2-D single record harmonic retrieval using higher-order statistics

Given a single record, we consider the problem of estimating the harmonics of a 2-D signal corrupted by Gaussian noise. Our method utilizes higher-order statistics (HOS), referred to as "mixed"-cumulants, which can be consistently estimated from a single record and are insensitive to colored Gaussian noise. In particular, we first estimate the frequencies in each dimension using an overdetermined Yule-Walker "type" approach based on 4th-order mixed-cumulants. Then, the 1-D frequencies are paired using a mixed-cumulant matching criterion. Simulations examining the performance of the proposed method are provided.<>