The Cricket Location-Support System

1
The Cricket Location-Support System

• By Min Chen
• 10/28/03

2
Goal

• User Privacy
• Decentralized administration
• Network heterogeneity
• Low cost
• Room-sized granularity

3
System Architecture

• Beacon
• Disseminate the string of space information
  about a geographic space to listeners.
• Listener Infer its current location from the set
  of beacons .

Approach

• Use combination of RF and ultrasound to provide
  a location-support service to users.

4
Related Problem and Method

• Collision of RF transmissions from different
  beacons
• ------------Randomization
• Wrong correlation of the RF data of one beacon
  with the ultrasonic signal of another
• ------ System parameters
• ------ Listener Inference Algorithms

5
System Parameters Selection

• use a relatively sluggish RF data transmission
  rate.

6
Interference Scenarios

• RF-AUS-RA
• Align the beacons
• RF-AUS-I
• Using RF signal with long range
• RF-AUS-RI
• Ensure less than 5 beacons within range of each
  other

7
Listener Inference Algorithm

• Majority
• Picks the beacon with highest frequency of
  occurrence
• MinMean
• Selects the beacon with the minimum mean value
• MinMode
• Compute the per-beacon statistical mode over the
  past n samples.

8
Implementation Beacon Positioning and
Configuration
9
Implementation Ultrasound Deployment
10
Experiment Boundary Performance
11
Experiment Static Performance
12
Experiment Static Performance (Cont)
13
Experiment Mobile Performance
14
Comments

• Concerns when apply to sensor network
• 2D Constrain
• Power Consideration
• Calibration On the Transmitter.
• Kalman Filter for the Mobile Algorithm