[Mathmesh] Quantum Resilience.

[Phillip Hallam-Baker <[e-mail]>]

NB: Please read to the end of this piece before beginning a response.


To expand on the brief comments I made at the BOF on Quantum Computing.

1) Of course Quantum Resilience is a serious concern. Any technology that
could potentially destroy the global trading system needs to be taken
seriously even if there is a less than 1% chance of it being realized
within the next 20 years.

2) The fact that Quantum Resilience is a concern does not make it the only
design criteria in new cryptographic security protocol designs.

My perspective here is that I did my doctorate in computer science in a
Nuclear Physics lab working on a large physics experiment. This leads me to
read the press releases from IBM and Google in a rather different light. Of
course the number of QBits that they can print on a chip is impressive but
their ability to make use of those QBits within the coherence time is quite
another matter. And at the end of the day, the technologies they are using
represent an attempt to scale a modality known to have scaling issues to
the very limit rather than break new ground in modalities (e.g. trapped
ion) that raise those limits entirely.

So while practical quantum cryptanalysis within the next ten years is
possible, it doesn't appear to be very probable. It certainly isn't such a
sure thing that we should drop all crypto protocol development until we
have a suite of quantum-safe algorithms.

3) It is not yet clear which 'post quantum' algorithms are secure, let
alone which are the algorithms that we should choose.

4) It is certainly not clear that the key splitting and combination
approaches used in the Mesh will be impossible.

5) Notwithstanding all the above, the existence of a solid framework
allowing us to make client side PKI practical while it is still effective
could well prove to be an invaluable resource for effecting the deployment
of post-quantum infrastructures.

For example, let us imagine that we establish a shared secret for each
device connected to a user's personal Mesh. Alternatively, consider
establishing a hash signature tree as a means of authenticating future
updates in circumstances where traditional public key signature has failed.

If we hit the type of situation we hit with Internet routing and RED where
we needed to lash together a scheme to save the Internet within 36 months,
then the Mesh provides quite a few of the tools I would want to make use of
in putting out a solution. In particular the fact that every Mesh device a
user connects is constantly re-synchronizing to the user's personal source
of ground truth via a Hash chain/block chain structure (DARE Sequence).

I am very interested in such issues and concerns and looking at ways in
which they might be addressed. But any such effort is very clearly an IRTF
effort that is entirely independent of any work that the IETF might be
doing on the Mesh.

If government or industry are seriously concerned about the threat of
Quantum Cryptanalysis, there is a very easy means by which that concern can
be demonstrated: Fund research into post quantum and Quantum Resilient Key
Infrastructures: QRKI. I would be more than willing to discuss writing a
SOW for such efforts.