highiq.calc.welchs_method

highiq.calc.welchs_method(complex_coeff, fs=50000000.0, nfft=32, window_skip=16, num_per_block=200)

This technique is an implementation of Welch’s method for processing a Doppler spectral density function from complex autocorrelation function data. Welch’s method calculates the Doppler spectral density function by calculating the FFT over (potentially overlapping) windows and then taking the average of the magnitudes of each FFT to create the Doppler spectra.

Parameters

complex_coeff: complex nD array

An n-D array of complex floats representing the value of the autocorrelation function. The first dimension is the number of entries in time, and the second is the number of entries in the range dimension.

fs:

The pulse repetition frequency of the radar.

nfft:

The number of points to include in the FFT for calculating spectra. This must be an even number.

window_skip:

The number of points to go forward for each window. This cannot exceed the size of the FFT.

num_per_block:

The number of time periods to send to the GPU at one time. This is implemented because the GPU can only handle a limited number of data points at one time.

Returns

freq: 1D array

The frequencies corresponding to each value in the spectra.

power: nD array

The power spectral densities for each frequency.