Survey of Simulation Tools for Sensor Networks

1
Survey of Simulation Tools for Sensor Networks

• Presentation 1
• 22/05/2006
• Shaoping,Yuan

2
WSNWireless Sensor Network

• Noninteractive, data and event centric
  applications
• Application specific
• Deployed in uncontrolled, often inaccessible
• Energy usage must be carefully managed
• Closed environment, sensors rarely change
• Fixed sensors but dynamic topology
• Some need for real time delivery and time
  synchronization

3
Challenges

• Distributed processing
• Low bandwidth communication
• Large scale coordination
• Real time computation

4
Operation system support

• TinyOS
• Mantis
• Contiki
• SOS

5
Other commercial platforms

• Cube
• Phidgets

6
TOSSIM

• Emulator run actual application code
• Component-based, discrete event simulator
• Focus on TinyOS and its execution
• Simulate at bit-level
• Four key points scalability, completeness,
  fidelity, bridging
• Static simulator
• TinyViz a Java visualzation and actuation
  environment for TOSSIM

7
Problems of TOSSIM

• Probabilistic bit error model leads to
  inaccuracies and reduces the effectiveness in
  anylyzing low level protocols
• Built directly from TinyOS code
• API application programming interface

8
TOSSIM (with MULE)Mobile Ubiquitous LAN
Extensions

• Motivation sensors may be spread over a large
  geographical area resulting in a sparse network
• Possible approaches install multiple base
  stations to relay the data from sensor nodes in
  their coverage area or deploy enough low-power
  relay nodes to form a dense connected network
• Advantage large power saving
• Disadvantage increased latency
• Conclusion MULE archtecture is suitable for a
  rich set of environmental monitoring applications
  in urban environments, both in low and high
  mobility scenarios

9
OPNET

• Simulator
• Discrete event, object-oriented
• Modeling and simulatio of ultra large networks
• Hierachical model define each aspect of the
  system
• Top level network model, topology is
  designed
• Next level node model, data flow models are
  defined
• Third level process editor, handlles control
  flow models
• Fourth level parameter editor, to support
  the three higher
• layers

10
Problems of OPNET

• Scalability
• Not enough protocols available
• Only available in commercial form

11
Comparision of different tools
12
Comparision of different tools
13
Comparision of different tools
14
Comparision of different tools
15
Others

• Avrora a cycle-accuracy instruction-level sensor
  network simulator, be able to measure detailde
  time-critical phenomena
• Emstar an environment for developing wireless
  embedded system software
• SENS a sensor, environment and network
  simulator
• WSNS
• Ptolemy
• WISENES
• Sunflower

16
Reference

• www.elet.polimi.it/picco
• http//ieeexplore.ieee.org
• http//en.wikipedia.org
• http//www.binghamton.edu/