Re: Encryption on the AS/400

["Nathan M. Andelin" <nandelin@xxxxxxxxxxxxxxxxxxx>]

An interesting subtopic of encryption is cracking tools.  Software utilities
that help you discover the basis of encryption.  This assumes that an
adversary has captured the encrypted data.  The use of a text or hex editor
to display the data may be the first step.  But even cracking weak
encryption usually requires something more sophisticated.

Subsequent steps will likely be determined by whatever basis the cracker
has.  Is the algorithm known?  Is the originator of the data known?  Is the
purpose of the data known?  Is the reason for the encryption known?  Is it
known whether the data originated from an ASCII or EBCDIC system?  Is the
language of the original data known?  It seems to me that the more a cracker
knows, the better his basis, and the fewer steps he must conquer.

A subsequent step may be to read the encrypted data, convert each byte to
it's integer equivalent, and chart the results where Y represents the
integer value and X represents the byte's position in the file.  With some
algorithms a visual pattern may emerge.

If the algorithm is unknown, then the next step may be to determine if the
encrypted data reveals a signature.  A signature would be something in the
data that is associated with a known algorithm, or a type of algorithm.  The
cracker may turn to utilities that implement mathematical and statistical
models to discover any signatures.

My thinking is that with strong encryption, the final step is probably brute
force.  A utility that takes what is known and fills in the blanks with
every conceivable possibily.

Of course, if you have criminal intent, it may be a lot easier to threaten
the person who holds the keys to decryption.

Nathan M. Andelin
www.relational-data.com