Information Fusion via Importance Sampling

Augustin A. Saucan, Victor Elvira, Pramod K. Varshney, Moe Z. Win

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Information fusion is a procedure that merges information locally contained at the nodes of a network. Of high interest in the field of distributed estimation is the fusion of local probability distributions via a weighted geometrical average criterion. In numerous practical settings, the local distributions are only known through particle approximations, i.e., sets of samples with associated weights, such as obtained via importance sampling is methods. Thus, prohibiting any closed-form solution to the aforementioned fusion problem. This paper proposes a family of IS methods—called particle geometric-average fusion (PGAF)—that lead to consistent estimators for the geometrically-averaged density. The advantages of the proposed methods are threefold. First, the methods are agnostic of the mechanisms used to generate the local particle sets and, therefore, allow for the fusion of heterogeneous nodes. Second, consistency of estimators is guaranteed under generic conditions when the agents use IS-generated particles. Third, a low-communication overhead and agent privacy are achieved since local observations are not shared with the fusion center. Even more remarkably, for a sub-family of the proposed PGAF methods, the fusion center does not require the knowledge of the local priors used by the nodes. Implementation guidelines for the proposed methods are provided and theoretical results are numerically verified.

Original languageEnglish (US)
Pages (from-to)1-14
Number of pages14
JournalIEEE Transactions on Signal and Information Processing over Networks
DOIs
StateAccepted/In press - 2023
Externally publishedYes

Keywords

  • Atmospheric measurements
  • Kullback–Leibler divergence
  • Monte Carlo methods
  • Monte Carlo methods
  • Particle measurements
  • Probability density function
  • Proposals
  • Random variables
  • Robot sensing systems
  • information fusion
  • kernel density estimate
  • multiple importance sampling

ASJC Scopus subject areas

  • Signal Processing
  • Information Systems
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Information Fusion via Importance Sampling'. Together they form a unique fingerprint.

Cite this