In the article you cite, the researchers create entangled photon pairs via parametric down conversion. One of the photons is the signal, the other is the idler.
They shoot the signal photon towards the object they need to radar, and wait until it is reflected.
Now in the meantime, they need to keep the idler photon physically inside the lab. In your case, what they could do is they create special reflecting mirrors, and keep the idler photon reflecting, bouncing between the two mirrors.
Now when the signal photon returns, it interferes (the signal and idler are still entangled) with the idler photon, and the researchers can calculate the time and distance for the signal photon, thus getting the distance to the object.
Now in your article it says that this method is only effective for short distances. The reason for that is:
photons can be kept bouncing between mirrors only for a limited time (as per QM, the mirrors not being perfect, will elastically scatter the photon for a while, but since it is all probabilities, after a while the photon will be absorbed)
the photons need to be low frequency (otherwise they would melt the walls of the mirror when absorbed)