0
After how many years submissions are accepted for a new AES? Few months back I heard the news of attacks on AES and it is suggested that AES-128 bit is more secure than AES-256-bit.
RPK
Posted 2011-01-02T15:30:55.800
Reputation: 2 293
What is the minimum time-period for selecting a new AES?
Answers
2
For details read this: http://www.schneier.com/blog/archives/2009/07/another_new_aes.html
Bruce Schneier says we're safe...and I trust that.
I'm not seeing a "new" aes until this one is woefully broken
One more thing, to be clear AES 192 still takes longer to break that AES 128 or AES 256 per Schneier.
"In the case of AES-128, there is no known attack which is faster than the 2^128"
and
"However, AES-192 and AES-256 were recently shown to be breakable by attacks which require 2^176 and 2^119 time, respectively."
He goes on to say that there are related key attacks that really make AES-256 breaking almost feasible.
RobotHumans
Posted 2011-01-02T15:30:55.800
Reputation: 5 758
1Do you have a source for that claim (128 better than 256)? – AndrejaKo – 2011-01-02T15:35:51.653
@AndrejaKo: In the article provided in the post below, Bruce Schneier writes: "And for new applications I suggest that people don't use AES-256." – RPK – 2011-01-02T15:40:30.833
1@Rohit Yeah, I'm reading it right now. Also, wouldn't new encryption standard be better name for this question? AES already refers to a precisely defined standard and new standard would need to have its own name. – AndrejaKo – 2011-01-02T15:41:12.837
1@Andrej True enough, but they'll probably call it AES2 or something like that when they get around to taking applications for the competition. Personally, I think the best solution is to use two crypto-systems wrapped around one another. Two-fish and AES for example. Another point worth noting is that they probably won't do this until x64 is a government standard as AES is currently optimized for 32bit words and optimizing for 64 just makes sense at this point. – RobotHumans – 2011-01-02T15:48:52.907
@aking1012: I remember, long back I asked this same question to a cryptographer and he replied using multiple layers of algorithms may sometimes weaken the previous encryption. – RPK – 2011-01-02T16:30:06.683
1interested in a reference on that...i highly doubt it is often the case. there may be a corner case where that occurs, but the guys at the german cryptophone company are really good at what they do and they wrap AES128 in twofish. the only way i could see that statement being true is if you used the same key material or one had predictable output – RobotHumans – 2011-01-02T16:36:05.000