Stop brute force attacks

View: New views
2 Messages — Rating Filter:   Alert me  

Stop brute force attacks

by BobGeezer :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Hi,

DES Encryption is now viewed as weak having only 56-bits of key information.  However, keys are only weak if people can either guess them, or find them by brute-force.  In either case, the attacker has to be able to recognise the coded information when they happen across it.

Therefore scramble the content before encryption.  A suitable scrambler is a PRBSR convolutional coding using a pseudo-random key (derived one-way from the current time and encryption key) stored in the message.  This removes any signature in the message preventing and anti-DES machine spotting when it has the right key.  Then DES is good enough surely?

Regards, Bob Geezer

Re: Stop brute force attacks

by Nicholas Jordan :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

could be, for non-commercial or something. Often the human factors far exceed such matters and the idea is not to be discounted in an overal analysis ..... but what we get into is some sort of idea of what the information is supposed to look like is seemingly needful for brute-forcing to be effective. Scrambers or other approaches are of utility in resisting brute forcing, for example Clyde as in Circus could be scrambled prior to D.E.S.'ing ... eventually arriving at a point where even the cryptographer has no idea what is going on and becomes like Hollywood, where you cannot even talk to your own attorney.

I am not trying to be funny, I have lots of experience with this on large projects in the real world involving virtually no computer cryptography. I can tell you that I would enjoy someone scrambling protected matters a little before real protections are emplaced. It is just that I have been suprised by what a mathematician can extract from seemingly obliterated datastream.