Riemann-Langevin Particle Filtering in Track-Before-Detect

Fernando J. Iglesias-Garcia; Pranab K. Mandal; Mélanie Bocquel; Antonio G. Marques

doi:10.48550/arXiv.1705.00546

Riemann-Langevin Particle Filtering in Track-Before-Detect

Fernando J. Iglesias-Garcia, Pranab K. Mandal, Mélanie Bocquel, Antonio G. Marques

Hybrid Systems

Research output: Working paper › Preprint › Academic

1 Downloads (Pure)

Abstract

Track-before-detect (TBD) is a powerful approach that consists in providing the tracker with sensor measurements directly without pre-detection. Due to the measurement model non-linearities, online state estimation in TBD is most commonly solved via particle filtering. Existing particle filters for TBD do not incorporate measurement information in their proposal distribution. The Langevin Monte Carlo (LMC) is a sampling method whose proposal is able to exploit all available knowledge of the posterior (that is, both prior and measurement information). This letter synthesizes recent advances in LMC-based filtering to describe the Riemann-Langevin particle filter and introduces its novel application to TBD. The benefits of our approach are illustrated in a challenging low-noise scenario.

Original language	English
Publisher	ArXiv.org
DOIs	https://doi.org/10.48550/arXiv.1705.00546
Publication status	Published - 1 May 2017

Keywords

stat.AP
stat.CO

Access to Document

10.48550/arXiv.1705.00546

1705.00546v2Final published version, 374 KB

1 Article

Riemann–Langevin Particle Filtering in Track-Before-Detect
Iglesias Garcia, F. (Corresponding Author), Mandal, P. K., Bocquel, M. & Garcia Marques, A., Jul 2018, In: IEEE signal processing letters. 25, 7, p. 1039-1043 5 p.
Research output: Contribution to journal › Article › Academic › peer-review

Open Access
File
14 Citations (Scopus)

211 Downloads (Pure)

Cite this

@techreport{f70d12e8f6a5497d9b46599d29add7af,

title = "Riemann-Langevin Particle Filtering in Track-Before-Detect",

abstract = " Track-before-detect (TBD) is a powerful approach that consists in providing the tracker with sensor measurements directly without pre-detection. Due to the measurement model non-linearities, online state estimation in TBD is most commonly solved via particle filtering. Existing particle filters for TBD do not incorporate measurement information in their proposal distribution. The Langevin Monte Carlo (LMC) is a sampling method whose proposal is able to exploit all available knowledge of the posterior (that is, both prior and measurement information). This letter synthesizes recent advances in LMC-based filtering to describe the Riemann-Langevin particle filter and introduces its novel application to TBD. The benefits of our approach are illustrated in a challenging low-noise scenario. ",

keywords = "stat.AP, stat.CO",

author = "Iglesias-Garcia, {Fernando J.} and Mandal, {Pranab K.} and M{\'e}lanie Bocquel and Marques, {Antonio G.}",

note = "Minor grammatical updates",

year = "2017",

month = may,

day = "1",

doi = "10.48550/arXiv.1705.00546",

language = "English",

publisher = "ArXiv.org",

type = "WorkingPaper",

institution = "ArXiv.org",

}

TY - UNPB

T1 - Riemann-Langevin Particle Filtering in Track-Before-Detect

AU - Iglesias-Garcia, Fernando J.

AU - Mandal, Pranab K.

AU - Bocquel, Mélanie

AU - Marques, Antonio G.

N1 - Minor grammatical updates

PY - 2017/5/1

Y1 - 2017/5/1

N2 - Track-before-detect (TBD) is a powerful approach that consists in providing the tracker with sensor measurements directly without pre-detection. Due to the measurement model non-linearities, online state estimation in TBD is most commonly solved via particle filtering. Existing particle filters for TBD do not incorporate measurement information in their proposal distribution. The Langevin Monte Carlo (LMC) is a sampling method whose proposal is able to exploit all available knowledge of the posterior (that is, both prior and measurement information). This letter synthesizes recent advances in LMC-based filtering to describe the Riemann-Langevin particle filter and introduces its novel application to TBD. The benefits of our approach are illustrated in a challenging low-noise scenario.

AB - Track-before-detect (TBD) is a powerful approach that consists in providing the tracker with sensor measurements directly without pre-detection. Due to the measurement model non-linearities, online state estimation in TBD is most commonly solved via particle filtering. Existing particle filters for TBD do not incorporate measurement information in their proposal distribution. The Langevin Monte Carlo (LMC) is a sampling method whose proposal is able to exploit all available knowledge of the posterior (that is, both prior and measurement information). This letter synthesizes recent advances in LMC-based filtering to describe the Riemann-Langevin particle filter and introduces its novel application to TBD. The benefits of our approach are illustrated in a challenging low-noise scenario.

KW - stat.AP

KW - stat.CO

U2 - 10.48550/arXiv.1705.00546

DO - 10.48550/arXiv.1705.00546

M3 - Preprint

BT - Riemann-Langevin Particle Filtering in Track-Before-Detect

PB - ArXiv.org

ER -

Riemann-Langevin Particle Filtering in Track-Before-Detect

Abstract

Keywords

Access to Document

Fingerprint

Research output

Riemann–Langevin Particle Filtering in Track-Before-Detect

Cite this