3
$\begingroup$

Im new to RADAR signal processing.
I using this doc for ref: https://www.mathworks.com/help/phased/examples/ground-clutter-mitigation-with-moving-target-indication-mti-radar.html

I trying to rebuild the code there, I dont need all the definitions of the antennas and signals because I already have file with I/Q pulses example and I need to process them.

I stuck on the part where I should do [Matched Filter] -> [3-Pulse Canceller]
The way I do it, which i told is wrong, but the result just look better for me is:

  1. calculating Matched Filter for every pulse (assuming I working with 64 pulses) and I arrange the filtered data on Matrix like this:

enter image description here

  1. take the matched filter result row by row and entering the 3PC filter like this:

enter image description here

and than continue the processing.. the matlab code:

%% Matched Filter  
mf_buffer_Q = filter(coeff,1,buffer_Q,[],2);
mf_buffer_I = filter(coeff,1,buffer_I,[],2);

%% Three-Pulse Canceller
h = [1 -2 1];
TriPC_buffer_Q = filter(h,1,mf_buffer_Q,[],2);
TriPC_buffer_I = filter(h,1,mf_buffer_I,[],2);

the plot result: enter image description here

the other way to do it is what i think is wrong but I told is the right way is:

%% Matched Filter  
mf_buffer_Q = filter(coeff,1,buffer_Q,[],1);
mf_buffer_I = filter(coeff,1,buffer_I,[],1);

%% Three-Pulse Canceller
h = [1 -2 1];
TriPC_buffer_Q = filter(h,1,mf_buffer_Q,[],2);
TriPC_buffer_I = filter(h,1,mf_buffer_I,[],2);

that means: enter image description here

the plot result for this: enter image description here

what output seems better? what is the right way to do this 2 filters? one horizontal and the second vertical
or I just should do both filters horizontal.

thanks.

$\endgroup$
2
$\begingroup$

The second way is how it is done.

In the fast-time (or range-bin dimension) you are right to perform the matched filter.

However, Doppler information is gathered from sampling pulse-to-pulse. The MTI filter essentially subtracts pulses so that if they have similar phase, or the phase is not changing at all, then there is little to no Doppler being generated and thus the target is not moving.

In other words, the three-pulse canceller applies a filter across the Doppler dimension for each range bin, as the second pulse matrix suggests.

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.