ten or more Mbps for WiFi). Theachievable data rates for these radios depends upon the environ-ment, can vary widely, and are sometimes far less than the nominalvalues. Second, their energy-efficiency also differs by more thanan order of magnitude [4, 6]. While the power consumption on thetwo kinds of radios can be comparable, the energy usage for trans-mitting a fixed amount of data can differ an order of magnitude ormore because the achievable data rates on these interfaces differsignificantly. Finally, the availability characteristics of these twokinds of networks can vary significantly. At least as of this writing,the penetration of some form of cellular availability (EDGE or 3G)is significantly higher than WiFi, on average. A similar observa-tion has been made in [22] where the authors report 99% and 46%experienced availability,
View Full Document
Unlocking...