> -----Original Message-----
> From: COSE [mailto:[email protected]] On Behalf Of Adam Langley
> Sent: Saturday, July 09, 2016 4:38 PM
> To: Jim Schaad <[email protected]>
> Cc: [email protected]
> Subject: Re: [COSE] Binding of sender's key in ECDH-ES.
> 
> On Sat, Jul 9, 2016 at 3:40 PM, Jim Schaad <[email protected]> wrote:
> > I need to find out where you went wrong, but that is not the intent.
> >
> > Note that there is both an ECDH-ES and an ECDH-SS version in the
> > table.  One uses the ephemeral key and one uses the static key for doing
the
> key
> > agreement.   The paragraph that discusses these options is the
> > "Ephemeral-static or static-static" bullet above the table.
> >
> > There is a small number of common ECDH routines because, modulo how to
> > find the second key, the processing is the same.
> >
> > If you can identify where you got confused, please let me know so I
> > can look at doing clarifications.
> 
> Thank you for the clarification.
> 
> I came into the document expecting ECDH-ES to bind a sender's static key
(i.e. in
> the style of 6.2.1 of NIST SP 800-56A:
>
http://csrc.nist.gov/groups/ST/toolkit/documents/SP800-56Arev1_3-8-07.pdf),
> so the fault might well be entirely my own. There's much discussion of a
sender
> static key in that area of the document, which fed my incorrect
assumption, and
> there's the label -2 in table 19, which is defined as the sender's static
key for
> ECDH-ES. Should that be for ECDH-SS, or is it a "return address" for the
recipient
> in the case of ECDH-ES?

Now I understand exactly where you got confused.  Using the NIST terminology

ECDH-ES --> C(1, 1)
ECDH-SS --> C(0, 2)

Thus ECDH-ES should NEVER have any of the sender static fields and only have
the ephemeral key field.  Likewise, ECDH-SS should ONLY have one of the
sender static fields and never have the ephemeral key field.

> 
> I'm not sure that anything needs to change in the document, but would the
> expected structure for a sender-authenticated, one-way forward-secure
> message be an ECDH-ES message encrypted in an ECDH-SS message?

It would be an ECDH-SS message.

I will cogitate and see if I can come up with some clarifications.

Jim


> 
> 
> Cheers
> 
> AGL
> 
> _______________________________________________
> COSE mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/cose

_______________________________________________
COSE mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cose

Reply via email to