Robust subspace tracking algorithms using fast adaptive Mahalanobis distance

•A novel algorithm to estimate adaptive Mahalanobis distance for mitigating the effect of burst noise, used with subspace tracking algorithms.•New subspace tracking algorithms, RYAST and ROBUSTQR, robust to burst noise, redesigned from the corresponding state-of the-arts, YAST and LORAF.•Structure of signal model and noise characteristics exploited to estimate the covariance matrix, leading to low complexity of RYAST and ROBUSTQR.