Communication Support for NASA Enterprises:
Sensor Management and Routing for Sensor Networks
Faculty: Michael Hadjitheodosiou, John Baras Student: Yingyong Chen
• Sensors activate themselves with certain probability such
that expected value of number of active sensors equals to
the number required by the application.
0.4
Probability of undercoverage
Probability-based SAP
0.3
0.2
0.1
Number of sensors in
the region |>|= 20
• Pros: simple implementation, fully decentralized; no
communication overhead.
0.0
0
2
4
6
8
10
12
14
16
18
20
Minimum number of active sensors required, S
Puc
Vs. S for SAP-p
1.0
Probability of undercoverage
• Cons: no guarantee that enough number of sensors will be
active, measured by probability of undercoverage (Puc)
which is defined as the probability that number of active
sensors are fewer than required.
0.9
A Sample Target-tracking Wireless sensor
Network (WSN)
• Wireless sensor networks often feature high node
density and redundancy.
• High redundancy can lead to excessive flow of
redundant data in network resulting in waste of
precious bandwidth and energy.
• Redundancy can be exploited to prolong network
longevity.
|>V
=20, S=10
|>V
=16, S=8
|>V
=10, S=5
|>V
=6, S=3
|>V
=4, S=2
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Sensor activation probability, p
Declaration-based SAP
Puc Vs. activation probability for SAP-p
100
SAPp0
• Nodes activates itself by broadcasting an Activation Control
Message (ACM) after a random delay.
NO SAP
• Only the first S nodes which successfully transmitted their
ACMs will be activated.
60
40
20
• Pros: low Puc with proper parameter values.
• Cons: Introduce communication overhead
SAPdb
80
Network Lifetim
Motivation
0.8
0
1
2
3
4
5
Number of active sensors required, S
Network lifetime using different SAPs
2.5%
SAP-db (0,1,1)
• All are particularly important in challenging
environments such as planetary exploration.
• Developed two decentralized lightweight senor
activation protocols (SAPs) which activate sensors
based on local network redundancy information while
trying to meet the application’s quality of service
requirements.
• Proposed a new set of link metrics which can directly
incorporate network redundancy information into route
selection.
1.5%
1.0%
0.5%
0.0%
0
Link metric:
E (0)
Ci j (t ) = [1 + SIi (t )]a [ i ]b [erx + etx (i, j )]g
Ei (t )
Notations:
S: Number of active nodes required by the application
1
2
3
4
Number of active sensors required
5
Control overhead for SAP-db
100
Network lifetime
Achievements
Redundancy-Aware Routing
SAPdb Overhead
2.0%
90
(0,1,1)-SAPdb
80
(1,0,1)-SAPdb
70
(0,0,1)-SAPdb
60
(0,0,0)-SAPdb
50
40
30
20
10
N: Number of nodes available in the region
0
1
Degree of Redundancy: DoR=N/S; Strategic-Importance: SI=1/DoR
2
3
4
5
Number of active sensors required
Network lifetime using different link metric