Two-Dimensional Frequency Estimation in Correlative Noise Based on Genetic Algorithm

The problem of two-dimensional (2D) harmonics retrieval in the presence of correlative zero-mean multiplicative and additive noise was presented. We propose a genetic algorithm (GA)-based approach to this problem. Firstly, we defined a special 2D slice of sixth-order time-average moment spectrum, which has peaks value at the harmonics frequencies. Second, we converted the problem of harmonics retrieval into extremum problem of binary function. Finally, the GA was employed to estimate the frequencies of 2D harmonics. The effectiveness of our algorithms is demonstrated through some simulations