Sentinel-1 is not the best source to create a DEM, because of many reasons (perpendicular baseline, temporal decorrelation between two images, comparably short wavelength…).
What you can expect from your data is already indicated after the interferogram step where you get the coherence as a first quality measure. Furthermore, the interferogram should show clear and topographically-related fringes. If these are not given at an early stage, there is little hope that you will get a proper DEM in the end.
To be honest, I have rarely seen any DEM which was at least comparable to the quality of SRTM or AW3D30.
Please have a look at this commend: Phase to elevation error
Also some important remarks here: Using VV and VH to make DEM
And discussions on error sources
- atmosphere Interpreting Wrapped and Unwrapped Interferometric Phase of Philippine Active Volcanoes
- decorrelation How can we calculate the volume of displaced material with displacement band(displacement_VV)?
Can you please share the image dates of your products and a screenshot of your interferogram and coherence (including color scale)?