Dear Elwyn Davies
Thanks for your comments.
Here is my answer (<<) and send the revised draft.
Please check it again to finish this process, ^^
BR,
Seokung
----- 메일의 원문 -----
Sent: Fri Apr 17 07:27:39 KST 2009
To: David McGrew <[email protected]>
Cc: 윤석웅 <[email protected]>, [email protected], [email protected], General Area Review Team <[email protected]>
Subject: Re: Request for comments regarding SEED-SRTP
Hi, Seokung.
For section 2.2, I *still* think it would be good to be totally explicit
as I said: Something like:
- At the sender: perform step 7 before step 5.
- At the receiver: perform the second half of step 5 (perform
verification) after step 6.
This means that authentication is performed on the cleartext rather than
the ciphertext.
>> I reflected.
Presumably this applies equally to SRTCP? You should say so if it is true.
>> I reflected.
It occurs to me (belatedly) that this reordering apparently increases
the processing that has to be done on packets before the authentication
is verified, providing a greater impact for a DoS attack that modifies
the encrypted payload. I think this should be noted in the security
considerations (assuming I am correct). Is this a significant problem
given encryption is mandatory on some packets?
>> I already mentioned the security consideration to use GCM and CCM
like : See [RFC3610] and [GCM] for security considerations
regarding the CCM
and GCM Modes of Operation, respectively.
The other problem was fixed.
Regards,
Elwyn
David McGrew wrote:
> Hi Seokung
>
> It looks like almost all of my comments were addressed. I did notice a
> couple things, though. The first is important, the second is very minor.
>
>>
>> There is no definition of how CCM and GCM are to be used to protect
>> RTCP. It would not be possible to use this specification to protect
>> RTCP packets with those algorithms.
>
> The new draft says "In case of SRTCP, the Authentication Portion
> described in Fig.2 of [RFC3711] is treated as AAD." Is that true even
> if the E bit is equal zero? I suspect that what you want here is to
> have a definition such that, when E=0, the AAD format for SRTCP is
> specified differently. (Either that or a restriction to not use E=0.)
>
>>
>>
>> In Section 2, I don't understand the meaning of "and valuables" and I
>> suggest just removing the phrase.
>>
>
> I notice that the word is still there. Perhaps "variables" was the
> intended word?
>
> Apologies for the lag in my getting this review done.
>
> Best regards,
>
> David
>
>
> On Apr 2, 2009, at 6:14 AM, 윤석웅 wrote:
>
>> Dear David Mcgrew and Elwyn Davies.
>>
>>
>> Thanks for your lots of comments during AD last call.
>>
>> I really appreciated it.
>>
>>
>> Here is the revised draft to reflect whole your comments.
>>
>> Please check if thers is any problem or deficiency.
>>
>>
>> If you are OK, I will submit it to proceed next step.
>>
>>
>> BR,
>>
>> Seokung
>>
>> <draft-ietf-avt-seed-srtp-10(alpha).txt>
>
![]()
draft-ietf-avt-seed-srtp-11(alpha).txt