Grant: Discovery Grants Program - Individual
University of Saskatchewan
$15,000.00
Heterogeneous sensor network deployment and monitoring for scaler fields
FY 2012-13
Summary
Spurred by the growth of wireless networking, enabled by advanced in MEMS design, and inspired by robotic platforms like Spirit and Opportunity, heterogeneous sensor networks have moved to the forefront of many research agendas. In heterogeneous sensor networks, the hardware of the nodes or motes is not uniform, and classified by power consumption, computational capacity and mobility. Homogeneous networks achieve spatial resolution through high mote densities, while mobile heterogeneous sensor networks (MHSN) employ fewer more expensive mobile motes to reduce the node density of the network.
While even a randomly deployed sensor network will generate more data than a single sensor, distributions which are based on the expected behavior of the area to be measured are more effective than distributions which are deployed blindly. However, this raises a dilemma: how do you estimate the expected behavior of the system you are about to measure? Presumably if detailed information on the system is available, there is no need for a sensor network to make measurements. Recently, researchers have found that using statistical representations which estimate how one sensor's measurements correlate to one another; they can deploy nodes with more confidence. This technique can be extended to a sequential deposition of sensors, building the sensor network piece by piece, or destination targets for mobile robots to rove from point to point, making measurements.
This research proposal aims to add levels of sophistication to MHSN deployment strategies, which will allow the integration of physical models, the management of complex multivariate data and the inclusion of human experts in the measurement loop. It will balance the additional implementation complexity against maintaining the statistical rigor of simpler techniques. By adding physical models, multivariate representations and human experts, MHSNs can be developed into powerful tools for agricultural and environmental monitoring.
Program
Discovery Grants Program - Individual
Recipient
Location
Saskatchewan, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
356043-2008
Name nom
Stanley, Kevin
Department département
Computer Science
Organization id
16
Institution établissement
University of Saskatchewan
Province en
Saskatchewan
Province fr
Saskatchewan
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2008
Award amount
15000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
331
Committee name en
Computing and Information Sciences - B
Committee name fr
Informatique - B
Area of application code
1209
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Agriculture
Area of application fr
Agriculture
Research subject code
2600
Research subject group en
Robotics
Research subject group fr
Robotique
Research subject en
Robotics
Research subject fr
Robotique
Partie
5
Nb partie
5
Application title
Heterogeneous sensor network deployment and monitoring for scaler fields
Keyword
NSERC
Application summary
Spurred by the growth of wireless networking, enabled by advanced in MEMS design, and inspired by robotic platforms like Spirit and Opportunity, heterogeneous sensor networks have moved to the forefront of many research agendas. In heterogeneous sensor networks, the hardware of the nodes or motes is not uniform, and classified by power consumption, computational capacity and mobility. Homogeneous networks achieve spatial resolution through high mote densities, while mobile heterogeneous sensor networks (MHSN) employ fewer more expensive mobile motes to reduce the node density of the network.
While even a randomly deployed sensor network will generate more data than a single sensor, distributions which are based on the expected behavior of the area to be measured are more effective than distributions which are deployed blindly. However, this raises a dilemma: how do you estimate the expected behavior of the system you are about to measure? Presumably if detailed information on the system is available, there is no need for a sensor network to make measurements. Recently, researchers have found that using statistical representations which estimate how one sensor's measurements correlate to one another; they can deploy nodes with more confidence. This technique can be extended to a sequential deposition of sensors, building the sensor network piece by piece, or destination targets for mobile robots to rove from point to point, making measurements.
This research proposal aims to add levels of sophistication to MHSN deployment strategies, which will allow the integration of physical models, the management of complex multivariate data and the inclusion of human experts in the measurement loop. It will balance the additional implementation complexity against maintaining the statistical rigor of simpler techniques. By adding physical models, multivariate representations and human experts, MHSNs can be developed into powerful tools for agricultural and environmental monitoring.