Distributed principal subspace estimation in wireless sensor networks

Lin Li, Anna Scaglione, Jonathan H. Manton

Research output: Contribution to journalArticle

42 Citations (Scopus)

Abstract

Motivated by applications in multi-sensor array detection and estimation, this paper studies the problem of tracking the principal eigenvector and the principal subspace of a signal's covariance matrix adaptively in a fully decentralized wireless sensor network (WSN). Sensor networks are traditionally designed to simply gather raw data at a fusion center, where all the processing occurs. In large deployments, this model entails high networking cost and creates a computational and storage bottleneck for the system. By leveraging both sensors' abilities to communicate and their local computational power, our objective is to propose distributed algorithms for principal eigenvector and principal subspace tracking. We show that it is possible to have each sensor estimate only the corresponding entry of the principal eigenvector or corresponding row of the $p$-dimensional principal subspace matrix and do so by iterating a simple computation that combines data from its network neighbors only. This paper also examines the convergence properties of the proposed principal eigenvector and principal subspace tracking algorithms analytically and by simulations.

Original languageEnglish (US)
Article number5719545
Pages (from-to)725-738
Number of pages14
JournalIEEE Journal on Selected Topics in Signal Processing
Volume5
Issue number4
DOIs
StatePublished - Aug 2011
Externally publishedYes

Fingerprint

Eigenvalues and eigenfunctions
Wireless sensor networks
Sensors
Sensor arrays
Covariance matrix
Parallel algorithms
Sensor networks
Fusion reactions
Processing
Costs

Keywords

  • Asynchronous time
  • average consensus
  • distributed algorithm
  • gossiping
  • signal detection
  • stochastic approximation
  • subspace estimation
  • subspace tracking
  • synchronous time

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

Distributed principal subspace estimation in wireless sensor networks. / Li, Lin; Scaglione, Anna; Manton, Jonathan H.

In: IEEE Journal on Selected Topics in Signal Processing, Vol. 5, No. 4, 5719545, 08.2011, p. 725-738.

Research output: Contribution to journalArticle

@article{acde13d3a3044146811d9d6aee3e5e13,
title = "Distributed principal subspace estimation in wireless sensor networks",
abstract = "Motivated by applications in multi-sensor array detection and estimation, this paper studies the problem of tracking the principal eigenvector and the principal subspace of a signal's covariance matrix adaptively in a fully decentralized wireless sensor network (WSN). Sensor networks are traditionally designed to simply gather raw data at a fusion center, where all the processing occurs. In large deployments, this model entails high networking cost and creates a computational and storage bottleneck for the system. By leveraging both sensors' abilities to communicate and their local computational power, our objective is to propose distributed algorithms for principal eigenvector and principal subspace tracking. We show that it is possible to have each sensor estimate only the corresponding entry of the principal eigenvector or corresponding row of the $p$-dimensional principal subspace matrix and do so by iterating a simple computation that combines data from its network neighbors only. This paper also examines the convergence properties of the proposed principal eigenvector and principal subspace tracking algorithms analytically and by simulations.",
keywords = "Asynchronous time, average consensus, distributed algorithm, gossiping, signal detection, stochastic approximation, subspace estimation, subspace tracking, synchronous time",
author = "Lin Li and Anna Scaglione and Manton, {Jonathan H.}",
year = "2011",
month = "8",
doi = "10.1109/JSTSP.2011.2118742",
language = "English (US)",
volume = "5",
pages = "725--738",
journal = "IEEE Journal on Selected Topics in Signal Processing",
issn = "1932-4553",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Distributed principal subspace estimation in wireless sensor networks

AU - Li, Lin

AU - Scaglione, Anna

AU - Manton, Jonathan H.

PY - 2011/8

Y1 - 2011/8

N2 - Motivated by applications in multi-sensor array detection and estimation, this paper studies the problem of tracking the principal eigenvector and the principal subspace of a signal's covariance matrix adaptively in a fully decentralized wireless sensor network (WSN). Sensor networks are traditionally designed to simply gather raw data at a fusion center, where all the processing occurs. In large deployments, this model entails high networking cost and creates a computational and storage bottleneck for the system. By leveraging both sensors' abilities to communicate and their local computational power, our objective is to propose distributed algorithms for principal eigenvector and principal subspace tracking. We show that it is possible to have each sensor estimate only the corresponding entry of the principal eigenvector or corresponding row of the $p$-dimensional principal subspace matrix and do so by iterating a simple computation that combines data from its network neighbors only. This paper also examines the convergence properties of the proposed principal eigenvector and principal subspace tracking algorithms analytically and by simulations.

AB - Motivated by applications in multi-sensor array detection and estimation, this paper studies the problem of tracking the principal eigenvector and the principal subspace of a signal's covariance matrix adaptively in a fully decentralized wireless sensor network (WSN). Sensor networks are traditionally designed to simply gather raw data at a fusion center, where all the processing occurs. In large deployments, this model entails high networking cost and creates a computational and storage bottleneck for the system. By leveraging both sensors' abilities to communicate and their local computational power, our objective is to propose distributed algorithms for principal eigenvector and principal subspace tracking. We show that it is possible to have each sensor estimate only the corresponding entry of the principal eigenvector or corresponding row of the $p$-dimensional principal subspace matrix and do so by iterating a simple computation that combines data from its network neighbors only. This paper also examines the convergence properties of the proposed principal eigenvector and principal subspace tracking algorithms analytically and by simulations.

KW - Asynchronous time

KW - average consensus

KW - distributed algorithm

KW - gossiping

KW - signal detection

KW - stochastic approximation

KW - subspace estimation

KW - subspace tracking

KW - synchronous time

UR - http://www.scopus.com/inward/record.url?scp=79960692394&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960692394&partnerID=8YFLogxK

U2 - 10.1109/JSTSP.2011.2118742

DO - 10.1109/JSTSP.2011.2118742

M3 - Article

AN - SCOPUS:79960692394

VL - 5

SP - 725

EP - 738

JO - IEEE Journal on Selected Topics in Signal Processing

JF - IEEE Journal on Selected Topics in Signal Processing

SN - 1932-4553

IS - 4

M1 - 5719545

ER -