Re: [Cose] Key management for MACs (was Re: Review of draft-schaad-cose-msg-01)

[Brian Campbell <[e-mail]>]

The context around the overhead was signing vs. key agreement - so ECDH-SS
+ MAC vs. ECDSA. I was presuming that ECDH-SS would need a random MAC key
and key wrap it using the agreed key.  So the wrapped key was part of the
overhead I had in mind when I said I thought they'd be similar. But yeah,
okay, a  ECDH-SS with a KDF would be smaller.

On Fri, Jul 10, 2015 at 5:08 PM, Jim Schaad <[e-mail]> wrote:

> The overhead on key management is going to be based a lot on what you are
> doing.
>
>
>
> If you are doing direct shared secret –then you have a kid
>
> If you are doing ECDH-SS – then you need two kids and a nonce (which can
> be very short)
>
> If you are doing ECDSA – then you need a kid
>
>
>
> This means that to be the same size you have 24 bytes to express a second
> kid (which could be implicit) and a nonce (which can be a short as 2 or 3
> bytes).  I don’t think you need that much space – so yes, it would be
> smaller.
>
>
>
> The overhead from the basic messages (MAC vs Sign) is going to be about
> the same in both cases.
>
>
>
> *From:* Brian Campbell [mailto:[e-mail]]
> *Sent:* Friday, July 10, 2015 3:34 PM
> *To:* Jim Schaad
> *Cc:* [e-mail]
> *Subject:* Re: [Cose] Key management for MACs (was Re: Review of
> draft-schaad-cose-msg-01)
>
>
>
> Really, there's no overhead from the key management part?
>
>
>
>
>
> I've admittedly not done the comparison but the message size overhead
> seems like it'd be similar with something like ECDSA (admittedly not RSA).
>
>
>
> ECDSA = 2 * SHA-256 or 32 bytes
>
> Truncated AES-CMAC is 8 bytes (64 bits)
>
>
>
>