Importance Densities for Particle Filtering Using Iterated Conditional Expectations

Roland Hostettler*, Filip Tronarp, Angel F. Garcia-Fernandez, Simo Sarkka

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

7 Citations (Scopus)

Abstract

In this letter, we consider Gaussian approximations of the optimal importance density in sequential importance sampling for nonlinear, non-Gaussian state-space models. The proposed method is based on generalized statistical linear regression and posterior linearization using conditional expectations. Simulation results show that the method outperforms the compared methods in terms of the effective sample size and provides a better local approximation of the optimal importance density.

Original languageEnglish
Article number8951063
Pages (from-to)211-215
Number of pages5
JournalIEEE Signal Processing Letters
Volume27
DOIs
Publication statusPublished - 1 Jan 2020
MoE publication typeA1 Journal article-refereed

Funding

Manuscript received October 2, 2019; revised November 24, 2019; accepted December 30, 2019. Date of publication January 7, 2020; date of current version January 31, 2020. This work was supported in part by the Aalto ELEC Doctoral School and in part by the Academy of Finland. The associate editor coordinating the review of this manuscript and approving it for publication was Prof. S. Zozor. (Corresponding author: Roland Hostettler.) R. Hostettler is with the Department of Engineering Sciences, Uppsala University, 752 36 Uppsala, Sweden (e-mail: [email protected]).

Keywords

  • Monte Carlo methods
  • Nonlinear systems
  • Particle filters
  • Posterior linearization
  • State estimation

Fingerprint

Dive into the research topics of 'Importance Densities for Particle Filtering Using Iterated Conditional Expectations'. Together they form a unique fingerprint.

Cite this