Slashdot Mirror
Quasars Used for Encryption
space_mongoose writes "According to this NewScientistTech story, intergalactic radio signals from quasars could emerge as an exotic but effective new tool for securing terrestrial communications against eavesdropping"
The idea is that its impossible to record all of the transmissions of a quasar to retroactively decrypt a previously seen message (because the quasar is emitting 10^9 or 10^12 random bits/sec)
RTFAA they use them to extract random data(noise).Good for one-time pads.
The KGB used CBR (Cosmic Background Radiation) to produce reams and reams of codebooks/ciphers. They would create two copies, and dispatch one of the copies to the remote location for encryption, then keep one copy at Lubyanka Square.
Obviously, if one side of the cipher was intercepted, then the communication would be suspect - but for most communication, it was the most secure available to them. I don't see this quasar issue as being much different than that.
Now, if they were using quantum states to dynamically generate the ciphers in two seperate places at the same time, THAT would be something to behold.
All one time pads are recorded from random data. You record a long stream of truly random input, then make two copies of the recording. Tne sender gets one copy, the receiver gets the other. Starting at the beginning of the pad, the sender uses each bit of the pad exactly once, then discards it. When the sender runs out of bits, he can not send any more data. The receiver decrypts decrypts likewise, discarding each pad bit after it has been used once. As long as the sender and receiver start with the same pads and don't skip or reuse any bits, they stay in syncronization.
Many perfectly good one time pads are drawn off of data "that anyone can record." For example, many pads are created from atmospheric noise. Anyone can record the same data, but unless you know exactly where and when the recording was done, it is computationally infeasible to record all possibilities, let alone brute force them.
There are many, many quasars that we record in the sky. All of them give off constant streams of random data. So it would be computationally intractable to record all possibilities or brute force a particulr message, because the attacker would have to know exactly which quasar was recorded, and exactly which instant the recording began. He would also have to know exactly which bit of the pad the sender was on when the sender started sending the message that he intercepted. All theoretically possible, but computationally intractable.
The preceding comments reflect the author's personal opinion and are public domain, unless explicitly stated otherwise.
For one thing they are just using the data to create random numbers .. i.e. all.
Logic being that, any random number we create is *not* random. i.e all.
And most (all??) encryption (RSA onwards) is based on random numbers.
Even if there is only one quasar the freq of the signal at any time would be very random indeed. Even if it is as created using known scientific functions, chaos theory predict that we wont be able to regenerate the same any time soon.
For that matter, they could even use any of the natural process to generate random numbers -
Amp of any specfic freq from our dear sun, to , say the power of wind blowing outside itself, would make a good random number
They just chose an exotic one. i.e. all
rajmohan_h@yahoo.com
They are indeed unbreakable, with a theoretical proof of unbreakability -- in the land of spherical horses, where you're allowed to make huge assumptions.
One underappreciated assumption about one-time-pads is that the recipient will (and can!) destroy the keying material after use so thoroughly that the adversary can't reconstruct it. There are several other issues, of which key distribution is one of the easiest. Just put a 500GB external drive in the diplomatic bag once and you've covered communications for a long time.
Here's the problem. The only things secret here are which quasar (13, 14 bits of uncertainty), when the sampling started (?? There won't be very many possible seconds that the adversary has to scan but sampling could start on a fraction of a second), and the sampling algorithm (but you have to assume in crypto that the adversary knows your algorithms). It's going to be easier to brute-force than a 6-word Diceware passphrase unless atmospheric effects somehow make the quasar signal look different everywhere on earth.
There are two problems:
Having a source of noise such as a disconnected sound card or a CCD sensor in a black box can be useful in cryptography if you want to generate truly random bits. But this is not the only thing that this article is about: the signal from the quasars can be received by both parties, which provides a good one-time pad.
-Raphaël
From the article:
/dev/random, which, although influenced by random events, is largely computed using pseudo-random processes. The data stream from each quasar would be broadcast from a single collection point for all to use, but this to me brings up trust issues. Would you trust that single point of contact? Additionally, with enough horsepower, someone could simply record the bytestream coming from the quasar, and perform an off-line brute-force attack on any encrypted data with relative ease.
Umeno believes astronomical cryptography could appeal to anyone who requires high-security communications. He adds that the method does not require a large radio antenna or that the communicating parties be located in the same hemisphere, as radio signals can be broadcast over the internet at high speed.
Basically, he just wants to use the quasars as a source of true randomness, instead of
It may look like I'm doing nothing, but I'm actively waiting for my problems to go away.
--Scott Adams