Amit Saha April 13th, 2005

1
Amit SahaApril 13th, 2005
VANET VEHICULAR AD HOC NETWORKS
2
What is a VANET ?

• Ad hoc network composed of vehicles
• Individual nodes different from traditional
  wireless nodes
• No power constraint
• Nodes mostly mobile
• Complements existing infrastructure
• Extends existing infrastructure

3
Medium Access Control

• Dedicated Short Range Communication (DSRC)
• FCC regulated 5.9 GHz
• 10 MHz channels
• 1 control, 6 service
• MAC similar to 802.11, i.e., CSMA/CA
• OFDM used in PHY
• Speeds up to 120 mph
• Up to 1000 m transmission range

4
VANET 2004

• First conference of its kind
• 9 papers, 7 posters
• Safety applications (3)
• Security (2)
• Data dissemination (4)

5
Sensors in VANET Nodes

• Size and power no longer issues

6
Applications

• Traffic control
• Automatic speed limit enforcement
• Rerouting in traffic congestion
• Safety
• Notification of accident up ahead
• Decentralized 911 service
• Prioritized over non-safety applications
• Extended communication
• Infotainment

7
Snowplowing Project

• Project by Berkeley PATH
• Snowplowing in total whiteout
• Vehicle equipped with sensors for
• Measuring steering angle and vehicle motion
• Field of magnetic markers on roadway
• Obstacle detection (radar sensors)

8
Security Aspects

• Somewhat similar to sensor networks
• Sensors might report incorrect data
• Undetected error
• Detected error
• Corrected error

9
Data Collection

• No aggregation
• No hierarchy
• Simple model for data collection

10
Data Delivery

• Opportunistic forwarding
• Disconnection possible
• Geographic forwarding
• Trajectory based forwarding
• A coordinate system is assumed
• Nodes know position in this coordinate system
• Route is a trajectory in this coordinate system

11
Conclusion

• Similar to mobile ad hoc networks
• Has sensors but usage is mostly different
• Accident notification similar to fire !!! in
  sensor networks
• Not power or size constrained

12
Amit SahaApril 13th, 2005
MOBILITY MODELS
13
Different Mobility Models

• Random waypoint model
• City section model
• Obstacle-based model

14
Random Waypoint Model

• Convergence issues
• Nodes tend to converge to center
• Wrap around
• Reflection
• Average speed
• Minimum speed has to be greater than zero
• Many variations exist

15
Example
16
City Section Model

• Simulates movement of vehicles on streets
• Gets very complicated to make realistic
• Lane change
• Traffic light modeling
• Widely used versions are simplistic
• Grid structure

17
Obstacle-based Model

• Models obstacles
• Restricts movement
• Nodes do not reflect off surfaces
• Restricts wireless propagation
• Doors used to cross buildings
• Movement on Voronoi Graph

18
Example
19
Problems

• Steady state unknown
• Existence is not known as well
• When is steady state reached in simulation
• Infocom 2005 best paper, EPFL, Lausanne
• Conditions required for steady state behavior
• Several mobility models
• Steady state initialization