Particle filter adaptation for distributed sensors via set membership

Shahrokh Farahmand, Stergios I. Roumeliotis, Georgios B. Giannakis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

26 Scopus citations

Abstract

A distributed set-membership-constrained particle filter (SMC-PF) is developed for decentralized tracking applications using wireless sensor networks. Unlike existing PF alternatives, SMC-PF offers reduced overhead for inter-sensor communications because it requires only particle weights to be exchanged among sensors, instead of raw measurements or parameters of a Gaussian mixture model. SMC-PF relies on a novel distributed adaptation scheme based on successive set intersections that can afford reduced number of particles without sacrificing performance. Conditions are provided to quantify the variance reduction of the SMC-PF-based state estimator. Simulations corroborate the ability of the SMC-PF to considerably outperform the bootstrap PF for a fixed number of particles.

Original languageEnglish (US)
Title of host publication2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3374-3377
Number of pages4
ISBN (Print)9781424442966
DOIs
StatePublished - 2010
Event2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Dallas, TX, United States
Duration: Mar 14 2010Mar 19 2010

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN (Print)1520-6149

Other

Other2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010
Country/TerritoryUnited States
CityDallas, TX
Period3/14/103/19/10

Keywords

  • Adaptation
  • Distributed
  • Particle filtering
  • Set-membership

Fingerprint

Dive into the research topics of 'Particle filter adaptation for distributed sensors via set membership'. Together they form a unique fingerprint.

Cite this