Ambient Sound-Based Collaborative Localization of Indeterministic Devices

Research output: Contribution to journalArticleAcademicpeer-review

15 Downloads (Pure)

Abstract

Localization is essential in wireless sensor networks. To our knowledge, no prior work has utilized low-cost devices for collaborative localization based on only ambient sound, without the support of local infrastructure. The reason may be the fact that most low-cost devices are indeterministic and suffer from uncertain input latencies. This uncertainty makes accurate localization challenging. Therefore, we present a collaborative localization algorithm (Cooperative Localization on Android with ambient Sound Sources (CLASS)) that simultaneously localizes the position of indeterministic devices and ambient sound sources without local infrastructure. The CLASS algorithm deals with the uncertainty by splitting the devices into subsets so that outliers can be removed from the time difference of arrival values and localization results. Since Android is indeterministic, we select Android devices to evaluate our approach. The algorithm is evaluated with an outdoor experiment and achieves a mean Root Mean Square Error (RMSE) of 2.18 m with a standard deviation of 0.22 m. Estimated directions towards the sound sources have a mean RMSE of 17.5 ° and a standard deviation of 2.3 °. These results show that it is feasible to simultaneously achieve a relative positioning of both devices and sound sources with sufficient accuracy, even when using non-deterministic devices and platforms, such as Android.
Original languageUndefined
Pages (from-to)1478
Number of pages23
JournalSensors (Switserland)
Volume16
Issue number9
DOIs
Publication statusPublished - 14 Sep 2016

Keywords

  • EWI-27188
  • opportunistic localization
  • Smart phones
  • Sound localization
  • input latency
  • IR-101849
  • Android
  • Wireless Sensor Networks
  • METIS-318508
  • collaborative localization

Cite this

@article{8bb0a0620fd245d7a5864c674a0d3be0,
title = "Ambient Sound-Based Collaborative Localization of Indeterministic Devices",
abstract = "Localization is essential in wireless sensor networks. To our knowledge, no prior work has utilized low-cost devices for collaborative localization based on only ambient sound, without the support of local infrastructure. The reason may be the fact that most low-cost devices are indeterministic and suffer from uncertain input latencies. This uncertainty makes accurate localization challenging. Therefore, we present a collaborative localization algorithm (Cooperative Localization on Android with ambient Sound Sources (CLASS)) that simultaneously localizes the position of indeterministic devices and ambient sound sources without local infrastructure. The CLASS algorithm deals with the uncertainty by splitting the devices into subsets so that outliers can be removed from the time difference of arrival values and localization results. Since Android is indeterministic, we select Android devices to evaluate our approach. The algorithm is evaluated with an outdoor experiment and achieves a mean Root Mean Square Error (RMSE) of 2.18 m with a standard deviation of 0.22 m. Estimated directions towards the sound sources have a mean RMSE of 17.5 ° and a standard deviation of 2.3 °. These results show that it is feasible to simultaneously achieve a relative positioning of both devices and sound sources with sufficient accuracy, even when using non-deterministic devices and platforms, such as Android.",
keywords = "EWI-27188, opportunistic localization, Smart phones, Sound localization, input latency, IR-101849, Android, Wireless Sensor Networks, METIS-318508, collaborative localization",
author = "Kamminga, {Jacob Wilhelm} and {Le Viet Duc}, {L Duc} and Havinga, {Paul J.M.}",
note = "Open access",
year = "2016",
month = "9",
day = "14",
doi = "10.3390/s16091478",
language = "Undefined",
volume = "16",
pages = "1478",
journal = "Sensors (Switserland)",
issn = "1424-8220",
publisher = "Multidisciplinary Digital Publishing Institute",
number = "9",

}

Ambient Sound-Based Collaborative Localization of Indeterministic Devices. / Kamminga, Jacob Wilhelm; Le Viet Duc, L Duc; Havinga, Paul J.M.

In: Sensors (Switserland), Vol. 16, No. 9, 14.09.2016, p. 1478.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Ambient Sound-Based Collaborative Localization of Indeterministic Devices

AU - Kamminga, Jacob Wilhelm

AU - Le Viet Duc, L Duc

AU - Havinga, Paul J.M.

N1 - Open access

PY - 2016/9/14

Y1 - 2016/9/14

N2 - Localization is essential in wireless sensor networks. To our knowledge, no prior work has utilized low-cost devices for collaborative localization based on only ambient sound, without the support of local infrastructure. The reason may be the fact that most low-cost devices are indeterministic and suffer from uncertain input latencies. This uncertainty makes accurate localization challenging. Therefore, we present a collaborative localization algorithm (Cooperative Localization on Android with ambient Sound Sources (CLASS)) that simultaneously localizes the position of indeterministic devices and ambient sound sources without local infrastructure. The CLASS algorithm deals with the uncertainty by splitting the devices into subsets so that outliers can be removed from the time difference of arrival values and localization results. Since Android is indeterministic, we select Android devices to evaluate our approach. The algorithm is evaluated with an outdoor experiment and achieves a mean Root Mean Square Error (RMSE) of 2.18 m with a standard deviation of 0.22 m. Estimated directions towards the sound sources have a mean RMSE of 17.5 ° and a standard deviation of 2.3 °. These results show that it is feasible to simultaneously achieve a relative positioning of both devices and sound sources with sufficient accuracy, even when using non-deterministic devices and platforms, such as Android.

AB - Localization is essential in wireless sensor networks. To our knowledge, no prior work has utilized low-cost devices for collaborative localization based on only ambient sound, without the support of local infrastructure. The reason may be the fact that most low-cost devices are indeterministic and suffer from uncertain input latencies. This uncertainty makes accurate localization challenging. Therefore, we present a collaborative localization algorithm (Cooperative Localization on Android with ambient Sound Sources (CLASS)) that simultaneously localizes the position of indeterministic devices and ambient sound sources without local infrastructure. The CLASS algorithm deals with the uncertainty by splitting the devices into subsets so that outliers can be removed from the time difference of arrival values and localization results. Since Android is indeterministic, we select Android devices to evaluate our approach. The algorithm is evaluated with an outdoor experiment and achieves a mean Root Mean Square Error (RMSE) of 2.18 m with a standard deviation of 0.22 m. Estimated directions towards the sound sources have a mean RMSE of 17.5 ° and a standard deviation of 2.3 °. These results show that it is feasible to simultaneously achieve a relative positioning of both devices and sound sources with sufficient accuracy, even when using non-deterministic devices and platforms, such as Android.

KW - EWI-27188

KW - opportunistic localization

KW - Smart phones

KW - Sound localization

KW - input latency

KW - IR-101849

KW - Android

KW - Wireless Sensor Networks

KW - METIS-318508

KW - collaborative localization

U2 - 10.3390/s16091478

DO - 10.3390/s16091478

M3 - Article

VL - 16

SP - 1478

JO - Sensors (Switserland)

JF - Sensors (Switserland)

SN - 1424-8220

IS - 9

ER -