[Nfd-dev] Try NDNCERT (based on Interest-Data exchange) and get an NDN certificate today

Junxiao Shi shijunxiao at email.arizona.edu
Wed Oct 24 13:17:01 PDT 2018


Hi Zhiyi

I asked a crypto expert to look at the NDNCERT protocol, and we found a
serious flaw that would allow an attacker to obtain a certificate. It works
as follows:

   1. Bob owns email address bob at example.org. According to example.org's CA
   policy, it entitles him to request a certificate named /org/example/bob.
   2. Bob sends PROBE command with his email address, and the CA grants the
   namespace /org/example/bob.
   3. Bob sends NEW command with a certificate request CR_B for
identity /org/example/bob.
   Eve intercepts this command, and sends another NEW command with a different
   certificate request CR_E for the same identity, but it contains Eve's
   public key.
   4. The CA generates two different request-ids ID_B and ID_E, one for
   each certificate request.
   5. Bob sends SELECT command choosing the email challenge under ID_B. Eve
   does the same for ID_E. The CA sends two emails with PIN_B and PIN_E to
   bob at example.org.
   6. Bob receives two emails. Since the email template only contains the
   six-digit PIN code and the CA name, Bob has no idea which email corresponds
   to his certificate request.
   7. There's a 50% chance that Bob enters PIN_E to his ndncert-client, and
   sends a VALIDATE command that contains ID_B and PIN_E. Eve intercepts this
   command and prevents it from reaching the CA.
   8. Since PIN_E appears as plaintext in the VALIDATE command, Eve is now
   able to construct a VALIDATE command with ID_E and PIN_E, and send it to
   the CA.
   9. The CA issues a certificate to Eve, despite that Eve does not control
   the email address.

The expert has suggested encrypting the PIN code, so that Eve cannot see
PIN_E, stopping this attack.
What do you think about this, and do you have other solutions?

Yours, Junxiao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/nfd-dev/attachments/20181024/5414337f/attachment.html>


More information about the Nfd-dev mailing list