A vehicle to vehicle collision warning system.....a presentation done by SAMRIDDHA SHIL & MY TEAM.....

1. Presented by: 1)SANTANU NASKAR
2)SOUMYAJIT SINGHA
3)SAMRIDDHA SHIL
4)TANAY PAUL
5)SAMIMUL ISLAM
A VEHICLE-TO-VEHICLE
COLLISION WARNING SYSTEM

2. AGENDA

Motivation for the Project

Requirements of Collision Avoidance System

Existing Approaches

Proposed Architecture

Work Flow in Detail

Improving Vehicle to Roadside connectivity

Collision Detection Mechanism

Expected Results

Future Work

3. MOTIVATION
There were nearly 6,420,000 auto accidents in the
United States in 2005
The financial cost of these crashes was more than 230
billion dollars
6858 people were injured in road accidents in 2006
60% of collisions could be avoided given at least 0.5
sec warning

4. MOTIVATION
• Stop sign and Intersection Violation
• Left or Right turn at Intersection crashes

5. MOTIVATING EXAMPLE
Scenario: A, B, and C traveling in same direction. A suddenly brakes.
Being farther from A does not make C safer for 2 reasons:
Line of sight limitation
Large human processing/forwarding delay (reaction time)
Can we build a quicker warning system using
vehicle-to-vehicle (V2V) communication?

6. COOPERATIVE COLLISION
WARNING
Scenario: A, B, and C traveling in same direction. A suddenly brakes.
Using V2V the danger for all parties is alleviated:
A can send warning messages immediately once emergency is detected
Assuming little delay, B and C can receive the alerts and react

7. REQUIREMENTS FOR
COLLISION AVOIDANCE
SYSTEM

Critical response time for Alerts

Reliable data transfer

Energy efficiency

Accurate Speed Measurement

Security

Integrity and Authenticity

8. INTERSECTION CRASH
STATISTICS

9. EXISTING APPROACHES TO
COLLISION AVOIDANCE

Vision based system

Deploy Wireless sensor Network(WSN) on the roads

10. EXISTING APPROACHES TO
COLLISION AVOIDANCE
Co-operative Vehicle to Vehicle (V2V)

11. PROPOSED SYSTEM
ARCHITECTURE
•Combination of V2V and V2R approach
•DSRC- Dedicated Short Range Communication

12. END TO END WORKING
• RSU connects with OBU by WSA
(Wave Service Advertisement)
• Cars near Intersection connects
RSU by Control CH
• Cars communicate with RSU by
service channel
• Separate RSU for Control and
Service channel depending on the
arrival rate of the vehicles.
OBU
RSU
OBU
RSU
OBUCCHZ
SCHZ

13. STATE TRANSITION DIAGRAM
Vehicles connect to
RSU by Control
Channel
Vehicle
enters
Service
Channel
range
RSU adds vehicle
to polling list to
query status
RSU
sends
polling
message
OBU replies by sending
Status message eg:
Velocity,position
Broadcast to other
Vehicles by
RSU
RSU send its
Status after
Every T sec
RSU sends
warning
Message to
OBU for
Collision

14. Proposed Architecture to improve
Vehicle to RSU connectivity

15. SYSTEM MODEL
• The Roadside Unit (RSU) acts
asWAVE providers that keep
advertising their presence and
the offered services through
periodic broadcasts.
• WSA(Wave Service
Advertisements)
Control information sent by the
RSUs over CCHs

16. SYSTEM MODEL CONTD.
• WBSS(Wave-based Basic Service Set)
Set up after WSAs are sent
Data exchange over the SCHs can only occur after the vehicle
successfully receives the WSA
• The signal strength of the RSU should be tuned to the
network latency and the lane speed at that location, so as to
not miss any vehicles coming in its range
• To relieve the RSU of too much computation and keep it real-
time, the Vehicle is expected to send the GPS location using
the DSRC Service Channel.
• This channel is exclusively reserved for the GPS data from
vehicle and is informed of this and is tuned to sufficient
bandwidth.

17. SYSTEM MODEL CONTD.
• To avoid the shadowing effect sometimes when a larger
building or a larger vehicle shadows other vehicles, we use
the mechanism of piggybacking the WAVE based parameters
which would further reduce the communication gap between
the Vehicles and the RSUs.
• Piggybacking
Periodic short status messages(Beacons) used to transmit
WSAs
There is a chance that WSA information sent is missed by a
vehicle
For vehicles that miss a WSA, beacons that are piggybacked
with WSA fields are sent

19. COMMUNICATION MODEL FOR
COLLISION AVOIDANCE
Different Communication Model exist:
1) Based on Vehicles:
General warning :broadcast message to all vehicles in range(nearby accident
information, road condition, etc.)
Selective Warning: Message sent only to the affected vehicle(expecting
collision)
2) Based on Warning Initiator:
Push Method: Here the Vehicle will send its current data including
direction, speed , acceleration to the RSU for constant time interval
Pull Method: Here the RSU will initiate the status message by asking for
the individual vehicle information.
We Use Selective warning with push method since this will avoid extra
message overhead involved in pull method. Also, we are interested only in
collision avoidance.

20. SCENARIO FOR ISSUING
WARNING
• There should be a good threshold in the system to prevent false
alarms.
• Time-To-Collision (TC) is computed by the pair-wise collision
detection algorithm while Time-To- Avoidance (TA) in Miller and
Huang’s peer-to-peer collision warning system is computed based on
vehicle kinetics, network latency, and human response time. This
holds good for v2v communication.
• If TC>TA, then warning is not issued, however if TC~TA ,then the
nearby RSU issues warning to the driver.
• There exists a tradeoff between providing drivers valuable
information in time and avoiding distraction to driver due to huge
number of messages.

21. MINIMUM WARNING DISTANCE
• The minimum warning distance required to inform a driver before intersection
crossing is calculated using the below parameters:
• [1]
v0 - velocity of the vehicle
a –vehicle braking deceleration,
tdriver -driver’s response time to brake,
tmachine - combination of braking system and warning system response time and
tinformation -constant information time, which is a time determined by the
assistance system to allow the driver to react and prepare the driver to stop

22. RSU MODEL
The communication between
the RSU and vehicle agents
are regulated inside the
administration zone which is
the spatial domain that
determines the region of
authority of an intersection
agent to coordinate vehicle
agents in the approaching and
passing vehicles.

23. COMMUNICATION AND
COMPUTATION
• Initially, the Vehicle pushes its status information to the RSU
unit nearby.
• Next, the RSU asks for a register request and assigns a unique
id to the vehicle. This will be required only for security
architecutre.
• Finally the RSU sends collision message to the vehicle if
needed based on its own computation.
• Computation can be performed by pairwise detection or
preselection.
• In pairwise, the time for each car to reach the future collision
point is calculated with velocity of each car and r is the vector
of co-ordinate of car(x,y) and the size of the vehicle.This has
high computational cost.

24. CHALLENGES IN SIMULATION
Introducing Beaconing support in DSRC when simulating the
scenario’s using VANETMobiSIM and NS-2. There exists no
parameter to specify Beaconing support in NS-2.
Handling highly mobile nodes which don’t stay for longer
timeframe for the network to stabilize the topology.
Extracting the results from huge trace files produced by NS-2
to evaluate our metrics requires complex awe parsers to
process the necessary information.
Simulating shadowing scenarios.

25. RESULTS (Expected)
• The solution is fully compliant with 802.11p/WAVE
specifications and incurs little-to-none over-head, by
leveraging on packets already foreseen to be transmitted on the
Control Channel and on largely available self-positioning
capabilities of vehicles.
• The Dedicated Control Channel in the DSRC is highly secured
and is responsible for safely delivering the Alert messages to
the Vehicles.
• The Service channels apart from hosting WBSS (WAVE Based
Basic Service Set) can be used to communicate the Vehicle
Location effectively to the RSUs thereby reducing further
overhead.

26. FUTURE WORK
• Evaluate the metrics in the presence of Obstructions.
• Evaluate the metrics based on realistic Vehicular traces
available
• Make use of Simulators to simulate much more complex
Urban traffic scenarios and investigate the Network
parameters using NS-2.
• Try to simulate shadowing effects using Simulators as well
and evaluate the performance.

27. Thank you!!