Open-Destination Quantum Teleportation

Roland Piquepaille writes "An international team of physicists has entangled five photons for the first time in the world, reports Technology Research News in "Five photons linked." Why is this important? Because it's the minimum number of qubits needed for universal error correction in quantum computing. In other words, they found a way to check computational errors in future quantum computers. The physicists also demonstrated what they call 'open-destination teleportation,' a way to teleport quantum information within and between computers." "They teleported the unknown quantum state of a single photon onto a superposition of three photons. They were then able to read out this teleported state at any one of the three photons by performing a measurement on the other two photons," adds PhysicsWeb in "Entanglement breaks new record ". This will be used in about ten to twenty years to move information among quantum networks. You'll find more details and references in this overview."

Comments

The Wiki-Tome

[RabidChicken]

For those of us who failed High School physics, from Wikipedia: A qubit (quantum + bit; pronounced /kyoobit/ [1] ) is a unit of quantum information. That information is described by state in a 2-level quantum mechanical system.
To be perfectly honest, quantum computing scares me to some extent. Things like PGP encryption and other very sensitive operations could, quite literally overnight, be blown away and dangerously shift power quickly. Then again we will also usher in a new age of unlimited (well, from a 2004 perspective, matter itself ultimately has a limit for storage and processing) computing that can make engineering in all fields like nothing we have seen before. And, the best part, we will see it in our lifetimes.

Re:The Wiki-Tome

[metlin]

And, the best part, we will see it in our lifetimes.

While I appreciate your optimism, I must tell you that the chances of QC taking a giant leap within the next 25 years is quite low.

Sure, people will build preliminary quantum computation elements, and will perform simple operations. But to have a system comparable to existing computers will take a really, really long time.

For one, the resources needed to perform and control such operations is really expensive, and occupy enormous amounts of space. Even technologies used today to achieve the quantum hall effect (one of the primary requirements if you are building a q.c.) is really primitive. For instance, consider MIT's carbon-nanotube technology -- the problem is that while you can achieve q.h.e., not two systems can be duplicated perfectly. Other methods such as building solid state elements to do this (which is what I work on) have been quite unsuccessful.

That, and the fact that we are yet to develop a good enough quantum error correction system. The thing is that in order for QC to take off big time, other areas (material science, nanotech, theoretical CS and information theory, etc) need to progress significantly.

Sure, you may see some primitive QC within the next 40 years or so. But the probability of you seeing a QC capable of, say, solving Primes in P or one that can play you a DVD is quite low. Just my two cents. And yes, IAAQP (I'm a quantum physicist).

Re:Misunderstanding...

[metlin]

Not in the quantum world. You can transport the data, but you cannot copy the data. This is one of the primary premises of Quantum Computation, covered by the No Cloning Theorem.

Ofcourse, if you are talking about the inherent parallelism in q.c., you are right.

Re:Limited use?

[NonSequor]

This is a little tradition borrowed from cryptography. Whenever you describe some apparatus for transmitting information, you refer to the sender as Alice and the receiver as Bob. Other people have added a bunch of other characters, such as Mallory, who represents anyone who might maliciously try to intercept the message in transit.

Re:This is what a normal person just read above.

[metlin]

It's actually fairly simple. In QC, you can perform any quantum operations on the qubits, but you cannot look at the bits without losing some information. Therefore, what you do is use error correcting codes, by superimposing the quantum states onto a set of photons whose states you observe, but do not use. What they have done here is basically taken the unknown quantum state of a photon onto a superposition set of three photons, and you can find the state of any one photon by observing the other two photons.

This was predicted a while ago by Alexei Kitaev, and Anton Zeilinger had a preliminary demonstration of a basic q.t. system a while ago. I would imagine that this is just an extension of their works.

Re:Faster than Light

[metlin]

You are missing something. This has got nothing to do with faster than light communication, instead it's on how they were able to successfully entangle 5 photons, which is the minimum number needed to implement a universal error correction system in quantum computation.

Teleportation was achieved a long time ago by a bunch of folks at Innsbruck, led by Prof Anton Zeilinger.

Re:Misunderstanding...

[metlin]

(Disclaimer: IAAQP)

Yes. They can transmit the data, but they cannot preserve the data without losing information. This is one of the primary ideas behind Quantum Cryptography, which forbids eavesdroppers from creating copies of the transmitted data.

I'm not talking about approximation -- I'm talking of copying the basic qubit as a function of quantum states -- no two quantum states can be copied, and if this were possible it would result in some funny stuff like causality.

You don't have to believe me, see for yourself - No Cloning Theorem.

Re:What we don't know

[wass]

What we may be seeing is the physical evidence that space and time are not much at all like we think they are.

Actually, this is physical realization of quantum principles that have been known for about 70-80 years. And all of those quantum theories were already verified at the fundamental level. There's no new fundamental physics theory being discovered here, the strangeness of relativistic time/space at the quantum limit (ie, Quantum Field Theory) has been quite well developed and understood for a long time now.

This is more like an applied physics or engineering verification of a quantum applied physicists sketch for quantum error correction of quantum teleportation.

Now if physicsists were able to finally merge gravitation with quantum mechanics, that would be huge and just might float your battleships. But this quantum teleportation is certainly not that at all.

Re:Faster than Light

[metlin]

So here's the idea - quantum entanglement is when you have two quantum states that have to be given in reference to each other, even though the two states maybe contained in elements spatially separated.

But - no useful information can be transmitted between the two systems. This is because the information in itself is given by probabilistic superposition of the states. For instance, you have a Qubit defined as the superposition of states, given by |psi> = a|0> + b|1> - so you can only find out when they are absolute states (0) or (1), and not in between -- and that will not happen at speeds less than the speed of light. In order to find out what state the system is in (in between 0&1), you will need to be able to copy the state, which is prohibited by the No Cloning Theorem.

So, to answer your question - you *may* be able to achieve instantaneous transmission of information, but you can never observe that information in a causal fashion less than the speed of light. Did that make sense? :)

Re:This is what a normal person just read above.

[John Courtland]

Maybe this is better: You have a particle. It has a certain and definite state. However, according to Quantum Mechanics, the act of observing the particle changes the state of it. That's no good because you can't rely on that state now. What you do is 'entangle' the particle with other ones, so that they have the same states, and never perform operations on the 'observer' particles. Then you can deduce the state of the 'hidden' particle by the states of the 'observer' ones.

Re:This is what a normal person just read above.

[metlin]

Oh it does. It's just that upon observation, the state collapses and is no longer useful.

It can have any state, in between 0 & 1 -- just that you are not permitted to know what state it is in.

Re:This is what a normal person just read above.

[rufusdufus]

This and its parent are incorrect.

For the parent: the state of all bits become fixed when observation of any member is read; this is simply a noise correction for what is read, a sort of redundance.

For this: this effect does not supply long distance communication. All it does is supply uncrackable encryption. A signal (probably radio) still needs to be sent in order for information to actually be communicated.

Re:But they are entangled!

[qcomp]

If the particles are entangled, and it observe one of the observer ones, isn't that going to change all of them because they are still entangled?

yes, any observation on a set of entangled particles changes the state of the whole set.
However, if you do it appropriately it does change it in such a way, that (a) your measurement tells you nothing about the unknown state and (b) the unknown state is still encoded in the state of the unmeasured particles.

or do you unentangle them before you observe them?

not before - but the act of measurement disentangles the measured particle from the rest. It may lead to *all* particle being disentangled (e.g., if they were in a state |00000>+|11111> and you measure in the basis {|0>,|1>}) or it may leave the unmeasured particles entangled (e.g., if you measure in the basis {|+>=|0>+|1>, |->=|0>-|1>}).

Can you unentangle particles without changing their state?

no, since the state they are in is either entangled or not, disentangling them implies changing their state.
However, the 5-qubit state may be a *redundant* encoding of another state Psi (of fewer qubits). Then it is possible to change the overall state (either by measurements or normal time-evolution) such that one ends up with a single qubit in the state Psi.
This can be useful, since it may allow to if something has happened to the state encoded *without* learning anything about the state. This is the essential idea of quantum error correction: encode in a big (say 2^5-dimensional) space the state of a two-dimensional system. Detect, whether the state has moved out of this subspace (i.e. an error has occurred) but do it such that you do nott distinguish the two states in the subspace (thus leaving it untouched).

Re:This is what a normal person just read above.

[bytesmythe]

Just for the record, the change in one entangled particle does make an immediate effect on the other. They have verified this in laboratory experiments and concluded that the change occurs instantly, not merely at the speed of light.

The problem is this: you cannot actually transfer information using this scheme, only randomness. This is because when you're making the change in the original particle, you cannot control HOW the change is made.

Let's use pennies as an example, pretending that we can "entangle" them like we can subatomic particles so that if two spinning pennies are entangled, if one stops on heads, the other stops on tails, and vice versa. If you take two spinning entangled pennies, then send one of them a few light seconds away, you have a situtation similar to the way these experiments are set up.

So we have these two spinning pennies... Now let's just stop the one still in front of us. Ok, it landed on heads. Now we know the other has just landed on tails. Yet we have not transmitted useful information because we didn't FORCE the penny to land on heads, we just STOPPED the penny. There is no way of controlling how it was going to end up, so all we have transmitted is randomness. This is great for generating randomness for encryption, but you can't communicate with it.

Also, let's set up a different scenario. We'll say that instead of using the states of the tangled pennies to try to transfer information, we'll just use the fact that we stopped them. Now if we have, say, 1000 total entangled pennies (each side having 500), we can agree on a "pennies stoppped per second" rate that is used to transmit information. If we stop 1 penny per second, it's a ZERO bit, and if we stop 2 pennies per second, it's a ONE bit. This means we can transmit a series of 250 ones, or 500 zeroes. But this is instantaneous, so it violates the idea of faster-than-light communication, right?

Actually, it doesn't. However far apart those pennies are when you set up the communications, the "remote half" had to travel at most the speed of light to get there. So, you do not get any increase in the total communication speed.

(You can read more details about quantum entanglement on Wikipedia.)