29
Diversity in Smartphone Usage MobiSys ‘10 June 17, 2010 UCLA, Microsoft, USC Hossein Falaki, Ratul Mahajan, Srikanth Kandula Dimitrios Lymberopoulos, Ramesh Govindan, Deborah Estrin

Diversity in Smartphone Usage

  • Upload
    langer

  • View
    53

  • Download
    0

Embed Size (px)

DESCRIPTION

Diversity in Smartphone Usage. Hossein Falaki , Ratul Mahajan , Srikanth Kandula Dimitrios Lymberopoulos , Ramesh Govindan , Deborah Estrin. UCLA, Microsoft, USC. MobiSys ‘10 June 17, 2010. Smartphone Penetration Is on the Rise. Basic Facts about Smartphone Usage Are Unknown. - PowerPoint PPT Presentation

Citation preview

Page 1: Diversity in Smartphone Usage

Diversity in Smartphone Usage

MobiSys ‘10June 17, 2010

UCLA, Microsoft, USC

Hossein Falaki, Ratul Mahajan, Srikanth KandulaDimitrios Lymberopoulos, Ramesh Govindan, Deborah Estrin

Page 2: Diversity in Smartphone Usage

2

2006 2007 2008 2009 2010 2011 2012 2013 20140

50

100

150

200

250

300

350

400

Western Europe Asia & Pacific

North America

(Source: Park Associates, 2009)

Smar

tpho

ne u

sers

(mill

ions

)

Smartphone Penetration Is on the Rise

Page 3: Diversity in Smartphone Usage

3

Basic Facts about Smartphone Usage Are Unknown

Page 4: Diversity in Smartphone Usage

4

Why Do We Need to Know These Facts?

How can we improve smartphone performance and usability?

Identical usersEveryone is different

?Can we improve resource management on

smartphones through personalization?

Page 5: Diversity in Smartphone Usage

5

Main Findings

1. Users are quantitatively very diverse in their usage

2. But invariants exist and can be harnessed

Page 6: Diversity in Smartphone Usage

6

Platform Demographics

Android 16 high school students17 knowledge workers

WinMobile 16 Social Communicators56 Life Power Users59 Business Power Users37 Organizer Practicals

Platform Information Logged

Android Screen stateApp usageBattery levelNet traffic per appCall starts and ends

WinMobile Screen stateApplications used

Data SetsPlatform # Users Duration

Android 33 7-21 Weeks/user

WinMobile 222 8-28 Weeks/user

Page 7: Diversity in Smartphone Usage

7

Diversity in interactionInteraction model

Diversity in application usageApplication usage model

Diversity in battery usageEnergy drain model

Outline

Comprehensive system view

Interaction Application Energy

Page 8: Diversity in Smartphone Usage

8

Users have disparate interaction levels

Two orders

Page 9: Diversity in Smartphone Usage

9

Sources of Interaction Diversity

1. User demographics2. Session count3. Session length4. Application use 5. Number of applications per session

Page 10: Diversity in Smartphone Usage

10

User Demographics Do Not Explain Diversity

Page 11: Diversity in Smartphone Usage

11

Session Lengths Contribute to Diversity

Page 12: Diversity in Smartphone Usage

12

Number of Sessions Contribute to Diversity

Page 13: Diversity in Smartphone Usage

13

Session Length and Count Are Uncorrelated

Page 14: Diversity in Smartphone Usage

14

Close Look at Interaction Sessions

Most sessions are short

Sessions terminated by screen timeout

Few very long sessions

Exponential distribution

Shifted Pareto distribution

Page 15: Diversity in Smartphone Usage

15

Modeling Interaction Sessions

Extremely long sessions are

being modeled well

Page 16: Diversity in Smartphone Usage

16

Implications of Interaction Diversity

• System parameters such as timeouts can be tuned based on model parameters

• System can be designed with insights from the distributions

Diversity Interaction Models

System Design Implications

Page 17: Diversity in Smartphone Usage

17

Diversity in application usageApplication usage model

Outline

Interaction

Application

Energy

Diversity in interactionInteraction model

Page 18: Diversity in Smartphone Usage

18

Users Run Disparate Number of Applications

50% of users run more than

40 apps

Page 19: Diversity in Smartphone Usage

19

Application Breakdown

Page 20: Diversity in Smartphone Usage

20

Close Look at Application PopularityStraight line in semi-log plot

appears for all users

Different list for each user

Page 21: Diversity in Smartphone Usage

21

Exponential Distribution Models App Popularity Well

Page 22: Diversity in Smartphone Usage

22

Implications of Application Diversity

• Most of a user’s attention is focused on a few applications• Optimize the system for the top applications for each user

Diversity Application Models

System Design Implications

Page 23: Diversity in Smartphone Usage

23

Diversity in application usageApplication usage model

Outline

Interaction

Application

Energy

Diversity in interactionInteraction model

Diversity in energy drainPredicting energy drain

Page 24: Diversity in Smartphone Usage

24

Users Are Diverse in Energy Drain

Two orders

Page 25: Diversity in Smartphone Usage

25

Close Look at Energy Drain

Significant variation across

time

High variation within each hour

Page 26: Diversity in Smartphone Usage

26

“Trend Table” Based Framework to Model Energy Drain

Page 27: Diversity in Smartphone Usage

27

Modeling Energy Drain

Page 28: Diversity in Smartphone Usage

28

Conclusions

Users are quantitatively diverse in their usage

Invariants exist and can be harnessed

• Building effective systems for all users is challenging• Static policies cannot work well for all users

• Users have similar distributions with different parameters.• This significantly facilitates the adaptation task

Page 29: Diversity in Smartphone Usage

Diversity in Smartphone Usage

MobiSys ‘10June 17, 2010

Hossein [email protected]