Re: EAP-AKA Test Vectors
From: Pascal Urien (urienptele2.fr)
Date: Sun, 16 Oct 2005 13:09:48 -0400 (EDT)
Dear Mr Melovatsky,

  I greatly appreciate your very pertinent analysis for test #2.
  You are right , the MAC-S value that is wrong.

  MAC-S = f1*(AMF,RAND,SQNms)
  AK2 = f5*(RAND)
  AT_AUTS = AK2+SEQms | MAC-S

The correct value will be included in the next version of draft-urien-eap-smartcard

Please let me known if you find other errors.

Thank you again

Best regards

Pascal Urien

At 15:53 16/09/2005 +0400, you wrote:

Dear Mr. Urien:

I have studied the corrected version of tests for draft-urien-eap-smartcard-08.txt
and faced a small problem while doing test #2. Could you help us deal with it?


As far as I understand, when performing test #2, wrong SQN is set on the
smartcard ("ff9bb4d0b608"). Messages with valid SQN come from the network.
The first message is similar to EAP_AKA_CHALLENGE_REQUEST provided in Test #1
 (SQN="ff9bb4d0b607" is used to form it).

While processing EAP_AKA_CHALLENGE_REQUEST on USIM the contradiction of
SQN is determined. In this case we should send AKA-Synchronization-Failure
with the field AUTS. To compute AUTS, AK and MAC_S are used. To compute MAC_S,
SQN is used. According to 3GPP PP TS 33.102 V4.4.0 (2002-06)  SQNms is used
for this purpose. I suppose that in this case the AT_AUTS value (its MAC-S
part) will be not "BA 85 3F 3C 12 33 01 CF AF 9E C4 E8 71 E9 ", and when using
 SQNms="ff9bb4d0b608" -> "BA 85 3F 3C 12 33 7C D9 24 E7 39 F1 23 69".

Could you tell me what you think about it?


>>//========================================== >>// Test #2 : FULL AUTHENTICATION, WRONG #SEQ >>//========================================== >>Tx: A0 80 00 00 44 > 01 > A5 > 00 44 > 17 > 01 > 00 00 > 01 05 00 00 //AT_RAND > 23 55 3C BE 96 37 A8 9D 21 8A E6 4D AE 47 BF 35 > > 02 05 00 00 //AT_AUTN > 55 F3 28 B4 35 77 B9 B9 4A 9F FA C3 54 DF AF B3 > > 0B 05 00 00 //AT_MAC > C7 00 35 36 66 2D 52 01 B0 11 F2 0F E5 DD 8C E4 > >> >>// AT_AUTS = AK2+SEQ | MAC-S >>// AK2 = 45 1E 8B EC A4 3B >>// SEQ = ff 9b b4 d0 b6 08 (right value = ff9bb4d0b607) >>// MAC-S= 01 CF AF 9E C4 E8 71 E9 >> >>Rx: > 02 > A5 > 00 18 > > 17 > 04 //AKA_SYNCHRONIZATION_FAILURE > 00 00 > 04 04 //AT_AUTS > BA 85 3F 3C 12 33 01 CF AF 9E C4 E8 71 E9 >>90 00



Best regards,

A. Melovatsky


  • EAP-AKA Test Vectors Pascal Urien, August 7 2005
    • Re: EAP-AKA Test Vectors Pascal Urien, October 16 2005

Results generated by Tiger Technologies using MHonArc.