Minutes interim-2020-qirg-02: Tue 12:00
minutes-interim-2020-qirg-02-202006091200-00

IETF Note-Well applies(IPR rules, etc)
https://irtf.org/policies/note-well.html

Attendee list  from Webex list

Goals of the meeting
End of the work on Archictecture Principles
Last section (§6) under discussion today
Goals

Goal 1
- Even though quantum data transmission was the first goal, it requires long
distance quantum correlations, but more applications require this too. So the
gool is establish quantum correlations, with the applicable metrics in this
field (fidelity...)

- Add link to use case document

Goal 2
Raja = architecture as open as possible is desirable

Goal 3 support HW heterogeneity
- Understood to be manage at protocol level
- Add that architecture shoudl have capability negociation
discovery/negotiation mecanisms - Steve - does it mean DV and CV qubit? => to
be discussed in the mailing list - could be complicated

Goal 4
- could be merged with goal 3 and then reflected in the Principles part.

Goal 5
- Applications do their own security verification (check source, entanglement)
- Steve: what is network level?
- Security function include the authentication of users beyond detecting
intrusion. => Security is a big topic, to be put at the agenda of July meeting.
See https://arxiv.org/abs/2005.04617 as food for thought.

Goal 6
It is quantum-specific because monitoring cannot be done reading entangled pair.
So one has to resort to specific ways.

Goal 7 - same remark : should focus on quantum-specific aspects

Overall, on the goal section has to highlight with what is different from
classical networks.

What about non-goals?
- what can be done classically?
- Not support a specific application

Principles

1 - building block
- discussion on keeping Bell pairs or expand to multipartite entanglement:
  . Bell pairs are efficient and enough
  . but furure protocols might require more than pairs
  . a solution can be to have an inclusive wording
    - 'entanglement is the building block'
    - 'entangled p-uplets' p= 2 pairs, p > 2 : multipartite

2 - Inditiguishable
- "resource reuse" is misleading
3 - Fidelity
- Fidelity is not just another KPI : it is qualitatively needed above a
threshold
  for the function to make sense (below, classical do the job).
- Does this mean 'error correction' is part of the service? Mention it in the
paragraph.

4 - Time
No comment.

What about the link layer service? Link and network services need to be more
accurately defined, probably in this document.