Bulletproof quantum crypto dinged by implementation weakness
Trust but verify
Security researchers have identified possible weaknesses in quantum cryptography implementations. A team from Linköping University in Sweden has also come up with suggestions about how the attack could be blocked.
Quantum cryptography allows two users on an optical fibre network to exchange secret keys. It takes advantage of the particle-like nature of light. In quantum cryptography, each bit of the key is encoded upon a single light particle (photon). Intercepting this data randomly changes the polarisation of the light, irreversibly altering the data.
Because of this quantum mechanics effect, any attempt by an eavesdropper to determine a key corrupts the same key with noise. Quantum cryptography systems discard these corrupt keys and only use codes that are known to be secure. These quantum keys, once exchanged, can be used in a one-time pad.
The technology - long the stuff of cyberpunk novels and hi-tech spy stories - is leaving the laboratory and making its way into commercial markets, particularly in the banking sector.
The Swedish team's work does not undermine the principles that underpin quantum cryptography but rather highlight shortcomings in how systems are implemented in practice, a common source of cryptographic weaknesses. The potential weakness discovered by the Swedes involves how practical quantum cryptography systems authenticate that a received message has not been altered in transit.
Eavesdropping to obtain a quantum key isn't possible, as previously explained. But it may be possible to get clues about this quantum key. Using this information, practical quantum cryptography systems could be tricked into authenticating altered messages, the researchers warn in a paper published in the IEEE Transactions on Information Theory journal last month.
By accessing the quantum channel used in QC [quantum cryptography], the attacker can change the message to be authenticated. This, together with partial knowledge of the key, does incur a security weakness of the authentication. The underlying reason for this is that the authentication used, which is insensitive to such message changes when the key is unknown, becomes sensitive when used with a partially known key.
The research is largely based on a masters thesis by Jörgen Cederlof.
Jan-Åke Larsson, an associate professor of Applied Mathematics at Linköping University and one of the two principal authors of the study, told IT News Australia that adding a small number of random bits to the initial key exchange foils the attack, which he acknowledges would be tricky to pull off even without additional safeguards.
"We weren't expecting to find a problem in quantum cryptography but it is a really complicated system. The security of the current technology is not sufficient," Larsson said. "Authentication does not work as intended." ®
@CTG and AC
I like the JCB line of DoS tools myself, their fibre detecting units seem to be highly accurate....
If you have physical access to the cable you can cut the fscking thing. That would be a pretty effective DoS attack too.
Nearly right, in fact you send Keylength*3 (say) qbits. The first keylength qbits allow us to check if line is being tapped, we share publicly which filter we used and the result of using that filter with other party, if when using same filter we have different readings (more than can be accounted for by noise) we know comms are being intercepted. (This is due to quantum! ie observing the qbit affects it's state.
We can then use remaining Keylength*2 qbits to create the key, we form this by sharing with other party publicly the filters used for each qbit. Where we used the same filter we will both know what the polarisation of the qbit was and hence can use this info to create a key that can then be used for a more standard crypto system (ie a perfectly secure one time pad)