Distributed Estimation From Relative Measurements of Heterogeneous and Uncertain Quality

摘要: This paper studies the problem of estimation from relative measurements in a graph, which vector indexed over nodes has to be reconstructed pairwise differences between its components associated with connected by an edge. In order model heterogeneity and uncertainty measurements, we assume them affected additive noise distributed according Gaussian mixture. this original setup, formulate computing maximum-likelihood estimates design two novel algorithms, based on least squares (LS) regression expectation maximization (EM). The first algorithm (LS-EM) is centralized performs soft classification parameters. second (Distributed LS-EM) but requires knowledge We provide rigorous proofs convergence for both algorithms present numerical experiments evaluate their performance compare it solutions literature. show robustness proposed methods against different kinds and, Distributed LS-EM, errors