Robust Estimation in Signal Processing: A Tutorial-Style Treatment of Fundamental Concepts

摘要: The word robust has been used in many contexts signal processing. Our treatment concerns statistical robustness, which deals with deviations from the distributional assumptions. Many problems encountered engineering practice rely on Gaussian distribution of data, situations is well justified. This enables a simple derivation optimal estimators. Nominal optimality, however, useless if estimator was derived under assumptions noise and that do not hold practice. Even slight assumed may cause estimator's performance to drastically degrade or completely break down. processing practitioner should, therefore, ask whether acceptable where hold. Isn't it robustness major concern for practice? areas today show measurements far as contains outliers, be heavy tailed. Under such scenarios, we address single multichannel estimation linear univariate regression independently identically distributed (i.i.d.) data. A rather extensive important challenging case dependent data also included. For these problems, comparative analysis most methods carried out by evaluating their theoretically, using simulations real-world