Module 3 – Doppler Centroid Estimation
Program to run: ‘avg_phschg.m’ OR ‘madsen_algo.m’
Functions used : -
Input : range_compressed.dat
Output : Doppler centroid frequency (fdc)
Description :
Doppler centroid frequency value is non-zero due to squint angle and earth rotation.
Doppler centroid frequency is the frequency shift from the point target which is at the centre of
the antenna beam. For an ideal SAR system, its value should be zero but practically due to non-
zero squint angle of the sensor and the sensor trajectory, it is different from zero. In order to
process the range compressed data in azimuth direction, we need to know the Doppler frequency.
Two methods viz. the average phase change method and the Sign Doppler method i.e. the
Madsen’s method have been is used in this module to estimate the Doppler centroid frequency.
Run ‘avg_phschg.m’ for average phase change method and ‘madsen_algo.m’ for sign Doppler
estimator method.
Average phase change method:
The average phase change method calculates the Doppler centroid frequency by using the
range compressed image (raw data can also be used since azimuth spectrum will not be affected
by operations in range direction). In this method, the phase change between two azimuth values
for each range value is calculated and averaged. This calculation is done by multiplying one
azimuth value with the complex conjugate of the successive azimuth value for each range bin
and taking the inverse tangent of imaginary part upon the real part of the product. The mean of
these average phase changes is then used to estimate the Doppler centroid frequency using the
following relation.
( )
( )
Sign Doppler Estimator (Madsen’s method):
This time domain technique depends on the fact that the Fourier transform of
autocorrelation function of a signal is the power spectral density of the signal. So, a shift in
frequency domain of the signal will lead to corresponding linear phase shift in its autocorrelation
function. This phase shift is directly proportional to the Doppler shift.
( ( ))
Where ( ) is the autocorrelation function of the given signal. The algorithm estimates four
sign correlations and derives corresponding correlation coefficients. Then, the complex
correlation coefficients are derived after which the above relation is applied to get Doppler
frequency shift.
Results:
sign doppler estimator method
-0.35
-0.4
-0.45
average phase change [rad]
-0.5
-0.55
-0.6
-0.65
-0.7
-0.75
0 5 10 15 20 25 30 35
range bin
Doppler frequency is -140.21 Hz..
Average phase change method
0.5
-0.5
average phase change [rad]
-1
-1.5
-2
-2.5
-3
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
range bin
Doppler centroid -134.99 Hz