Robust error estimation based on factor-graph models for non-line-of-sight localization

Abstract This paper presents a method to estimate the covariances of inputs in factor-graph formulation for localization under non-line-of-sight conditions. A general solution based on covariance estimation and M-estimators linear regression problems, is presented that shown give unbiased estimators multiple variances are robust against outliers. An iteratively re-weighted least squares algorithm proposed jointly compute variance state estimates nonlinear factor graph optimization. The efficacy illustrated simulation study using robot problem various process measurement models outlier scenarios. case involving Global Positioning System an urban environment data containing multipath problems demonstrates application technique.