Re: [IPsec] IPsec with QKD

[Rodney Van Meter <[e-mail]>]

On Oct 31, 2014, at 7:49 AM, Greg Troxel <[e-mail]> wrote:

> 
> I implemented using QKD material for IPsec (in 2002, if I can remember
> the years right), which included an interface between QKD processing and
> racoon (on NetBSD) to include QKD bits in the Phase 2 hash.  What's been
> published on that effort is:
>   http://dx.doi.org/10.1145/863955.863982

Yes, we’re very familar with your work, the first in the world, as far as I’m aware!  It was one of our inspirations, although our actual implementation was done using QKD devices we borrowed from acquaintances at NEC.  I’ve known Chip since about 2004.  My memory fades already, but when we began this project in 2008, if memory serves, we inquired about simply using your code, but for some reason that wasn’t possible.  I think at the time Chip felt like he was unable to release it, but I can’t remember the details.

When we first wrote down what we had done (even before the I-D -00 was published), Chip looked at our technical approach and endorsed it as more flexible than what you guys had done, which I believe involved borrowing some bits in existing headers rather than creating new payload and transform types.  I don’t recall whether Chip asked you about it directly at that time.  I think Dave Pearson may have looked at it.

FWIW, as long as we’re talking history of projects, Sakane-san, one of the key implementers of racoon2 if I have it right, kibbitzed with us a bit on this project.

		—Rod