An improved method for the determination of airplane radial velocity using doppler shift

Abstract EN

In this research velocity of moving airplane from its recorded digital sound is introduced.

The data of sound file is sliced into several frames using overlapping partitions.

Then the array of each frame is transformed from time domain to frequency domain using Fourier Transform (FT).

To determine the characteristic frequency of the sound, a moving window mechanics is used, the size of that window is made linearly proportional with the value of the tracked frequency.

This proportionality is due to the existing linear relationship between the frequency and its Doppler shift.

An algorithm was introduced to select the characteristic frequencies, this algorithm allocates the frequencies which satisfy the Doppler relation, beside that the tracked frequencies was chosen to be the most occurred frequencies in the power spectrum.

Several sizes of the overlap windows have been tested to study the effect of window size on the accuracy of the determined radial velocity of the moving airplane.

The obtained results have been compared with the corresponding results of non-overlapped window.

The accuracy of the results was assessed using two criteria, the first is the closeness of the curves of the determined radial velocities versus time for the case of overlapped and non-overlapped window; the closer curves indicate better results.

The second used criteria is the Root Mean Square Error (RMSE), it is used to determine the degree a method for determining the radial of linearity of the relationship between Doppler shift and characteristic frequencies, the least values of (RMSE) signify better results are obtained.