Directional sensor control for maximizing information gain

Shankarachary Ragi, Hans Mittelmann, Edwin K P Chong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

We develop tractable solutions to the problem of controlling the directions of 2-D directional sensors for max- imizing information gain corresponding to multiple targets in 2-D. The target locations are known with some uncertainty given by a joint prior distribution (Gaussian). A sensor generates a (noisy) measurement of a target only if the target lies within the field-of-view of the sensor, and the measurements from all the sensors are fused to form global estimates of target locations. This problem is hard to solve exactly|the computation time increases exponentially with the number of sensors. We develop heuristic methods to solve the problem approximately and provide lower and upper bounds on the optimal information gain. We improve the solutions from these heuristic approaches by formulating the problem as a dynamic programming problem and solving it using a rollout approach.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
Volume8857
DOIs
StatePublished - 2013
EventSignal and Data Processing of Small Targets 2013 - San Diego, CA, United States
Duration: Aug 28 2013Aug 29 2013

Other

OtherSignal and Data Processing of Small Targets 2013
CountryUnited States
CitySan Diego, CA
Period8/28/138/29/13

Fingerprint

Information Gain
Sensor
Target
sensors
Sensors
heuristic methods
dynamic programming
Heuristic methods
Gaussian distribution
Heuristic Method
Prior distribution
Field of View
Dynamic programming
normal density functions
Joint Distribution
field of view
Dynamic Programming
Upper and Lower Bounds
Heuristics
Uncertainty

Keywords

  • Directional sensor control
  • maximizing information gain
  • rollout on heuristic methods

ASJC Scopus subject areas

  • Applied Mathematics
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics

Cite this

Ragi, S., Mittelmann, H., & Chong, E. K. P. (2013). Directional sensor control for maximizing information gain. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 8857). [88570J] https://doi.org/10.1117/12.2022451

Directional sensor control for maximizing information gain. / Ragi, Shankarachary; Mittelmann, Hans; Chong, Edwin K P.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8857 2013. 88570J.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ragi, S, Mittelmann, H & Chong, EKP 2013, Directional sensor control for maximizing information gain. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 8857, 88570J, Signal and Data Processing of Small Targets 2013, San Diego, CA, United States, 8/28/13. https://doi.org/10.1117/12.2022451
Ragi S, Mittelmann H, Chong EKP. Directional sensor control for maximizing information gain. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8857. 2013. 88570J https://doi.org/10.1117/12.2022451
Ragi, Shankarachary ; Mittelmann, Hans ; Chong, Edwin K P. / Directional sensor control for maximizing information gain. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 8857 2013.
@inproceedings{bb3fc70156a94c23a56f5444d8b29f79,
title = "Directional sensor control for maximizing information gain",
abstract = "We develop tractable solutions to the problem of controlling the directions of 2-D directional sensors for max- imizing information gain corresponding to multiple targets in 2-D. The target locations are known with some uncertainty given by a joint prior distribution (Gaussian). A sensor generates a (noisy) measurement of a target only if the target lies within the field-of-view of the sensor, and the measurements from all the sensors are fused to form global estimates of target locations. This problem is hard to solve exactly|the computation time increases exponentially with the number of sensors. We develop heuristic methods to solve the problem approximately and provide lower and upper bounds on the optimal information gain. We improve the solutions from these heuristic approaches by formulating the problem as a dynamic programming problem and solving it using a rollout approach.",
keywords = "Directional sensor control, maximizing information gain, rollout on heuristic methods",
author = "Shankarachary Ragi and Hans Mittelmann and Chong, {Edwin K P}",
year = "2013",
doi = "10.1117/12.2022451",
language = "English (US)",
isbn = "9780819497079",
volume = "8857",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - Directional sensor control for maximizing information gain

AU - Ragi, Shankarachary

AU - Mittelmann, Hans

AU - Chong, Edwin K P

PY - 2013

Y1 - 2013

N2 - We develop tractable solutions to the problem of controlling the directions of 2-D directional sensors for max- imizing information gain corresponding to multiple targets in 2-D. The target locations are known with some uncertainty given by a joint prior distribution (Gaussian). A sensor generates a (noisy) measurement of a target only if the target lies within the field-of-view of the sensor, and the measurements from all the sensors are fused to form global estimates of target locations. This problem is hard to solve exactly|the computation time increases exponentially with the number of sensors. We develop heuristic methods to solve the problem approximately and provide lower and upper bounds on the optimal information gain. We improve the solutions from these heuristic approaches by formulating the problem as a dynamic programming problem and solving it using a rollout approach.

AB - We develop tractable solutions to the problem of controlling the directions of 2-D directional sensors for max- imizing information gain corresponding to multiple targets in 2-D. The target locations are known with some uncertainty given by a joint prior distribution (Gaussian). A sensor generates a (noisy) measurement of a target only if the target lies within the field-of-view of the sensor, and the measurements from all the sensors are fused to form global estimates of target locations. This problem is hard to solve exactly|the computation time increases exponentially with the number of sensors. We develop heuristic methods to solve the problem approximately and provide lower and upper bounds on the optimal information gain. We improve the solutions from these heuristic approaches by formulating the problem as a dynamic programming problem and solving it using a rollout approach.

KW - Directional sensor control

KW - maximizing information gain

KW - rollout on heuristic methods

UR - http://www.scopus.com/inward/record.url?scp=84886997452&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84886997452&partnerID=8YFLogxK

U2 - 10.1117/12.2022451

DO - 10.1117/12.2022451

M3 - Conference contribution

AN - SCOPUS:84886997452

SN - 9780819497079

VL - 8857

BT - Proceedings of SPIE - The International Society for Optical Engineering

ER -