Greedy sensor selection based on QR factorization
Abstract We address the problem of selecting a given number of sensor nodes in wireless sensor networks where noise-corrupted linear measurements are collected at the selected nodes to estimate the unknown parameter. Noting that this problem is combinatorial in nature and selection of sensor nodes f...
Guardado en:
Autor principal: | |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
SpringerOpen
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/c79239bb16b34a3fa90ba52b6cfe7dc9 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:c79239bb16b34a3fa90ba52b6cfe7dc9 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:c79239bb16b34a3fa90ba52b6cfe7dc92021-12-05T12:10:30ZGreedy sensor selection based on QR factorization10.1186/s13634-021-00824-51687-6180https://doaj.org/article/c79239bb16b34a3fa90ba52b6cfe7dc92021-12-01T00:00:00Zhttps://doi.org/10.1186/s13634-021-00824-5https://doaj.org/toc/1687-6180Abstract We address the problem of selecting a given number of sensor nodes in wireless sensor networks where noise-corrupted linear measurements are collected at the selected nodes to estimate the unknown parameter. Noting that this problem is combinatorial in nature and selection of sensor nodes from a large number of nodes would require unfeasible computational cost, we propose a greedy sensor selection method that seeks to choose one node at each iteration until the desired number of sensor nodes are selected. We first apply the QR factorization to make the mean squared error (MSE) of estimation a simplified metric which is iteratively minimized. We present a simple criterion which enables selection of the next sensor node minimizing the MSE at iterations. We discuss that a near-optimality of the proposed method is guaranteed by using the approximate supermodularity and also make a complexity analysis for the proposed algorithm in comparison with different greedy selection methods, showing a reasonable complexity of the proposed method. We finally run extensive experiments to investigate the estimation performance of the different selection methods in various situations and demonstrate that the proposed algorithm provides a good estimation accuracy with a competitive complexity when compared with the other novel greedy methods.Yoon Hak KimSpringerOpenarticleGreedy algorithmSensor selectionLinear inverse problemNear-optimalityQR factorizationTelecommunicationTK5101-6720ElectronicsTK7800-8360ENEURASIP Journal on Advances in Signal Processing, Vol 2021, Iss 1, Pp 1-13 (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
Greedy algorithm Sensor selection Linear inverse problem Near-optimality QR factorization Telecommunication TK5101-6720 Electronics TK7800-8360 |
spellingShingle |
Greedy algorithm Sensor selection Linear inverse problem Near-optimality QR factorization Telecommunication TK5101-6720 Electronics TK7800-8360 Yoon Hak Kim Greedy sensor selection based on QR factorization |
description |
Abstract We address the problem of selecting a given number of sensor nodes in wireless sensor networks where noise-corrupted linear measurements are collected at the selected nodes to estimate the unknown parameter. Noting that this problem is combinatorial in nature and selection of sensor nodes from a large number of nodes would require unfeasible computational cost, we propose a greedy sensor selection method that seeks to choose one node at each iteration until the desired number of sensor nodes are selected. We first apply the QR factorization to make the mean squared error (MSE) of estimation a simplified metric which is iteratively minimized. We present a simple criterion which enables selection of the next sensor node minimizing the MSE at iterations. We discuss that a near-optimality of the proposed method is guaranteed by using the approximate supermodularity and also make a complexity analysis for the proposed algorithm in comparison with different greedy selection methods, showing a reasonable complexity of the proposed method. We finally run extensive experiments to investigate the estimation performance of the different selection methods in various situations and demonstrate that the proposed algorithm provides a good estimation accuracy with a competitive complexity when compared with the other novel greedy methods. |
format |
article |
author |
Yoon Hak Kim |
author_facet |
Yoon Hak Kim |
author_sort |
Yoon Hak Kim |
title |
Greedy sensor selection based on QR factorization |
title_short |
Greedy sensor selection based on QR factorization |
title_full |
Greedy sensor selection based on QR factorization |
title_fullStr |
Greedy sensor selection based on QR factorization |
title_full_unstemmed |
Greedy sensor selection based on QR factorization |
title_sort |
greedy sensor selection based on qr factorization |
publisher |
SpringerOpen |
publishDate |
2021 |
url |
https://doaj.org/article/c79239bb16b34a3fa90ba52b6cfe7dc9 |
work_keys_str_mv |
AT yoonhakkim greedysensorselectionbasedonqrfactorization |
_version_ |
1718372218498449408 |