Re: [Qirg] Survey paper about the Quantum Internet Protocol Stack

[Joseph D Touch <[e-mail]>]

Hi, Alex,

...
On 6/29/22, 5:14 AM, "Alexandre Petrescu" <[e-mail]> wrote:

    Le 27/06/2022 à 14:35, Joseph D Touch a écrit :
    > Hi, Alex
...
    > Avian carries does not introduce an intermediate address at the network layer. Its addrsses would be no different to IP than are Ethernet addresses; they represent the link addresses.

    A pidgeon equipped with a small computer (maybe a small ARM or STMicro
    board) would need to have an IP address.  

In the intended example (from the RFC), it acts just like an Ethernet or other link layer, as follows:

	- pidgin coops would have some 'link address' that the IP packet would be sent to, in anticipation of using the pidgin for message transfer
	- pidgins act like photons; they transfer the message

    ...The source desktop PC would
    record data on the pidgeon computer, by sending it IP packets.  Pidgeon
    would then fly.  At arrival, the destination desktop PC would read the
    message from the pidgeon computer's IP address.

Let's imagine the approach you're suggesting. That's no different from a router having an IP address. It has nothing to do with the address of the messages that go through the router, though. The messages through a router arrive at a router using the router's link layer address, not its IP address.

    This is src-intermediary-dst communication that is different than
    src-dst communication.

Your model does not represent the Internet.

    This 'intermediary' is actually the flight.  The flight does not involve
    electromagnetic waves.  It is some kind of communication that is outside
    the scope of classical communications.

Strictly, we could transfer either classical or quantum via physically moving fermions (see Rod's now infamous 'container ship' approach). For classical nets, fermions remain in use in many places, but for quantum nets they appear to be more straw-man arguments.

    Because of this blackbox nature of the flight (something happens during
    the flight, but we dont really know what, from the em standpoint), one
    might assimilate the avian carrying to the quantum communications.
    There too something happens during the reading of a particle that the
    counterpart reads the same - but we dont know what from an
    electromagnetic wave standpoint, yet they do communicate.

    > 
    > Additionally, IP isn't source-destination only; both IPv4 and IPv6 have 'source route' options.

    Fair enough.  However, the source route concept of IP does not change a 
    higher level's protocol workings, such as TCP.

Yes, that's by design.

    On another hand, with src-intermediary-dst communications the normal TCP 
    would not work.  There would be two distinct TCP connections, long 
    separated in time - maybe 1 day.  One TCP connection between src and 
    intermediary, and one between intermediary (pidgeon) and dst.

That's only because you created a model that doesn't represent Internet IP packet forwarding; it represents proxying, as you note.

If the IP packets arrive at a router, the router is moved, then the IP packets are emitted, the router need not have an IP address and there need not be two TCP connections.

Over that net, which does model the pidgin RFC, TCP would work fine; it would be slow, but it would work.

    >
    >      Further, it can be considered that the data carried by a pigeon would be
    >      encrypted, and if that used a quantum-resistant algorithm then we had an
    >      approach between quantum networking and avian carriers.
    > 
    > This group and discussion has been focused on quantum systems, not quantum-resistant classical systems.

    I agree then.

    > 
    >      >> I suspect the qubits dont need electromagnetic waves in order to
    >      >> appear elsewhere (aka 'to be transmitted'), hence there is a
    >      >> difference.
    >      >
    >      > They do too. The collapse of entanglement of the superposition can
    >      > happen later, BUT it also always relies on classical communication to
    >      > be meaningful.
    > 
    >      This is entirely unclear to me, I dont know(?)
    > 
    >      I thought, really, that quantum communication did not involve
    >      electromagnetic waves at all, but some other way of transmitting data.
    > 
    > Photons are EM waves.

    I agree photons are EM waves in some experiments.

    But I meant this 'entanglement' mentioned in draft-irtf-qirg-principles.

    I understand this entanglement works without photons, and hence without 
    electromagnetic waves.

    I understand this entanglement is able to communicate information from 
    one place to another.

That is incorrect. Entanglement can coordinate qubits, but they cannot be used to communicate information. Information communication requires that state can be controlled; entanglement requires that the state be 'random', i.e., not controlled.

There is no way to use entanglement for classical communication.

    Hence - quantum networks (entanglement) dont use electromagnetic waves, 
    whereas classical networks do.

    Hopefully I dont miss something.

    > And the "classical communication" I'm referring to is the classical network used to coordinate the quantum one, i.e., to exchange basis pairs, coordinate bell pair use, etc.

    For that 'classical communication' network used to coordinate a quantum 
    network - I agree it uses electromagnetic waves.

    Do you mean that a practical quantum network necessarily involves a 
    classical network to coordinate it, and hence a practical quantum 
    network necessarily involves electromagnetic waves, indirectly?

No; as per above, a quantum network alone does not perform classical communication.

Joe

    Alex

    > 
    > Joe
    >   
    >