Comments on "Estimation of frequencies and damping factors by two-dimensional ESPRIT type methods"

For original paper, see Rouquette and Najim (IEEE Trans. Signal Process., vol.49, no.1, p.237-45, 2001).The original authors present two high-resolution approaches to estimate two-dimensional (2-D) signal parameters. The first approach decomposes the 2-D estimation problem into two one-dimensional (1-D) ones, estimates the 1-D frequencies with the MEMP method, and then forms correct pairs using a new pairing scheme introduced in Vanpoucke et al. (1994). The second method (the 2-D ESPRIT method) directly estimates the 2-D frequencies. It is shown that both methods are more accurate than the related methods. The present authors note that the original paper appears to have a key error in the permutation matrix specification of the first method. In this comment, they derive the correct one with the performance simulations.