Mobi-Sync: Efficient Time Synchronization for Mobile Underwater Sensor Networks Jun Liu, Zhong Zhou,...

Preview:

Citation preview

Mobi-Sync: Efficient Time Synchronization forMobile Underwater Sensor NetworksJun Liu, Zhong Zhou, Zheng Peng and Jun-Hong CuiComputer Science & Engineering Department, University of ConnecticutIEEE Globecom 2010

Outline

• Introduction

• Challenge

• Goals

• Networks architecture

• Mobi-sync

• Simulation

• Conclusion

Introduction

• The world's oceans cover over 70 % of its surface• Underwater Wireless Sensor Networks (UWSNs)

Introduction

• Transmission rate• WSN: 3 x 108 m/s

• UWSN: 1500 m/s

• Propagation delay

• Time synchronization

A3:00

B3:00

3:10

Introduction

• Time synchronization

• Clock divergence• Clock drift

• Clock offset

iii btatC )(

Challenge

• Long delays significantly affect the synchronization accuracy

• Propagation delays are changing continuously in mobile UWSNs

• synchronization even more difficult

• Energy consumption is heavy of acoustic transmissions, so energy efficiency is very important

Goals

• A time synchronization scheme for mobile UWSNs

• high accuracy

• energy efficiency

Networks architecture• Surface Buoys: • GPS receivers to obtain global time references

• Super Nodes: • communicate with the surface buoys and get synchronized

• offer time and speed information to ordinary nodes

• Ordinary Nodes:

• synchronized with neighbors

Mobi-sync

• Mobi-Sync consists of three phases

• Phase 1: propagation delay estimation

• Message exchange and delay calculation

• Phase 2: linear regression

• Phase 3: calibration

Mobi-sync

• Phase 1

• Message exchange

Tim

e

T1

T4

T2

T3

T5

T6

Ordinary node Super node ”A”

tr1

tr2

SR

RS1

RS2

One run message exchange

T1,T3,T5 : sending time of SR,RS1,RS2

T2,T4,T6 : receiving time of SR,RS1,RS2

tr1,tr2 : the first ,second response time

Mobi-sync

• Phase 1

• Delay calculation

Ordinary nodeSuper node ”A”

d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

β

θ

d1 : are propagation distance of SRd2 : are propagation distance of RS1

d3 : are propagation distance of RS2

L1,L2 : straight-line distance super node “A” move relatively to the ordinary node during tr1,tr

Mobi-sync

Vp : propagation speed

Ordinary nodeSuper node ”A”

d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

β

θ

T1 T2 T3 T4 T5 T6

1ms 5ms 7ms 11ms13ms 17ms time

tr1

tr2

iii btatC )(

Mobi-sync

Super node ”A”d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

β

θ

Ordinary nodeSuper node ”A”

d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

θ

Ordinary node

assumption

β=0

α

Mobi-sync

Super node ”A”d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

θ

Ordinary node

α

L1x =Vx*ti L2x =Vx*ti

L1y =Vy*ti L2y =Vy*ti

Vx=0.3m/msVy=0.4m/ms

L1tr1=2 L1=1mL2tr1+tr2=8 L2=4m

T1 T2 T3 T4 T5 T6

1ms 5ms 7ms 11ms13ms 17ms time

tr1

tr2

Mobi-sync

Super node ”A”d1

d2

d3

T1

T2

T3

T4

T5

T6

L2

L1

θ

Ordinary node

α

Combine Cosine theorem for common angle α

L1=1 , L2=4 , h1=18 , h2=36τ1= fd(L1,L2,h1,h2)=1

Mobi-sync

• Phase 2

• Linear regression

Tim

e

T1

T4

T2

T3

T5

T6

Ordinary node Super node ”A”

SR

RS1

RS2

One run message exchange

i is the index of the messages exchange round

Mobi-sync

• Phase 3

• Calibration

• Due to nodes mobility, the distance d1might be different from the distance d2, so the initial distance

• Update initial distance “r” and re-calculation the speed vectors

• We assign the initial skew as “1”. And since the first estimated skew has been obtained, we can update it and re-calculation.

r : initial distance between an ordinary node and a super node

Simulation

Simulation

• Related works

• TSHL• TSHL combines one-way and two way MAC-layer message

delivery. One-way communication allows TSHL to estimate the clock skew, and Two-way is to compute the clock offset.

• MU-Sync• The first linear regression allows the cluster head to estimate

the draft skew by totally ignoring the variance of propagation delays. And the second one is used to correct the estimated skew and calculate the offset.

Simulation

Simulation

Conclusion

• we presented Mobi-Sync, a novel time synchronization scheme for mobile UWSNs.

• Mobi-Sync objects to improve the synchronization accuracy as well as the energy efficiency.