Question about passive radar implementation

[nkaramolegos]

Hello, nice work.

I have two questions:

1. Can passive radar detect the distance of the target? If yes, how is this achieved?
2. Why do we need coherent SDRs in the same location? I can understand the concept of two coherent SDRs in different locations (due to TDoA estimation) but for the same location it is not obvious to me

[nkaramolegos]

Hmm for the second question, is this due to the need of reference path?

[mhassell]

Hi,

1. Yes passive radar can detect range to the target based on time-difference-of-arrival or TDoA. If we Rx the reference signal at time T and the reflected signal at (T + T_r) we can derive the difference in range based on knowing: a. the distance between the receiver and b. the speed of light. We can, with multiple measurements, fit an estimate to the position of the object and not just range.

2. Yes, we need a reference path to infer TDoA. If the two Rx units at the Rx site are not coherent then there will be no way to infer TDoA between the signals. Even a slight variance in phase will give huge errors in range.

I don't remember where I found the equations for the range and the position estimates but some googling this morning gave me some places to look:

https://www.mathworks.com/help/fusion/ug/track-using-bistatic-range-detections.html

https://dopplerfish.com/passive-radar/

Note that passive radar is bistatic radar.

[nkaramolegos]

Thank you. Have you measured how accurate is the distance estimation? Also, I believe if you deploy 2 coherent RX pairs (in different location) you can improve the accuracy by using something like triangulation. Have you tried something like that? However, I believe that again the distributed RXs pair should be coherent too which as I know it is difficult task and even more challenging for RTL SDR

[mhassell]

I have not measured the accuracy, I think in the dopplerfish website Max Manning points out that the range estimates were good but the frequency shift was more accurate:

Comparing the experimental results to the predicted results from the plane trajectory, there is a very good match! Due to the nature of the FM signal that was used, the Doppler resolution is more precise than the range resolution.

I have not tried multistatic equations for this, but it probably would be more accurate. If you derive the bistatic equations (hint: the answer is an ellipsoid) you could throw in more coherent Rx units and see where it goes. It would be a system of nonlinear equations where you seek the intersection of multiple ellipsoids (one ellipsoid per Rx unit). The Rx units would, as you said, need to be coherent with the Tx site, and hence would need two rtl-sdrs with the mod on the clocks.

[nkaramolegos]

Thank you. Just to make sure that we are in the same page, the first coherent RTL pair in position1 should be also coherent with the second coherent RTL pair in position2. Correct? Also, the transmitter should be reaching both RX positions.

[mhassell]

I haven't done the derivation for multistatic problems but my first thought is

• position1 coherent with Tx
• position2 coherent with Tx

but I dont think 1&2 will be coherent since they do not Tx waveforms for TDoA calculations. I think all they would communicate would be range estimates to the object for data fusion