An alternative approach for distributed parameter estimation under Gaussian settings

TL;DR

This paper introduces a novel distributed Gaussian parameter estimation algorithm for multi-agent networks that combines consensus and innovation strategies to ensure consistency and rapid convergence.

Contribution

It proposes a new distributed estimation method that integrates consensus and innovation, with optimal gain design under a novel observability condition.

Findings

Estimates are consistent and converge quickly.

The method outperforms traditional approaches in Gaussian settings.

Optimal gain matrices are derived for improved performance.

Abstract

This paper takes a different approach for the distributed linear parameter estimation over a multi-agent network. The parameter vector is considered to be stochastic with a Gaussian distribution. The sensor measurements at each agent are linear and corrupted with additive white Gaussian noise. Under such settings, this paper presents a novel distributed estimation algorithm that fuses the the concepts of consensus and innovations by incorporating the consensus terms (of neighboring estimates) into the innovation terms. Under the assumption of distributed parameter observability, introduced in this paper, we design the optimal gain matrices such that the distributed estimates are consistent and achieves fast convergence.