ANSI-C program: dop_mlcc.c
NAME
dop_mlcc - Doppler ambiguity
estimation from SAR raw data using the MLCC algorithm
SYNOPSIS
dop_mlcc <SAR_param> <PROC_param>
<signal_data> [output_plot] [loff] [nlines]
| <SAR_param> | (input) SAR sensor parameter file |
| <PROC_param> | (input) MSP Processing Parameter File |
| <signal_data> | (input) uncompressed raw SAR signal data (I/Q complex) |
| [output_plot] | (output) plot file of the MLCC phase differences |
| [loff] | number of lines offset (enter - for default=parameter file value) |
| [nlines] | number of lines to process |
EXAMPLES
dop_mlcc JERS_1.par p921106.slc.par 921106.raw
921106.mlcc
DESCRIPTION
The Doppler centroid is a necessary parameter for SAR processing.
Due to the fact that the radar is a sampled data system, i.e. the
SAR acquires echoes at the rate of the pulse repetition frequency
(PRF), the Doppler spectrum is aliased modulo the PRF. This means
that some other method must be used to determine an unambiguous
value of the Doppler centroid apart from simple fourier analysis
of the azimuth spectrum.
The raw SAR data (IQ complex data)can be used to estimate the ambiguity number of the Doppler centroid using the methods described by Wong and Cumming (A Combined SAR Doppler Centroid Estimation Scheme based upon Signal Phase, IEEE Trans. Geosci. Remote Sensing, Vol. 34, No. 3, May 1996, pp. 696-707.) The signal phase from different complex range looks is used to obtain an unambiguous estimate of the Doppler centroid. This value is stored in the Processing Parameter file with the DAR_doppler keyword. The programs doppler and azsp_IQ can use the unambiguous Doppler estimate to determine the Doppler ambiguity number and add it to the accurate, but ambiguous estimate.
The Multi-Look Cross Correlation (MLCC) algorithm is the most robust and measures the difference in Doppler centroid as a function of frequency. The correlation method of centroid estimation is implemented as in doppler for two segments of the range spectrum. The rate of change of Doppler with respect to frequency yields an unambiguous estimate of the Doppler centroid. See discussion in the Users Guide. The data segment used for the estimation can be specified by the user by setting the offset in lines relative to the beginning of the file.
The MLCC plots the Doppler in radians of phase shift between range lines versus the slant range pixel. If the average of the correlation coefficient between range pulses is PI/2, then the centroid is PRF/2 + N*PRF where N is the number of ambiguities in the Doppler centroid. The phase of the correlation is plotted for both the upper and lower segments of the range bandwidth. The last column is the difference between the Doppler centroid estimates of the upper and lower band range bandwidth. This difference can be used to determine a noisy estimate of the unambiguous Doppler centroid. The avarage value of the difference is used to determine the unambiguous estimate of the Doppler centroid.
| Column | Parameter |
|---|---|
|
|
range sample number |
|
|
line to line corrlation phase dp1 for lower half of the range spectrum in units of radians. Doppler centroid = fd1 = dp1*PRF/(2*PI) |
|
|
line to line corrlation phase dp2 for upper half of the range spectrum in units of radians. Doppler centroid = fd2 = dp2*PRF/(2*PI) |
|
|
difference of line to line correlation in units of radians (=dp3). DAR_dop = (dp2-dp1)*PRF/(2*PI) unambiguous Doppler centroid estimate: DAR_dop = -f_c*PRF*dp3/(2*PI*df) where df = f_adc/2 and f_adc is the range sampling frequency and f_c is the radar carrier frequency. |
SEE ALSO
Users Guide, dop_ambig, typedef_MSP.h, Sensor parameter file, MSP Processing Parameter File.