Re: [Qirg] New Version Notification for draft-wang-qirg-quantum-internet-use-cases-00.txt

[Joseph Touch <[e-mail]>]

Hi, Rod,

> On Mar 4, 2020, at 2:47 PM, Rodney Van Meter <rdv=[e-mail]> wrote:
> 
> 
> 
>> On Mar 5, 2020, at 0:02, Joseph Touch <[e-mail] <mailto:[e-mail]>> wrote:
>> 
>> Hi, Rod (et al.),
>> 
>> AFAICT, sensing is measuring a property other than that of the transmitter. 
> 
> Sorry Joe, but I have no idea what that sentence means. Can you explain more?

We could say that communication are two parties “sensing” (which is a kind of measurement) each other’s transmitter.

Sensing (in a way that isn’t communication) is a one party communicating with a different party that isn’t intending to communicate. I.e., when you sense the earth, you’re measuring something about the earth even though the earth has no explicit “communication transmitter”.

I.e., 

	Communication = two or more parties explicitly sharing state
	
	Sensing = two or more parties implicitly sharing state

>> 
>> Measuring properties of a transmitter is (IMO) communication. In that way, time sync is comm, not sensing.
> 
> Not following this, either, but maybe a little better…
> 
> I refer to clock sync as use of entangled states for some purpose that interacts with the physical world, as opposed to purely digital communication/computation.

There are many kinds of clock sync. 

If you’re syncing your clock to sunrise on the solstice at Greenwich, England then you are sensing.

If you’re syncing your clock to mine you’re not sensing my clock. You’re communicating with me to explicitly share state,

> I chose “sensing” rather than “cybernetic” or some or term because it’s used to measure the environment, or participate in it, rather than control it.  In a sense-calculate-actuate loop, this is definitely on the sense side, IMO, but feel free to suggest another name. I don’t think “comm” here is enlightening, since all of this involves communication.

Comm is sharing state. You’re not really sensing my clock - you could, but that’s not how most time sync protocols actually work.

You receive messages from my transmitter. That’s comm, not sensing.

> 
>> 
>> (The only difference between QKD and comm to me is how errors are recovered; in comms they’re directly corrected where in QKD the “correction” is to restart)
> 
> Real-world QKD actually invests a *TON* of effort in trying to distinguish environmental and spontaneous errors from an actual eavesdropper. It’s pretty hard :-).

Not sure I know why it matters; an error is an error. The point about eavesdroppers is that they cannot avoid creating a certain amount of errors; if you don’t see that amount, they (statistically) cannot be there.

But seeing a higher amount of errors doesn’t mean they ARE there; it just means you don’t believe (with high probability) that they’re not "NOT there".

> 
>> 
>> Note: that perspective explains why quantum computing isn’t also sensing. It also is consistent with the existing continuum between computing and communication. 
>> 
>> I.e., why is crypto in this list vs. being a special case of communications?
> 
> Since we’re talking about a network, all of these are special cases of communication: communication for cryptographic functions, communication for sensing, and communication for distributed computing.

Communication only needs a FSM; computation takes a TM (Turing machine). Yes, any computer needs to have some communication internally but a comm system cannot itself compute.

Joe