Self-localization from landmark bearings using pseudolinear estimation techniques

This correspondence presents a pseudolinear estimator (PLE) and its improved variants for self-localization from landmark bearings. Despite its simplicity a well-known problem with the PLE is its large estimation bias. Based on the PLE, low-complexity, and reduced-bias, weighted instrumental variable (WIV) estimators are developed. It is shown by simulation examples that these new estimators are capable of achieving localization performance on par with the computationally demanding maximum likelihood estimator (MLE).