2
$\begingroup$

I'm trying to apply the hash function combination method described in https://eprint.iacr.org/2013/210.pdf to a HKDF PRNG construction. Thing is it needs 4 uniformly random parameters/keys chosen by the signer during the combination step, but that's what I was already using the PRNG for so it feels like I'm running into some chicken-egg problem.

What would be the recommended way to generate these keys? Is it at all possible to strengthen PRNGs with this trick?

$\endgroup$
  • $\begingroup$ In general collision resistance is not something that a PRF should be protected against, if I'm not mistaken. $\endgroup$ – Maarten Bodewes Mar 5 '18 at 22:31
  • $\begingroup$ It's about improving pre-image resistance to the strongest of the selected hash functions. The keys should be uncorrelated and are known to any attacker, so it does require pre-image resistance if I understand correctly... $\endgroup$ – user56697 Mar 5 '18 at 22:37
  • $\begingroup$ Simply XORing the results is also not an option, even if collision resistance is not important, the functions may be correlated. $\endgroup$ – user56697 Mar 5 '18 at 23:02

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.