Final Presentation Approved 10-14-09.ppt · 2010-02-26 · Motivation Mobile phones are readily available and have lots of functionality ... Microsoft PowerPoint - Final Presentation

ALMONDAcoustic Localization for Mobile Open-Source

N k D lNetwork DeploymentScott Almquist

Muhammad SaleemMuhammad SaleemDaniel Skehan

Worcester Polytechnic InstituteMajor Qualifying Project

14 October 2009

Embedded Digital Systems – Group 102

ALMOND10/14/2009

MIT Lincoln LaboratoryThis work is sponsored by the Department of the Air Force under Air Force contract FA8721-

05-C-0002. Opinions, interpretations, conclusions and recommendations are those of the author and are not necessarily endorsed by the United States Government.

Approved for public release - distribution is unlimited

Acknowledgements

• WPI Advisors:– Professor ClancyProfessor Clancy– Professor Heineman

• Lincoln Laboratory Advisors– Albert Reuther– Glenn Schrader

MIT Lincoln LaboratoryALMOND10/14/2009 2

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 3

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 4

Acoustic Localization

• Determining the location of a

sound source by using an

array of sensorsSound

Source

• The difference in arrival times

is used to calculate an

approximate position of the

t1t2

approximate position of the

sound sourceMicrophones

Time Differences of Two Microphones Result in a

Hyperbola

MIT Lincoln LaboratoryALMOND10/14/2009 5

yp

Acoustic Localization

Microphone 3

Microphone 2

t2t3

Sound Source

t1

Microphone 1

MIT Lincoln LaboratoryALMOND10/14/2009 6

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 7

Motivation

Mobile phones are readily available and have lots of functionality

• GPS

• Camera

• Wireless Communication

• MicrophoneMicrophone

• Operating System

Phone architecture suitable for many applications

MIT Lincoln LaboratoryALMOND10/14/2009 8

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 9

Setup and Challenges

ComputationSignal DetectionSignal DetectionTime Synchronization

OpenmokoPhonePhone

Sound SourceLaptop PC

Wireless Router

OpenmokoPhone

MIT Lincoln LaboratoryALMOND10/14/2009 10

Processing Overview

Phones are Unable to Perform Necessary Processing y g

MIT Lincoln LaboratoryALMOND10/14/2009 11

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 12

Signal Detection

• Chirps are signals that can be

easily detected because of

their unique characteristics

Th b d t t d i• They can be detected using a

cross correlation

• Cross correlation of two• Cross correlation of two

similar chirps will create a

peak

Detection of Chirps is Simple

MIT Lincoln LaboratoryALMOND10/14/2009 13

Signal Detection

Chirp

Threshold

ChirpVentilation Noise

MIT Lincoln LaboratoryALMOND10/14/2009 14

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 15

Time Synchronization

Time synchronization errors are the greatest source of positional error

Mic 1 Mic 2

1.7 meters minimum error

MIT Lincoln LaboratoryALMOND10/14/2009 16

GPS Time Synchronization

Global Positioning System (GPS)• P i i f 100 d 99% f th ti• Precision of 100 nanoseconds 99% of the time• Can be used with Network Time Protocol (NTP) to discipline

the phones’ clocks

Openmoko hardware is not built to take advantage of precise GPS timing

MIT Lincoln LaboratoryALMOND10/14/2009 17

Time Synchronization Methods

Network Broadcast Signal• A message is broadcasted across the network• A message is broadcasted across the network• When the phones receive the message time is reset• Both phones receive the signal at the same time resulting in

synchronizationy

Calibration Chirp• An initial chirp is produced for calibrationAn initial chirp is produced for calibration• The time offset from the two phones is calculated based on

known positions• Future detection calculations adjust time based on the

ff toffset

MIT Lincoln LaboratoryALMOND10/14/2009 18

Time Synchronization Test Setup

If the phones are equidistant from the sound source, the time difference of arrival should be zerodifference of arrival should be zero

MIT Lincoln LaboratoryALMOND10/14/2009 19

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 20

Results

Timing Error Using Network Broadcast

50

60

70

econ

ds)

10

20

30

40

50

Diff

eren

ce (m

illis

e

-10

0

10

0 200 400 600 800 1000 1200 1400

Time from Phone One After Synchronization (seconds)

Tim

e D

Timing Error Using Chirp Calibration

2

4

6

renc

es

nds)

-6

-4

-2

00 50 100 150 200 250 300

Tim

e D

iffer

(mill

isec

on

MIT Lincoln LaboratoryALMOND10/14/2009 21

-8Time on Phone Two After Synchronization (seconds)

Outline

• Acoustic Localization

• Motivation

• Setup and Processingp g

• Signal Detection

• Time SynchronizationTime Synchronization

• Results

• Conclusions and Future Work• Conclusions and Future Work

MIT Lincoln LaboratoryALMOND10/14/2009 22

Conclusions and Future Work

• Sound detection is possible with cell phones

• With the current setup, time can be synchronized to within 10 milliseconds for several minutes

• A more refined method of time synchronization would probably be required for future applications

– Correction for drift rateMore precise calibrations– More precise calibrations

– Periodic synchronizations

• Bandwidth can be saved by performing processing on moreBandwidth can be saved by performing processing on more capable phones

MIT Lincoln LaboratoryALMOND10/14/2009 23