5 ms·

Successful Attack Against a Quantum Cryptography System

38 points by soundsop 16y ago

JCThoughtscream 16y ago
Should be pointed out that it was a hack against the hardware, not the theory. Hardware vulnerabilities will probably remain the key weakness of quantum encryption, not that this is necessarily anything new.
- amalcon 16y ago
  You're absolutely correct, but the irony is that the whole point of quantum encryption is to eliminate certain classes of hardware vulnerability. You could skip the whole quantum thing and send a conventionally generated key over a dedicated line (regular fiber or even copper), and you'd be no more vulnerable except to hardware attacks (cable splicing or other eavesdropping tricks).
- eru 16y ago
  I vote for human failings as the key weaknesses in any cryptosystem.
beloch 16y ago
Quantum crypto and classical crypto are both susceptible to faults in implementation. Big freakin' deal. Here's why Quantum crypto is the future: If you encode something via any classical crypto system (e.g. RSA) it's usually based on something that is computationally difficult to crack rather than something that is impossible to crack. You can usually find somebody willing to tell you it will somewhere on the order of thousands of years to crack that message, even assuming Moore's law growing resources. That's utter crap. In reality, algorithms continually evolve such that encryption algorithms become insecure in a matter of decades, and that's without disruptive algorithms/tech. (e.g. Quantum computing will shake things up bigtime if and when it happens, but 4096 bit RSA will probably be toast long before that happens.) Here's where things get scary. Any encrypted message you send today that is encrypted classically can be archived by an eavesdropper without your knowledge and cracked at any point in the future. Your credit card info is going to change before that's likely to happen, but what about your medical records? If information is going to remain sensitive for decades or more, it's already idiotic to protect it with classical encryption. So how is quantum crypto different? Why can't an eavesdropper just archive a quantum encrypted message and crack it at their leisure the way they can with a classically encrypted message? The answer boils down to the no-cloning theorem, which I encourage you to look into if you're interested. To be brief, quantum crypto algorithms rely on exchanges of quantum states that, if an eavesdropper attempts to intercept and copy, will introduce errors that will alert the sender/receiver to the fact that they are being eavesdropped upon. With quantum crypto, an eavesdropper can obtain no information unless she has an attack that works at the time the message is sent. The attack mentioned in the article is no different. This isn't like DES, where the vulnerability means that all past messages sent can now be trivially cracked. It means that future messages might be compromised if upgrades are not made, but that's it. With quantum crypto you only have to worry about attacks that can be made with existing knowledge and technology, not all attacks that could be made in the future. If you send a message via quantum crypto and it's safe when you send it, it will be safe for all time. Not just for a while like with classical crypto. Classical encryption is not safe. It's safe for a while. Quantum crypto is actually safe.
- recampbell 16y ago
  So it sounds like Quantum cryptography is only useful for data being transmitted between two points, and would not apply to data at rest? Or is this a false distinction?
  - sp332 16y ago
    Yeah, that's true. It depends on entanglement of particles (or photons in this case, which are definitely never at rest!), which is hard to maintain for more than a fraction of a second. Edit: If they ever get this working that might change! http://www.physorg.com/news196613216.html http://www.physorg.com/news196613216.html
Blunt 16y ago
help me understand Quantum Mechanics. As I've read, with great interest, QM is a no longer a theory but a provable science that electron spin or rotation is equally opposite in two different places at the same time? What hardware do we use to detect electron spin direction and such... Can anyone explain this in easier terms?
- Estragon 16y ago
  QM is a no longer a theory but a provable science that electron spin or rotation is equally opposite in two different places at the same time? No, this "superpositioning" aspect of QM is still a completely untested theory, al beit one which is accepted without question by most contemporary physicists. It's called the Copenhagen Interpretation. There are other possibilities, and no empirical basis for distinguishing between them, at this stage.
  - eru 16y ago
    Don't most scientist prefer the many worlds interpretation now?
    - sp332 16y ago
      Neither of them is really accepted. Scientists might have a preference toward one or the other, but so far they both have inconsistencies.
      - eru 16y ago
        Both may not be very convincing, depending on your biases. But I found them consistent (especially the many worlds interpretation). Could you explain?
        sp332 16y ago
        I'm not an expert, but this guy is: http://quantummoxie.wordpress.com/2007/04/18/inconsistencies-in-the-many-worlds-interpretation-of-quantum-mechanics/ http://quantummoxie.wordpress.com/2007/04/18/inconsistencies...

38 points by soundsop 16y ago

Should be pointed out that it was a hack against the hardware, not the theory. Hardware vulnerabilities will probably remain the key weakness of quantum encryption, not that this is necessarily anything new.

amalcon 16y ago
You're absolutely correct, but the irony is that the whole point of quantum encryption is to eliminate certain classes of hardware vulnerability. You could skip the whole quantum thing and send a conventionally generated key over a dedicated line (regular fiber or even copper), and you'd be no more vulnerable except to hardware attacks (cable splicing or other eavesdropping tricks).
eru 16y ago
I vote for human failings as the key weaknesses in any cryptosystem.

beloch 16y ago

Quantum crypto and classical crypto are both susceptible to faults in implementation. Big freakin' deal. Here's why Quantum crypto is the future: If you encode something via any classical crypto system (e.g. RSA) it's usually based on something that is computationally difficult to crack rather than something that is impossible to crack. You can usually find somebody willing to tell you it will somewhere on the order of thousands of years to crack that message, even assuming Moore's law growing resources. That's utter crap. In reality, algorithms continually evolve such that encryption algorithms become insecure in a matter of decades, and that's without disruptive algorithms/tech. (e.g. Quantum computing will shake things up bigtime if and when it happens, but 4096 bit RSA will probably be toast long before that happens.) Here's where things get scary. Any encrypted message you send today that is encrypted classically can be archived by an eavesdropper without your knowledge and cracked at any point in the future. Your credit card info is going to change before that's likely to happen, but what about your medical records? If information is going to remain sensitive for decades or more, it's already idiotic to protect it with classical encryption. So how is quantum crypto different? Why can't an eavesdropper just archive a quantum encrypted message and crack it at their leisure the way they can with a classically encrypted message? The answer boils down to the no-cloning theorem, which I encourage you to look into if you're interested. To be brief, quantum crypto algorithms rely on exchanges of quantum states that, if an eavesdropper attempts to intercept and copy, will introduce errors that will alert the sender/receiver to the fact that they are being eavesdropped upon. With quantum crypto, an eavesdropper can obtain no information unless she has an attack that works at the time the message is sent. The attack mentioned in the article is no different. This isn't like DES, where the vulnerability means that all past messages sent can now be trivially cracked. It means that future messages might be compromised if upgrades are not made, but that's it. With quantum crypto you only have to worry about attacks that can be made with existing knowledge and technology, not all attacks that could be made in the future. If you send a message via quantum crypto and it's safe when you send it, it will be safe for all time. Not just for a while like with classical crypto. Classical encryption is not safe. It's safe for a while. Quantum crypto is actually safe.

recampbell 16y ago
So it sounds like Quantum cryptography is only useful for data being transmitted between two points, and would not apply to data at rest? Or is this a false distinction?
- sp332 16y ago
  Yeah, that's true. It depends on entanglement of particles (or photons in this case, which are definitely never at rest!), which is hard to maintain for more than a fraction of a second. Edit: If they ever get this working that might change! http://www.physorg.com/news196613216.html http://www.physorg.com/news196613216.html

Blunt 16y ago

help me understand Quantum Mechanics. As I've read, with great interest, QM is a no longer a theory but a provable science that electron spin or rotation is equally opposite in two different places at the same time? What hardware do we use to detect electron spin direction and such... Can anyone explain this in easier terms?

Estragon 16y ago
QM is a no longer a theory but a provable science that electron spin or rotation is equally opposite in two different places at the same time? No, this "superpositioning" aspect of QM is still a completely untested theory, al beit one which is accepted without question by most contemporary physicists. It's called the Copenhagen Interpretation. There are other possibilities, and no empirical basis for distinguishing between them, at this stage.
- eru 16y ago
  Don't most scientist prefer the many worlds interpretation now?
  - sp332 16y ago
    Neither of them is really accepted. Scientists might have a preference toward one or the other, but so far they both have inconsistencies.
    - eru 16y ago
      Both may not be very convincing, depending on your biases. But I found them consistent (especially the many worlds interpretation). Could you explain?
      - sp332 16y ago
        I'm not an expert, but this guy is: http://quantummoxie.wordpress.com/2007/04/18/inconsistencies-in-the-many-worlds-interpretation-of-quantum-mechanics/ http://quantummoxie.wordpress.com/2007/04/18/inconsistencies...