NLCSystems

Discussion in 'other software & services' started by DmitriyIT, Sep 26, 2004.

Thread Status:
Not open for further replies.
  1. DmitriyIT

    DmitriyIT Registered Member

    Joined:
    Sep 26, 2004
    Posts:
    1
    I have found info about new non-linear dynamics cryptosystem for file protection, etc. at http://www.mercurysoftworks.com/nlc.htm. It showed to be using principially new algorithm for encryption, allowing to customize security strength from "fast, but normal" with 500 Kb/sec to "incredibly strong" with 1 Kb/sec ! Maybe its good to look at it...
     
  2. Paranoid2000

    Paranoid2000 Registered Member

    Joined:
    May 2, 2004
    Posts:
    2,839
    Location:
    North West, United Kingdom
    Welcome to the forums DmitriyIT,

    Since this product is doing encryption rather than acting as a firewall, I'm sure a moderator will move this thread to a more appropriate forum.

    While I do not claim any real expertise on encryption, your method (adding a pseudo-random "noise" signal to plaintext) has been considered a trivial example for cryptanalysis and your paper seems to have some statements I think are errors. Specifically you use (approximations of) differential functions in the encryption process but not the reverse - integrals - for the decryption. While there are functions where taking the second or higher differential returns the original function, these are uncommon and do not produce chaotic results.

    In addition, all the existing algorithms (that are currently considered secure) work by repeating operations (called rounds) on the plaintext to produce the ciphertext. Your algorithm seems to work by just doing one round.

    Are you making the source code available for this? If not, then how can people be sure that you have implemented your algorithm in a secure way? (e.g. using a strong RNG, preventing copies of the key or plaintext from being written to the Windows swap file, etc).

    While I cannot prove or disprove the effectiveness of your program, I would point out that the existing algorithms (Triple-DES, Serpent, Rijndael) have undergone extensive analysis and it is always safer to use techniques that have undergone the test of time rather than a new one with potential undiscovered loopholes - suggesting that your program is better used for experimentation than for real security.

    For more on selecting encryption algorithms, I would suggest anyone interested check out NIST's PDF document Report on the AES Selection (which details the criteria used in choosing the AES algorithm) and the Slashdot thread On Choosing Encryption ....
     
  3. Detox

    Detox Retired Moderator

    Joined:
    Feb 9, 2002
    Posts:
    8,507
    Location:
    Texas, USA
    Moved to the appropriate section.
     
  4. Bubba

    Bubba Updates Team

    Joined:
    Apr 15, 2002
    Posts:
    11,271
    Hi DmitriyIT,

    I'll assume for a moment you are Dmitriy Tabakov....and if my assumption is correct....it seems you developed this program NOT found the info.

    Wilders TOS/Privacy policy states that....Spam, flooding, advertisements, chain letters, pyramid schemes, and solicitations are also inappropriate in this forum.

    Wilders Security Forums - Terms Of Service

    If my assumption is in-correct....Please feel free to IM me and we will discuss the matter.

    Having said that....I'll close this thread and leave your post intact for review by the members in case they wish to discuss this program in another thread.

    Found on another Forum---> New cryptosystem

    Code:
    DmitriyIT....2 posts
     I am Dmitriy Tabakov, Russian Federation. Following text describes
    cryptosystem NLCS (Non-linear crypting system) developed by me and my
    colleagues from MercurySoftworks. I showed the idea of the algorithm on the
    conference on non-linear dynamics; from that point of view I get first rank
    diploma, but I want to know how it look from the cryptography’s position. And
    also I thought about an opportunity of shareware selling of program based on
    this cryptoalgorithm. And please excuse me for my English…
     
Thread Status:
Not open for further replies.