Analysis of weighted subspace fitting and subspace-based eigenvector techniques for frequency estimation for the...

Area of Science:

• Atmospheric Science
• Signal Processing
• Remote Sensing Technology

Background:

• Periodogram Maximum (PM) algorithm lacks consistent estimates, particularly in low signal-to-noise ratio (SNR) environments.
• Subspace fitting frameworks offer robust alternatives for spectral estimation.

Purpose of the Study:

• To develop and validate a more robust wind estimation technique for coherent Doppler lidar.
• To enhance accuracy and extend the detectable range in low SNR conditions.

Main Methods:

• Formulation within a subspace fitting framework.
• Introduction of a Weighted Subspace Fitting (WSF) method utilizing an optimal weighting matrix.
• Exploitation of low-rank properties of the covariance matrix for coherent Doppler lidar echo data.

Main Results:

• WSF yields a higher number of reliable estimates and the smallest standard deviation compared to PM and Eigenvector (EV) methods.
• WSF demonstrates a narrower spectral width in the probability density function of estimates.
• Experimental validation confirms WSF outperforms PM and EV algorithms in furthest detectable range, improving it by up to 14.2% (vs. EV) and 26.6% (vs. PM).

Conclusions:

• The proposed WSF method significantly reduces statistical uncertainties in wind estimation.
• WSF enhances accuracy and extends the detection range for coherent Doppler lidar, especially in low SNR regimes.