Decentralized quickest change detection

Research output: Contribution to journalLetter

Abstract

A decentralized formulation of the quickest change detection problem is studied, where the distributions of the observations at all of the sensors in the system change at the time of disruption, and the sensors communicate with a common fusion center. A Bayesian setting is considered in which a priori knowledge of the change time distribution is available. The observations are assumed to be independent from sensor to sensor, conditioned on the change hypothesis. An optimal solution to the problem is derived under a quasi-classical information structure, where each sensor retains only its messages from the past (restricted local memory), and receives feedback from the fusion center about the past messages of the other sensors (full feedback). A technique for implementation of the optimal solution is given, and the solution is extended to the situation where a priori change time distribution information is not available. The structure of the optimal solution is then used to arrive at a simple suboptimal policy that does not require any past message information. Numerical examples are given, which illustrate that the optimal solution offers little improvement over the suboptimal one, i.e., that feedback from the fusion center cannot be exploited to improve performance.

Original languageEnglish (US)
Pages (from-to)1657-1665
Number of pages9
JournalIEEE Transactions on Information Theory
Volume47
Issue number4
DOIs
StatePublished - May 1 2001

Fingerprint

Sensors
Fusion reactions
system change
Feedback
knowledge
performance
time
Data storage equipment

Keywords

  • Bayes problem
  • Detection
  • Distributed decision-making
  • Monitoring
  • Multisensor systems
  • Quickest change

ASJC Scopus subject areas

  • Information Systems
  • Electrical and Electronic Engineering

Cite this

Decentralized quickest change detection. / Veeravalli, Venugopal Varadachari.

In: IEEE Transactions on Information Theory, Vol. 47, No. 4, 01.05.2001, p. 1657-1665.

Research output: Contribution to journalLetter

@article{4078c73914aa4d4c93a865f5df533a32,
title = "Decentralized quickest change detection",
abstract = "A decentralized formulation of the quickest change detection problem is studied, where the distributions of the observations at all of the sensors in the system change at the time of disruption, and the sensors communicate with a common fusion center. A Bayesian setting is considered in which a priori knowledge of the change time distribution is available. The observations are assumed to be independent from sensor to sensor, conditioned on the change hypothesis. An optimal solution to the problem is derived under a quasi-classical information structure, where each sensor retains only its messages from the past (restricted local memory), and receives feedback from the fusion center about the past messages of the other sensors (full feedback). A technique for implementation of the optimal solution is given, and the solution is extended to the situation where a priori change time distribution information is not available. The structure of the optimal solution is then used to arrive at a simple suboptimal policy that does not require any past message information. Numerical examples are given, which illustrate that the optimal solution offers little improvement over the suboptimal one, i.e., that feedback from the fusion center cannot be exploited to improve performance.",
keywords = "Bayes problem, Detection, Distributed decision-making, Monitoring, Multisensor systems, Quickest change",
author = "Veeravalli, {Venugopal Varadachari}",
year = "2001",
month = "5",
day = "1",
doi = "10.1109/18.923755",
language = "English (US)",
volume = "47",
pages = "1657--1665",
journal = "IEEE Transactions on Information Theory",
issn = "0018-9448",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Decentralized quickest change detection

AU - Veeravalli, Venugopal Varadachari

PY - 2001/5/1

Y1 - 2001/5/1

N2 - A decentralized formulation of the quickest change detection problem is studied, where the distributions of the observations at all of the sensors in the system change at the time of disruption, and the sensors communicate with a common fusion center. A Bayesian setting is considered in which a priori knowledge of the change time distribution is available. The observations are assumed to be independent from sensor to sensor, conditioned on the change hypothesis. An optimal solution to the problem is derived under a quasi-classical information structure, where each sensor retains only its messages from the past (restricted local memory), and receives feedback from the fusion center about the past messages of the other sensors (full feedback). A technique for implementation of the optimal solution is given, and the solution is extended to the situation where a priori change time distribution information is not available. The structure of the optimal solution is then used to arrive at a simple suboptimal policy that does not require any past message information. Numerical examples are given, which illustrate that the optimal solution offers little improvement over the suboptimal one, i.e., that feedback from the fusion center cannot be exploited to improve performance.

AB - A decentralized formulation of the quickest change detection problem is studied, where the distributions of the observations at all of the sensors in the system change at the time of disruption, and the sensors communicate with a common fusion center. A Bayesian setting is considered in which a priori knowledge of the change time distribution is available. The observations are assumed to be independent from sensor to sensor, conditioned on the change hypothesis. An optimal solution to the problem is derived under a quasi-classical information structure, where each sensor retains only its messages from the past (restricted local memory), and receives feedback from the fusion center about the past messages of the other sensors (full feedback). A technique for implementation of the optimal solution is given, and the solution is extended to the situation where a priori change time distribution information is not available. The structure of the optimal solution is then used to arrive at a simple suboptimal policy that does not require any past message information. Numerical examples are given, which illustrate that the optimal solution offers little improvement over the suboptimal one, i.e., that feedback from the fusion center cannot be exploited to improve performance.

KW - Bayes problem

KW - Detection

KW - Distributed decision-making

KW - Monitoring

KW - Multisensor systems

KW - Quickest change

UR - http://www.scopus.com/inward/record.url?scp=0035334718&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035334718&partnerID=8YFLogxK

U2 - 10.1109/18.923755

DO - 10.1109/18.923755

M3 - Letter

VL - 47

SP - 1657

EP - 1665

JO - IEEE Transactions on Information Theory

JF - IEEE Transactions on Information Theory

SN - 0018-9448

IS - 4

ER -