Emergency Navigation by Wireless Sensor Networks in 2D and 3D Indoor Environments

1
Emergency Navigationby Wireless Sensor Networks
in 2D and 3D Indoor Environments

• Yu-Chee Tseng
• Deptment of Computer Science
• National Chiao Tung University

2
Outline

• Introduction
• System Overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

3
Outline

• Introduction
• System Overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

4
Introduction

• Wireless Sensor Network
• Each sensor has
• Limited Memory?Limited CPU?Wireless
  Transceiver?Sensing Unit
• Each sensor can
• Sense environments
• Communicate with others
• Do simple computations

5
Introduction

• Traditional Navigation Devices
• Advantage
• Cheap
• Easy deployment
• Disadvantage
• Fixed direction.
• Can not adapt to actual emergency situations.

6
Introduction

• Motivation
• According to the statistic report of the NFA of
  Taiwan(??????), 228 people died in fire accidents
  in 2003.
• The main reason is that people can not find
  right escaping paths to exits.
• Our Goal
• to develop an emergency navigation system
• for indoor 2D and 3D environments

7
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

8
System Overview

• Our system is composed of 3 parts
• Environment setting
• Regular reporting
• Emergency Navigation
• Two network graphs
• Communication graph and guidance graph

Communication graph
Guidance graph
9
Environment Setting

• Deploy sensors
• Construct reporting tree
• Setup initial navigation paths

10
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

11
Deployment of Sensors

• Plan locations of sensors
• Define the roles of sensors
• Sink
• Exit sensors
• Normal sensors
• Decide navigation links

navigation links (for human)
12
Construct a Reporting Tree

• Step 1. Discover symmetric links
• Each sensor periodically broadcasts HELLOs
• When receiving a HELLO, sensors reply ACKs
• After receiving an ACK, sensors record the sender
  ID in its link table

HELLO
2
ACK
ACK
0
1
3
ACK
Link table
2 3
13
Construct a reporting tree (cont.)

• Step 2. Construct a spanning tree
• Sink floods a BEACON.
• For a sensor receives a BEACON, it checks if the
  sender is in its link table
• If yes, it sends a REG(ister) to sink and
  rebroadcasts BEACON.
• Else, drops it

BEACON
REG
BEACON
14
communication links (for packets)
15
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

16
Reporting Issues

• How often a report should be sent?
• Will each sensor report individually?
• Is there any inaccuracy?
• False alarm?
• How to save energy of sensors?

17
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation in 2D environment
• Simulation results
• Demonstration
• Conclusion

18
Design Principle

• When a sensor detects an emergency event, it
  forms a hazardous region
• The navigation algorithm will try to guide people
  as farther away from hazardous regions as possible

19
Problem Formulation

• Each sensor has an altitude.
• Sensors in hazardous regions will raise their
  altitudes.
• Each sensor guides people to the neighbor with
  the lowest altitude
• After forming hazardous regions, some sensors may
  become local minimum ones
• A partial link reversal operation is performed to
  solve this problem

20
Phases of Navigation

• Initialization phase
• Initial phase is started by Exit sensor
• After this phase, every sensor has a default
  guiding direction.
• Navigation phase
• This phase starts by the sensor which detects an
  emergency event.

21
Terminology

• DThe radius of the hazardous region
• Aemg A large constant which represents the
  maximum altitude
• AiThe altitude of sensor i
• IiThe altitude obtained in the initialization
  phase
• ej,iThe hop count from emergency sensor j to
  sensor i

22
Initialization phase

• Every exit sensor sets its altitude to 0 and
  broadcasts an initialization packet.
• When receiving an initialization packet, a sensor
  adds its hop count by 1.
• Then, it compares the hop count with its current
  altitude

8
8
8
0
8
8
8
8
8
8
23
Initialization phase (cont.)

• If the hop count is smaller than its altitude, it
  resets its altitude and setups its initial
  guiding direction to that sender.
• Then, it rebroadcasts this packet.

8
8
0
1
2
8
8
8
1
2
3
8
8
8
2
3
4
24
Navigation phase

• When a sensor x detects an emergency, it will set
  its altitude to the maximum altitude Aemg (let it
  be 200).
• Then it broadcasts an emergency packet EMG(seq,
  x, x, Aemg, 0)
• seqsequence number
• xemergency ID
• w sender ID
• Awaltitude of sender
• hhop count to emg. location

10
11
12
11
12
13
200
12
13
14
25
Navigation phase (cont.)

• When a sensor node y receives a EMG packet
  originated from node x, it will do the following
  steps.
• Step1
• Decide that the emergency is a new one or not
• If its a new emergency, record this event and
  set the hop count ex,y to h1.
• Else, compare the h and ex,y. If h is smaller
  than ex,y , set ex,y to h1.
• Record the altitude (Aw) in the navigation link
  table.

10
11
12
13
11
200
12
13
14
26
Navigation phase (cont.)

• Step 2
• If eX,Y was changed in step1 and eX,Y ?D, y
  considers itself within hazardous region. Then it
  re-calculates its altitude as follows

10
11
12
61
13
11
200
61
63
12
13
14
63
27
Navigation phase (cont.)

• Step 3
• If y has a local minimum altitude and its not an
  exit, it must adjust its altitude as follows
• altitudes of ys neighbors
• STA standard deviation
• A bigger value means closer to the hazardous
  region. So we need to adjust the altitude faster.
• Ny number of neighbors of y.
• A smaller Ny means less escape ways. So we
  need to adjust the altitude faster.
• dis a small constant.

Static adjustment
61
12
10
Five iterations
200
61
63
Our scheme
Three iterations
63
12
14
63.1
28
Navigation phase (cont.)

• Step 4
• y has to broadcast an EMG(seq, x, y, Ay, ex,y)
  packet if any of the following conditions
  matches.
• Its a new emergency
• y has changes its altitude or ex,y in the
  previous steps.
• Step 5
• If y is in hazardous regions and it sees an exit
  sensor which is in Ny and which is also in
  hazardous regions, then y chooses this exit
  sensor
• In all other cases, y directs users to a safer
  sensor first, and then gradually to a safe exit.

29
ExampleAltitude after initial phase
Exit
10x10 Grid Network
30
One emergency event after step 1, 2 4
Local minimum
31
One emergency eventfinal result
32
Two emergency eventsafter step 1, 2 4
Local minimum
33
Two emergency eventsfinal result
34
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

35
Simulation results

• We compare our navigation algorithm with
  Distributed algorithm for guiding navigation
  across a sensor network (MobiCom 03)
• This algorithm guides people to the nearest exits
• However, nearest exits may not be good choices

36
Simulation results

• Case1. Our algorithm will choose to pass
  hazardous region areas as farther away from
  emergency locations as possible.
• Case2. Our algorithm will not guide people
  passing through the hazardous region.
• Case3. Only the sensors near the exit in the
  hazardous region will guide people to that exit.

37
Outline

• Introduction
• System overview
• Environment setting
• Regular report
• Emergency navigation service
• Simulation results
• Demonstration
• Conclusion

38
Demonstration

• System Components
• MICAz sensors
• Environment monitoring
• Navigation
• Sink
• MIB510 serial Gateway
• Gateway between wireless sensor network and PC
• PC
• Control Host

39
Demonstration
second event (emergency time)
first event (emergency time)
exit (normal time)
40
A Short Summary (2D)

• Novel indoor monitoring and navigation services
  based on wireless sensor network technolgoies
• emergency will raise sensors altitudes
• navigation similar to TORA protocol, but
  different in that emergencies will disturb
  altitudes
• altitude adjustment is designed for quicker
  convergence
• navigation in emergency applications requires
  safer paths, but not necessarily longer paths

41
Emergency Navigation in Indoor 3D Environments
by Wireless Sensor Networks

• Yu-Chee Tseng
• Department of Computer Science
• National Chiao Tung University

42
Introduction

• Why 2D guiding algorithms cant directly apply to
  3D environments

Rooftop
3F
room
room
2F
room
room
room
room
room
room
2F
room
room
room
room
1F
room
room
room
room
1F
room
room
room
room
43
System Architecture
44
Guidance initialization
(1, 1)
2F
e
(1, 0)
(1, 1)
d
f
(0, 0)
(0, 1)
b
(0, 2)
(0, 1)
1F
a
(0, 2)
(0, 3)
c
45
Guidance initialization
(
3
,
0
)
(
3
,
2
)
(
3
,
2
)
(
3
,
1
)
(
3
,
1
)
room
room
(
3
,
1
)
(
3
,
1
)
(
3
,
1
)
(
3
,
0
)
(
3
,
1
)
room
room
4
F
(
3
,
1
)
(
3
,
2
)
(
3
,
1
)
(
3
,
2
)
(
3
,
0
)
(
2
,
0
)
(
2
,
2
)
(
2
,
3
)
(
2
,
2
)
(
2
,
1
)
room
room
(
2
,
1
)
(
2
,
1
)
(
2
,
2
)
(
2
,
3
)
(
2
,
2
)
room
room
3
F
(
2
,
1
)
(
2
,
2
)
(
2
,
2
)
(
2
,
1
)
(
2
,
0
)
(
1
,
0
)
(
1
,
2
)
(
1
,
3
)
(
1
,
2
)
(
1
,
1
)
room
room
(
1
,
1
)
(
1
,
1
)
(
1
,
2
)
(
1
,
3
)
(
1
,
2
)
room
room
2
F
(
1
,
0
)
(
1
,
1
)
(
1
,
2
)
(
1
,
2
)
(
1
,
3
)
(
0
,
0
)
(
0
,
2
)
(
0
,
3
)
(
0
,
2
)
(
0
,
1
)
room
room
(
0
,
1
)
(
0
,
1
)
(
0
,
2
)
(
0
,
1
)
(
0
,
2
)
room
room
1
F
(
0
,
0
)
(
0
,
1
)
(
0
,
2
)
(
0
,
0
)
(
0
,
1
)
46
Principles of 3D guidance

• A sensor is located in a hazardous region if
• it is D hop away from the emergency point or
• its a stair sensor and its downstair sensor is
  in a hazardous region
• When guiding
• Avoid to guide people through hazardous regions
• Try to guide people to the exits on the ground
  floor
• Guide people to rooftop if there is no proper
  ways to downstairs

47
Simulation results
48
Prototyping

• We have implemented our system using MICAz motes
  and MTS310 sensors on TinyOS.
• Protocol stack

49
JAVA GUI
50
Guidance UI
51
Demonstration

• Environment
• A virtual 2-store building

52
Demonstration

• Vedio

53
More Results
54
Conclusions

• Extending 2D navigation to 3D navigation
• on each floor, the navigation is similar to 2D
• stair and gateway sensors are paid of special
  attention
• roof is also paid of special attention

55
References

• Q. Li, and et. al, Distributed algorithm for
  guiding navigation across a sensor network,
  MobiCom 03.
• Y.-C. Tseng, M.-S. Pan, and Y.-Y. Tsai, A
  Distributed Emergency Navigation Algorithm for
  Wireless Sensor Networks, IEEE Computers, Vol.
  39, No. 7, July 2006, pp. 55-62.
• M.-S. Pan, C.-H. Tsai, and Y.-C. Tseng,
  Emergency Guiding and Monitoring Applications in
  Indoor 3D Environments by Wireless Sensor
  Networks, Intl Journal of Sensor Networks, Vol.
  1, Nos. 1/2, pp. 2-10, 2006.