Processing files

[nkaramolegos]

Hello, nice work. Are all the estimation of distance of the target taking place in the goship.m file? I mean is this the overall code for the passive radar?

[jmfriedt]

Yes that's all: https://github.com/jmfriedt/passive_radar/blob/main/171210ship/goship.m#L94 computes the cross-correlation between the frequency shifted reference channel https://github.com/jmfriedt/passive_radar/blob/main/171210ship/goship.m#L92 and the surveillance channel, and time is converted to distance when plotting the result https://github.com/jmfriedt/passive_radar/blob/main/171210ship/goship.m#L98 (*3e8 for the speed of light, /fs for the time interval between two samples, /2 for the two way trip, /1000 for km instead of meters since the frequency is in Hz).

[nkaramolegos]

Thank you. I am impressed with your overall work in the area. Perfect. Have you measured the range that it can detect targets? For example the maximum distance that it can detect ships from the coastline?

[jmfriedt]

At some point we will end up being limited by the curvature of the Earth ! In this experiment we had the antennas on the beach at sea level facing the sea, so that a 20 m [1] high ship would vanish under the horizon at a distance of (acos(6400/(6400+0.02)))*6400=16 km (assuming a 6400-km Earth radius). At the bottom of the videos of https://github.com/jmfriedt/passive_radar/tree/main/171210ship you will see some dots switching on and off at 8-km range at least so that sounds realistic since grazing angle illumination is probably not the most favorable condition.

[1] somewhat optimistic according to https://transportgeography.org/contents/chapter5/maritime-transportation/evolution-containerships-classes/ which seems to refer to 13-m high container ships leading to (acos(6400/(6400+0.013)))*6400=13 km

[nkaramolegos]

How easy would be to use antenna arrays to rotate the receive beam for both target and reference signal? Would this affect the software of the passive radar? if I'm not missing something, it should not affect the equation and the most of the work is on the beamforming and collecting the signal to feed to the .m file

[jmfriedt]

If I understand your question correctly, this is what we did in http://jmfriedt.free.fr/dvbt_hardware.pdf figure 5. In this article I argue that in order to keep the time delay calibration valid for the whole measurement, the dual SDR receiver system must never stop streaming so I use a 0-MQ publish (UDP like) data transfer from GNU Radio to the Octave processing software, which only records data once the antenna has settled to a stable azimuth and throws the data away otherwise.

[nkaramolegos]

Yes, in this case the antenna azimuth was changed by hand. I am thinking if it is possible to do something like beam-forming/beam-stearing. But as far as the passive radar we are ok.

[jmfriedt]

You will need some spatial diversity for beamforming. This was done by moving the ground-based passive radar receiver by W. Feng in http://jmfriedt.free.fr/IGARSS2019.pdf and by using the motion of the spaceborne source in http://jmfriedt.free.fr/SDR-implemented_passive_bistatic_SAR_system_using_Sentinel-1_signal_and_its_experiment_results.pdf

[nkaramolegos]

I'm unsure if I fully grasp the concept. My idea, perhaps simplistic, is to employ an array of stationary antennas rather than physically reorienting the receive antenna to target a specific position. The approach involves utilizing beam-forming or beam-steering techniques to direct focus towards the intended target (during RX). Am I overlooking something critical?