Simulate multipath bouncing for passive radar sensing

I would like to simulate the passive sensing of radar pulses from multiple radars, which also includes multi-path effects. I have been adopting this MATLAB example involving a single radar and direct path: https://www.mathworks.com/help/phased/ug/signal-parameter-estimations-in-a-radar-warning-receiver.html

I figured out how to include multiple radars, but not multi-path. The phased.ScatteringMIMOChannel function models multipath, but it is designed for one transmitter and one receiver, with a different signal for each transmitting element. In the passive sensing example, there is only one pulse train per radar, while phased.ScatteringMIMOChannel expects a pulse train for each transmitting element.

As with the example, I want to model the full waveforms themselves, not just abstract pulses. Can someone please point me in the right direction? I'm content with the signal bouncing off clutter just one time (radar-clutter-receiver).

Answer 1

[Aditya]

Hi ,

You're correct: phased.ScatteringMIMOChannel expects a signal per transmitter element, which doesn't directly match your scenario (multiple, independent radars, each with their own pulse train, and one receiver). But you can still model multipath (including clutter bounces) for each radar pulse train, with some workarounds.

1. Model Each Path Explicitly

Since you want to model direct and single-bounce (clutter) paths for each radar, you can explicitly sum the delayed, attenuated copies of the radar pulse train at your receiver.Steps:

a. For Each Radar:

• Generate its pulse train (as you already do).

b. For Each Path (Direct, and Each Multipath/Clutter Path):

• Compute path delay (distance/speed of light).
• Compute path loss (free-space or with additional attenuation for clutter).
• Apply Doppler shift if needed (for moving clutter).
• Sum delayed, attenuated, and Doppler-shifted copies of the pulse train.

c. Sum All Radars' Contributions

• Add up all the contributions at the receiver.

2)Suppose you have N radars and for each, you want to model the direct path and M clutter paths:

fs = ...; % sample rate
c = physconst('LightSpeed');

% For each radar
for radarIdx = 1:N
    txSignal = ... % the radar's pulse train
    
    % For each path (direct + multipath)
    for pathIdx = 1:(1+M)
        % Set up path parameters
        if pathIdx == 1
            % Direct path
            pathLen = ...;
            attenuation = ...;
            doppler = ...;
        else
            % Clutter path
            pathLen = ...; % radar->clutter->receiver
            attenuation = ...; % include clutter RCS, etc.
            doppler = ...; % if clutter is moving
        end
        
        % Compute delay in samples
        delay = pathLen/c;
        delaySamples = round(delay*fs);
        
        % Apply delay, attenuation, Doppler
        sig = attenuation * txSignal;
        sig = sig .* exp(1j*2*pi*doppler*(0:length(sig)-1)/fs);
        sig = [zeros(1,delaySamples), sig]; % delay
        sig = sig(1:length(txSignal)); % truncate to original length
        
        % Add to received signal
        rxSignal = rxSignal + sig;
    end
end

You can wrap the above in functions for clarity.

You can refer to following documentation asw ell:

https://www.mathworks.com/help/phased/ug/signal-parameter-estimations-in-a-radar-warning-receiver.html

[black cat]

Thank you for explaining! I also found this example to be helpful as well: https://www.mathworks.com/help/phased/ug/local-and-global-coordinate-systems-example-in-radar.html