CCM Mode

[despot at crosswinds.net]

On Thu, 15 Aug 2002, Housley, Russ wrote:

> http://ftp.ietf.org/internet-drafts/draft-housley-ccm-mode-00.txt
>
> It contains a specification for an authenticated encryption mode.

While this merging is alluded to in the OCB paper and elsewhere, I still 
found the idea of the CCM mode interesting. It is taking two separate modes 
and merging them into one. It is performing authentication (CBC-MAC) then 
encryption (CTR), and, while I have not seen the details of the security of 
this scheme, I imagine it is somewhat focused on the notions outlined in the 
Krawczyk papers last year.

I think this "expansion" of modes is a beneficial move. Instead of allowing 
protocol designers to attempt to figure out the proper ways to merge 
authentication and encryption modes, modes are being designed that cover the 
proper use of both. This is a good thing.

Of course, I am not ignoring modes like OCB that use "blended constructs" to 
perform both encryption and authentication. Such modes can achieve the 
benefits of "merged modes" with potentially more efficiency.

-Andrew


---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at wasabisystems.com