Author Topic: Simple Encryption Algorithm  (Read 8560 times)

jj2007

  • Member
  • *****
  • Posts: 10556
  • Assembler is fun ;-)
    • MasmBasic
Re: Simple Encryption Algorithm
« Reply #15 on: November 28, 2015, 05:22:20 AM »
The effort took almost 2000 2.2GHz-Opteron-CPU years according to the submitters, just short of 3 years of calendar time.

Could be worse ;)

xanatose

  • Member
  • ***
  • Posts: 398
Re: Simple Encryption Algorithm
« Reply #16 on: December 01, 2015, 01:30:02 AM »
Assuming binary file output and that the size is large enough to bother:

1. Compress the file with a custom header. The result will look like garbage unless you know the header you are using. (for a good compression library http://7-zip.org/sdk.html).

2. Encrypt the result with whatever encryption you want.

That means that the attacker would need to know how to decrypt the file. And how to uncompress it.
Is important that the compression takes place BEFORE the encryption. Otherwise the file will not compress.

To recover the file.
1. Decrypt the file.
2. Decompress the file.

Not perfect, but should be enough for most cases. Not the NSA. But most cases.

K_F

  • Member
  • *****
  • Posts: 1609
  • Anybody out there?
Re: Simple Encryption Algorithm
« Reply #17 on: December 01, 2015, 08:45:50 AM »
All it takes.. is one person to think of something different.. and there's panic in the 'control room'.

PGP is the perfect example.. one person with an idea - and the CIA went ballistic (probably coz they were stupid then ;) )
Now there are 6 billion+ people around with as many ideas.... enough to cause CIA/KBG/Whatever good many well deserved heart attacks.
 :biggrin:
'Sire, Sire!... the peasants are Revolting !!!'
'Yes, they are.. aren't they....'

hutch--

  • Administrator
  • Member
  • ******
  • Posts: 7553
  • Mnemonic Driven API Grinder
    • The MASM32 SDK
Re: Simple Encryption Algorithm
« Reply #18 on: December 01, 2015, 09:39:47 AM »
I have a highly cynical view on popular encryption methods, if it has a bit count, its already dead. Comparisons to some clapped out old Opteron is not telling you what is going on. National security agencies all over the world have dedicated hardware, massive multi-parallel processor capacity for brute forcing ordinary methods of encryption. There are not that many decent alternatives but any that are vaguely useful are conceptual, not just winding up the bit count.

Anyone who has ever done cryptic crosswords (my mum was a whizz at them but I hated them) knows that the more abstract the association, the more difficult it is to get the result. Then you can have single symbols that represent phrases,  non word character collections that represent ideas and there is a massive range of variation available to do things like this. My favourite that can at least be done on a computer is high quality random pads that are then simply xorred to the text message. The high quality random pad is very large and must be securely transferred to both ends of the communications but once it is done, its very hard to beat.

Look at the stats for every character having the range of 0 to 255 and with no consistent method of transferring one character to another. 1st character is one in 256, second character is one in 256. Just with 2 characters you have the range of 256 * 256. Now this accelerating level of complexity comes with a couple of provisos, the reason why you need a massive random pad is so you can always choose an unused part of the pad for the encryption as reuse of the identical pad weakens the security of the technique. The other consideration is verification of attempts to break the encryption, if the encrypted data has no format or header then there is no routine way of testing if there has been a successful result.

Now xanatose has a good idea that makes verification even more complex, if the data is scrambled with compression then even a successful breaking of some or part of the encrypted data still looks like garbage. Using a different encryption method before running the data against a random pad will do much the same.

Now if you were a Mossad agent prying secret data out of Iran, you would take a CD of the speeches of the Ayotollah Khomenie  as your pad, if you were a Canadian business spy stealing trade secrets from the US, you would take a CD or more probably a blue ray data disk of Joe Biden's speeches on foreign policy and use that for your secret transmissions. The drift is to do something that is different and you may keep them in the dark for a millenium or two.
hutch at movsd dot com
http://www.masm32.com    :biggrin:  :skrewy:

MichaelW

  • Global Moderator
  • Member
  • *****
  • Posts: 1209
Re: Simple Encryption Algorithm
« Reply #19 on: December 01, 2015, 12:07:10 PM »
There is an improved version of TEA here, but it looks like an efficient assembly implementation will require more registers than you have available in 32-bit code.
Well Microsoft, here’s another nice mess you’ve gotten us into.

Farabi

  • Member
  • ****
  • Posts: 969
  • Neuroscience Fans
Re: Simple Encryption Algorithm
« Reply #20 on: December 01, 2015, 07:41:18 PM »
I think all webserver is secure enough. I saw M-SQL server, I had no idea how to break in. So if I were CIA, NSA, KGB, I`ll focused on brute-force. But I saw social engineering can still be worked. I was wondered how to make it more secure, from social engineering.

I can see that lots of mistake was
1. Distraction
2. Trust
3. Appearance.

That is my personal weakness. I thought a good tablet was that have a screen that can see from any direction, but I was mistaken. My old Toshiba laptop, which I own at 1999 was having a good screen. Where no one can see it from another angle. No wonder my dad bought it.
http://farabidatacenter.url.ph/MySoftware/
My 3D Game Engine Demo.

Contact me at Whatsapp: 6283818314165