Wireless Sensor Networks for Habitat Monitoring

1
Wireless Sensor Networks for Habitat Monitoring

• Jennifer Yick
• Network Seminar
• October 10, 2003

2
Outline

• Habitat monitoring
• Requirements
• System Architecture
• Hardware and Design
• Network and Application Interface
• Results

3
Habitat monitoring

• Objective monitor seabird nesting environment
  and behavior
• Key fact Seabird colonies are very sensitive to
  disturbance. Repeated disturbance will lead to
  abandonment of the colony.
• Major Problem The impact of human presence
  repeatedly can distort results by changing
  behavioral patterns and destroy sensitive
  populations.
• Solution Deployment of a sensor network

4
Great Duck Island and requirements

• Great Duck Island is a remote island south of
  Mount Desert Island, Maine
• Requirements
• Internet access to allow remote interactions
• Hierarchical network to provide enough
  connectivity over several kilometer
• Sensor network duration at least 9 months on
  non-rechargeable battery

5
Requirements cont

• Inconspicuous operation habitat monitoring
  should not disrupt the natural process or
  behavior under study
• Sensor network should be stable, predictable,
  and repeatable behavior if necessary
• Local interaction PDA assist with direct sensor
  query, adjust operational parameters or assist in
  locating devices
• Sensor type sense light, temperature, infrared,
  relative humidity, and barometric pressure
• Data compression archive sensor readings

6
System Architecture
7
Sensor Node

• UC Berkeley mote family called Mica
• Uses a single channel 916 MHz radio from RF
  monolithic to provide bidirectional communication
• Uses 2 AA batteries and a DC boost converter to
  stable voltage source
• Small in size (approx. 2.0x1.5x0.5 inches)

8
Sensor node functionality

• Each sensor node contains a computational module
  (a programmable unit) which provides computation
  ability, storage, and bidirectional communication
  with other nodes in the system
• Two advantages
• They can be re-task in the field
• Easily communicate with the rest of the network

9
Gateway

• Each sensor patch contains a gateway node
• Each gateway node can communicate with the sensor
  network and provides connectivity to the transit
  network

10
Transit Network

• Can consist of a single hop link or a series of
  networked wireless nodes
• Each transit network design has different
  characteristics
• Robustness
• Bandwidth
• Energy efficiency
• Cost
• Manageability

11
Base Station

• Data storage for the collection of sensor patches
• WAN connectivity will be wireless
• Base-Remote link connection to the internet

12
Possibility of Disconnection

• There is a possibility of disconnection at every
  level
• Each layer (sensor nodes, gateways, base station)
  has storage against data loss (power failure)
• Each layer has data management
• Sensor nodes data logging
• Gateway some database storing services
• Base station database services

13
Sensor Deployment

• 32 sensor nodes were deployed using Mica motes
  with Mica Weather Boards
• Mica motes
• Battery case integrated with the main processing
  and sensor boards
• Weather proof case
• First enclosed by a 10 micron parylene sealant to
  protect it from electrical contact
• Then enclosed by a transparent acrylic enclosure

14
(No Transcript)
15
Chosen hardware

• Gateway a single hop mote to mote network over
  CerfCube1 (an embedded linux system equipped with
  a compact flash adapter
• 1Cerfcube embedded StrongARM system.
  http//www.intrinsyc.com/products/cerfcube/.
  Intrinsyc Corporation, Vancouver BC Canada

16
Base Station

• Base station a laptop
• Each group of sensors communicates with a laptop
  which stores the data in a database

17
Base Station

• An external antenna permits communication from
  the sensor patch to the laptop-a distance of over
  350 feet.
• A satellite transciever is then used to replicate
  the data in the database to the research
  facilities a distance of over 50,000 miles

18
Data Management and User Interface

• Base station uses a SQL database which stores the
  time-stamped readings from the sensors
• Database is replicated every fifteen minutes
• User Interface is implemented on top of the
  sensor network database
• Matlab to analyze the data
• Java Applet to display

19
RESULTS

• http//www.greatduckisland.net

20
Conclusion

• Define the core application requirements
• Constraints on the sensor nodes have defined
  energy budget for all other services
• This study can be a representation of many
  applications in this area

21
References

• www.greatduckisland.net
• A. Mainwaring, J. Polastre, R. Szewczyk, D.
  Culler, J. Anderson, Wireless Sensor Networks
  for Habitat Monitoring, ACM International
  Workshop on Wireless Sensor Networks and
  Applications, Sept. 2002