custom encryption and decryption

There are lots of good code. But for some reason you have already refused to use it when you decided that you don't want to use "aes or the other ones".

So why don't you like to use one of the established crypto algorithms?

yo per westmerk.

everybody knows the ones like aes and knows how to reverse them. if i have my own they wont know how to reverse and my data is protected more.

"everybody knows the ones like aes and knows how to reverse them"

Ohm really?!
Hmm.
Encryption is not for the common man. Use a well establish, open method and READ why is it safe. Ho, and stop using this sms language. that's a start.

if i have my own they wont know how to reverse and my data is protected more.

The problem is that if you get the code here, it won't be your own.

"everybody knows the ones like aes and knows how to reverse them"

Aha.

"AES has been adopted by the U.S. government and is now used worldwide. It supersedes the Data Encryption Standard (DES),[5] which was published in 1977. The algorithm described by AES is a symmetric-key algorithm, meaning the same key is used for both encrypting and decrypting the data."

Hmm.

The established algorithms are established because people have looked at the code and not seen any insecurities. They are secure because of the access to the keys not the accesss to the code.

Encryption cracking is normally done using statistical methods. Write your own code and make it complex. Have statistical methods unroll it within seconds or minutes. Just because your complexities aren't good enough.

An encryption algorithm have very much in common with a good random number generator (RNG). It needs to permute the data in ways that removes the correlations you have between characters in the original data from the encrypted data. So you can't see patterns in the output even when you know that there are lots of spaces and lots of dual-characters in the input data.

Sorry, but you better invest 10-20 years with pure math and statistics before you decide to work on crypto algorithms. And the 10-20 years better be in the correct areas of math and statistics, picking up the latest knowledge about different methods to analyze data.

"AES has been adopted by the U.S. government and is now used worldwide"

It makes you think, doesn't it? What if the NSA left a backdoor in this cipher?

And what if the Chinese knows about this back door without NSA knowing about it...

en.wikipedia.org/.../Blowfish_(cipher)

I helped implement this in an earler project - seemed easy enough to do and was secure enough for what I needed ( settop box)

=> And what if the Chinese knows about this back door without NSA knowing about it...

=====================================================================================

August 17, 2005
Chinese Cryptographers Denied U.S. Visas

www.schneier.com/.../chinese_cryptog.html

=>

Chinese cryptographer Xiaoyun Wang, the woman who broke SHA-1 last year, was unable to attend the Crypto conference to present her paper on Monday. The U.S. government didn't give her a visa in time:

• On Monday, she was scheduled to explain her discovery in a keynote address to an international group of researchers meeting in California.
• But a stand-in had to take her place, because she was not able to enter the country. Indeed, only one of nine Chinese researchers who sought to enter the country for the conference received a visa in time to attend.

Sadly, this is now common:

• Although none of the scientists were officially denied visas by the United States Consulate, officials at the State Department and National Academy of Sciences said this week that the situation was not uncommon.
• Lengthy delays in issuing visas are now routine, they said, particularly for those involved in sensitive scientific and technical fields.

I really dont think its right forum to post this question but since you have worked extensively in this domain your inputs would be of great help.

Requirement: Avoiding repeated decryption of file/data
Expected behavior of Algorithm:

Example: consider file(f0) on encryption gives encrypted file(ef0) and key(ek0) i.e
f0 ----> ef0 and ek0.(repeating encryption of encrypted data)
ef0 ----> ef1 and ek1.
ef1 ----> ef2 and ek2.
. .
. efn-1 ----> efn and ekn.

On applying decryption on efn using ekn I should be able to get f0 bypassing decryption of (efn-1, ...ef2, ef1). Hence key is expected to have delta value during each successive encryption.

Is there any encryption and decryption algorithm currently displaying this behavior?

Looking forward for your suggestion.