Localization problem with Sensor Networks

1
Localization problem with Sensor Networks

• By
• Jayanth Patil
• Ganesh Godavari

2
Localization

• Localization The process of determining or
  marking the location or site
• Why is Localization a tough problem
• ubiquitous computing requires high security and
  high resolution indoors (MIT CRICKET)
• military planning requires high security and low
  resolution outdoors .
• Problem goals changes from one scenario to another

3
Localization for Sensor Networks

• Sensor networks generally require
• low security and low resolution
• assume minimal infrastructure, hardware and
  deployment-time overhead.
• Problem complexity increases?

4
Related Work

• Calamari Univ. of Berkeley
• aims to explore the space of localization
  solutions in the sensor network domain
• tries to give each sensor node in an ad-hoc
  wireless sensor network an (x,y) coordinate with
  respect to some known coordinate system.
• http//www.cs.berkeley.edu/kamin/calamari/

5
Mobi Loc

• Mobility Enhanced Localization Univ. of
  Berkeley
• aims to study relationship between mobility,
  navigation, and localization in the context of
  wireless sensor networks and mobile objects.
• http//www.eecs.berkeley.edu/prabal/projects/cs29
  4-1/

6
Localization Questions

• How to determine where the object is ?
• Need some reference object to determine.
• How to measure how far is the object ?
• Use radio signal strength, Round Trip Time
• Is the radio signal strength or RTT a good
  measure?
• No Disruptive forces of Nature

7
Signal Strength vs. Distance
8
Noise Vs. Distance
9
How to determine Where the object is ?

• Existing technologies
• Global Positioning System
• Designed for outdoor localization
• High power consumption
• Provides about 2-3 meter resolution
• Not suitable for sensor networks
• Cricket
• Used for Ubiquitous computing.
• Passive listeners self localization
• Not scalable for ad-hoc needs

10
Contd..

• AHLoS under development _at_ UCLA
• Similar to cricket but uses Ultra sound
• Millibots
• Miniature robots
• Use Leap Frog approach
• Uses centralized approach, group leader
• Not feasible for large ad-hoc deployment.

11
Lessons Learned

• Could not get calamari to work
• Calamari needs a separate ultrasonic transciever
  board.
• Looked into calamari project and learnt a lot
  about tinyos, mica nodes
• Currently working on repeating the testing of
  radio signal strength with distance

12
References

• This document is prepared based on the thesis
  work done by Cameron Dean Whitehouse,
• All information presented here is taken from
  http//www.cs.berkeley.edu/kamin/calamari/
• The authors claim no right. All quoted
  products/work belong to the respective companies
  working on it.