We will be attempting to use AFR to characterize both sides of a test fixture to which our card attaches with blind mate connectors (GPO or GMS depending on the card). We will be using semi-rigid cable assemblies with the correct mating connector as an unknown through. My concern is that trying to use a bare open to create the reflection for separation will be prone to error due to radiation. We will need to characterize up to 20 GHz. Would it be viable to create a short or open with a mating connector? Would this need to be characterized itself or would it be enough to be a non-radiating open/short only for use with AFR?