Frequency Estimation Algorithms for Monotone Signals

This article discusses several frequency estimation algorithms for monotone signals, such as the Maximum Likelihood method and Pisarenko harmonic decomposition technique. These algorithms enable accurate frequency estimation of signal components. The Maximum Likelihood method represents a statistically optimal approach that maximizes the probability of observing the given signal data, typically implemented through grid search or gradient-based optimization over possible frequency values. The Pisarenko method utilizes eigenvalue decomposition of the signal's autocorrelation matrix to estimate frequencies from noise subspaces, particularly effective for harmonic signals in additive noise. By employing these techniques with proper MATLAB implementations involving functions like fminsearch for optimization or eig for matrix decomposition, researchers can achieve precise frequency estimates that facilitate better understanding of signal characteristics and underlying physical phenomena.