Wireless Sensor Networks and Real-World Applications

1
Wireless Sensor Networks and Real-World
Applications

• Nirupama Bulusu
• Portland State University
• http//www.cs.pdx.edu/nbulusu

2
On Sensor Networks

• One of the 10 technologies that will change the
  world,
• MIT Technology Review, 2003
• More than half a billion sensor nodes will ship
  for wireless sensor applications in 2010 for an
  end-user market worth at least 7 billion
• Demand growing at 300 between 2004 and 2005.
• ON World, a wireless research firm.

3
Burgeoning Research and Commercial Activity

• NSF Research Centers
• Center for Embedded Networked Sensing
• More than 100 Companies (many started after 2003)
• Crossbow, Sensoria, Millennial Net, Ember
  Corporation, Dust Networks, Chipcon, Arched Rock
  Corporation, Moteiv

4
Push Technology Trends

• Moores law
• Energy capacity miniaturization
• Micro-electro Mechanical Systems (MEMS)
• System-on-chip Integration

5
Wireless Sensor Networks

• Micro-sensors, on-board processing, wireless
  interfaces feasible at very small scale--can
  monitor phenomena up close
• Enables spatially and temporally dense
  environmental monitoring

Sensing
Computing
Communication
6
State-of-the-Art

• Telos Mote
• (Source David Culler, Berkeley)

7
Pull Real-world Applications

• Most applications fall into of one of three
  categories
• Monitoring Space
• Monitoring Objects
• Monitoring Interactions of Objects and Space

Classification due to Culler, Estrin, Srivastava
8
Monitoring Space

• Environmental and Habitat Monitoring
• Precision Agriculture
• Indoor Climate Control
• Military Surveillance
• Treaty Verification
• Intelligent Alarms

9
Example Precision Agriculture

• The Wireless Vineyard
• Sensors monitor temperature, moisture
• Roger the dog collects the data
• Source Richard Beckwith,
• Intel Corporation

10
Monitoring Objects

• Structural Monitoring
• Eco-physiology
• Condition-based Maintenance
• Medical Diagnostics
• Urban terrain mapping

11
Example Condition-based Maintenance

• Intel fabrication plants
• Sensors collect vibration data, monitor wear and
  tear report data in real-time
• Reduces need for a team of engineers cutting
  costs by several orders of magnitude

12
Monitoring Interactions between Objects and Space

• Wildlife Habitats
• Disaster Management
• Emergency Response
• Ubiquitous Computing
• Asset Tracking
• Health Care
• Manufacturing Process Flows

13
Example Habitat Monitoring

• The ZebraNet Project
• Collar-mounted sensors monitor zebra movement in
  Kenya

Source Margaret Martonosi, Princeton University
14
Tracking node with CPU, FLASH, radio and GPS
Data
Store-and-forward communications
Data
Base station (car or plane)
Data
Data
Sensor Network Attributes ZebraNet Other Sensor Networks
Node mobility Highly mobile Static or moderate mobile
Communication range Miles Meters
Sensing frequency Constant sensing Sporadic sensing
Sensing device power Hundreds of mW Tens of mW
15
The Computing Challenge

• Build Robust, Long-lived systems that can be
  un-tethered (wireless) and unattended
• Communication will be the persistent primary
  consumer of scarce energy resources (MICA Mote
  720nJ/bit xmit, 4nJ/op)
• Autonomy requires robust, adaptive,
  self-configuring systems
• Leverage data processing inside the network
• Exploit computation near data to reduce
  communication, achieve scalability
• Collaborative signal processing
• Achieve desired global behavior with localized
  algorithms (distributed control)

16
Some Problems
Power-aware Networking low-power media access
power-aware routing of data packets Macro-program
ming high-level program for a sensor network
not low-level programs for individual sensors

• Calibration correcting systematic errors in
  sensor data
• Causes manufacturing, environment, age, crud
• Localization establish spatial coordinates for
  sensors and target objects

17
In-depth Localization
18
Mathematically
11
10

• Given xi, cij for some i, j 1, N
• Estimate xs for any s

C5.11 5
9
7
5
6
8
1 (0,0,0)
C23 5
3
2
4(100,0,0)
19
Localization System Components
Stitching and Refinement
This step applies to distributed construction of
large-scale coordinate systems
This step estimates target coordinates (and often
other parameters simultaneously)
Coordinate System Synthesis
Coordinate System Synthesis

• Parameters might include
• Range between nodes
• Angle between nodes
• Psuedo-range to target (TDOA)
• Bearing to target (TDOA)
• Absolute orientation of node
• Absolute location of node (GPS)

20
Example of a Localization System

• SHM system, developed at Sensoria Corp.

Each node has 4 speaker/ microphone pairs,
arranged along the circumference of the
enclosure. The node also has a radio system and
an absolute orientation sensor that senses
magnetic north.
Microphone
Speaker
12 cm
Source Lewis Girod, UCLA
21
System Architecture

• Ranging between nodes based on detection of coded
  acoustic signals, with radio synchronization to
  measure time of flight
• Angle of arrival is determined through TDOA and
  is used to estimate bearing, referenced from the
  absolute orientation sensor
• An onboard temperature sensor is used to
  compensate for the effect of environmental
  conditions on the speed of sound

22
System Architecture

• Nodes periodically emit acoustic pulses. Other
  nodes detect these pulses and compute a range and
  angle of arrival.
• Range data, angle data, and absolute orientation
  are broadcast N hops away.
• Based on this table of ranges, angles, and
  orientations, each node applies a
  multi-lateration algorithm with iterative outlier
  rejection to compute a consistent coordinate
  system.

23
In-depth Cane-toad Monitoring

• Joint work with colleagues at
• University of New South Wales, Australia

24
Figures of Cane Toad
Cane Toads Distribution in Australia (2003)
25
Objective
In-expensive real-time monitoring system (set up
and maintenance cost) to detect Cane toads and
their impact (Presence and Area)
26
Detecting Frogs by Their Calls

• Acoustic features can be used to distinguish the
  vocalizations of different amphibians. (call
  rate, call duration, amplitude-time envelope,
  waveform periodicity, pulse-repetition rate,
  frequency modulation, frequency and spectral
  patterns.)

Frog 1
Frog 2
Frog 3 (Cane toad)
Waveform Figures of Three Different Frogs Calls
27
How Our System Works

• Input acoustic signal is converted into a
  spectrogram of time-frequency pixels by a Fast
  Fourier Transform (FFT) algorithm.

• Our system examines each slice of the
  spectrogram (1 millisecond) and tries to estimate
  frequency local peaks.

• Frog species are identified based on the
  comparisons of these frequency local peaks with
  some classifiers.

• Quinlans machine learning system, C4.5 used
  to build classifiers.

Frog 1
Frog 2
Frog 3 (Cane toad)
Spectrogram Figures of Three Different Frogs
Calls
28
Application Challenges on Device Resources

• Very High Frequency Sampling (gt 10 KHZ, the rule
  of double the highest frequency)

Machine Learning
Acoustic Signal Processing
29
Hybrid Architecture
Motivation Increased sensing coverage at
comparable cost
30
Design Features

• In-network Reasoning

Achieve (Very) High Sampling Rate in Mica motes
through sampling scheduling
Acoustic Signal of a frogs call collected
from the field (Top). The same signal after
compression and decompression (Bottom) .
Compression and noise-reduction.
31
The Future Participatory Sensor Networks

• Sensor networks for urban applications will form
  the next tier of the Internet
• Leverage Cell phone installed base of acoustic
  and image sensors
• Using internet search, blog, and personal feeds,
  along with automated location tags, to achieve
  context, and in network processing for privacy
  and personal control
• Source Deborah Estrin, UCLA
• Source David Culler Berkeley

32
For more information

• Wireless Sensor Networks A Systems
  Perspective, Nirupama Bulusu and Sanjay Jha
  (editors),
• Artech House, Norwood, MA,
• August 2005.