Eigendecomposition methods for frequency estimation: a unified approach

A unified approach to three eigendecomposition-based methods for frequency estimation in the presence of noise is presented. These are the Tufts-Kumaresan (TK) method, the minimum-norm (MN) method, and the total-least-squares (TLS) method. It is shown that the MN method is a modified version of the TK method and the TLS method is a generalization of the MN methods. The TLS solution vector is expressed in matrix form, and an alternate way of computing it is presented. The MN methods exhibit some improvement over the TK method.<>