Mobile Phone Sensing Broadly Defined Mahesh K. Marina School of - - PowerPoint PPT Presentation

Mobile Phone Sensing Broadly Defined

Mahesh K. Marina

School of Informatics The University of Edinburgh mahesh@ed.ac.uk http://homepages.inf.ed.ac.uk/mmarina/

Joint work with Matt Calder, Marinos Argyrou, Arsham Farshad, Richard Mortier, Ciprian Balas and Valentin Radu

Modern Smartphone

iPhone 4

• Sensors:

– Touchscreen – 3-axis Accelerometer – (Dual) Camera (Video) – (Dual) Microphone – Digital Compass – 3-axis Gyroscope – Proximity sensor – Ambient light sensor

• A-GPS

Nexus S (Android)

• Sensors:

– Touchscreen – 3-axis Accelerometer – (Dual) Camera – (Dual) Microphone – Digital Compass – 3-axis Gyroscope – Proximity sensor – Ambient light sensor

• A-GPS, NFC reader

3

Applications are key drivers for new sensors getting added to phones

Interesting Examples

CellScope (Berkeley) NETRA (MIT)

4

Can we expand the scope of sensing applications

• f mobile phones relying on what is already built

in or likely to be integrated in near future?

5

Our Focus

• UHF RFID reader integration for making the phone

into a generic sensor data acquisition device

– Environmental monitoring, esp. for low cost sensing in indoor environments with dense sensor tag deployments

• (Explicitly) View Network Interface as a Sensor on

Phone

– Wireless monitoring (indoor, outdoor, WiFi, 3G) for client side view

6

What is the Energy Consumption Impact of UHF RFID Reader Integration with Mobile Phones?

7

Experimental Methodology (1)

Three types of tags, incl. a semi-passive temperature logger tag Two types of compact readers

8

Quark A528

Experimental Methodology (2)

• HTC Magic (Android) phone
• Digital multimeter + power supply

setup

• PC application for receiving & plotting

power measurements over time

9

Impact of Reader Operations and Reader-Tag Separation

10

Impact of Reader “Q” Parameter Setting

11

Power vs. Energy Consumption

12 .

13

Further Issues

• Impact of non-ideal channel and environmental

conditions (e.g., non-line-of-sight, orientation)

• Adaptively choosing the Q parameter based on a

predictive model

• Reader scheduling / duty cycling

14

Summary

• UHF RFID reader integration with mobile phones

for low cost environmental monitoring

• (Explicitly) View Network Interface as a Sensor
• n Phone for wireless monitoring
• Crowdsourcing based sensing apps

– Incentives and addressing privacy concerns key to real world adoption

15

Thanks!