(Received June 2, 1988; revised April 13, 1990)
Abstract. An iterative algorithm for the robust M-estimation of the dispersion matrix of the form Gamma + sigma2Ip has been given. This algorithm converges after some steps and reduces the effect of outliers on the covariance matrix. The consistency and asymptotic normality of the estimator are established.
Key words and phrases: Robust estimation, M-estimation, elliptically symmetric distribution, outliers, signal processing.