Let's open an issue for the choice of PRF.
Originally posted by @franziskuskiefer in #11 (comment)
ScrambleDB requires a PRF / PRP to convert "raw pseudonyms", i.e. group elements that are output by the coPRF into "hardened pseudonyms". We should make a well-motivated choice.
