Using GCP to improve the accuracy of the DEM

hi any one
If I have control points (GCP) well defined. I want to use them to improve the accuracy of (Phase to Elevation@Geocoding) as in the Figure. Any layer of the product through which the points can be used because I do not see the features in all layers.

Do the points have elevation information?

I’m not sure if I understand. Maybe you can describe your problem a bit more, e.g. with a screenshot or a specific question

Thank you for the answer.Yes, I have well defined 3D control points on a set of features. I want to use them to improve the accuracy of the DEM.I do not know the way to use these points because they did not appear in any band of radar images compared to visible images.

In which coordinate reference system are they stored?
It could be that they can only be shown in the radar image after its geocoding. Additionally, as far as I know, the GCPs will not affect the “phase to elevation” progress.
What you can do is to geocode the InSAR elevations, then import the GCPs and remove trends or over-estimations by a linear regression (I discussed this in another topic: here). Do you think that would help?

A first step is to plot the heights from the InSAR DEM against the heights of the GCO points. This will show you if you can increase the quality of the DEM by a mathematical expression to bring the values into a more reasonable range.

thanks for your care.
the coordinate reference system for GCP was WGS1984,EGM2008.
I read in some paperstampubolon-reinhardt_isprs-archives-xlii-4-615-2018.pdf (1.1 MB) .The main missing part in SNAP desktop is the height calibration based on GCP input. Currently the GCPs are only involved in the geocoding step (Step 7) and not in the absolute phase offset estimation (Step 6).

Poor results do not help me search the high accuracy and the use of accurate images on it. Perhaps one of the most asked questions is why we use these products in spite of the high prices despite the existence of free products finer ones.