MMI banner

Home

Order Online

Contact Dynacrypt

Operations

DynaInformation

DynaHelp

Overview

DynaTools

Setup

Encryption

Decryption

Analysis Tools

Entropy

Roughness

Coincidence

Randomness

Compression

Probability

Text Editing

Event Logging

Administration

DynaSolutions

MMI


DynaHelp
Overview DynaTools Setup Encryption Decryption Analysis Tools Text Editing Event Logging Administration

Measures of Randomness

Dynacrypt® gives its users the option to calculate the Measures of Randomness for any file that will be encrypted using Dynacrypt®. To find out the degree of randomness of the file if encrypted using Dynacrypt® do the following:

1.   Select Open File for Analysis from the File Menu or select the Icon on the DynaTool Bar. A File Browser dialog box will appear as shown below:

2.   Select the file to analyze from the File Browser dialog box and select the OK button.

3.   Select ROM Measures from the Analysis Menu on the Menu bar or the ROM Icon DynaTools Bar.

4.   The ROM Measures Dialog Box will appear showing the ROM Measures if the file is encrypted.

In order for an encryption algorithm to be of use it must produce cipher text that is as close to random text as possible. For a particular language like English, it is known that in the basic theory of coincidences there will be 66 or 67 cases per 1000 pairs of letters where letters drawn at random will be identical. This is known as the Kappa Test.

The calculated Cipher Text File Randomness, &psi(o), produced by the encryption method should be close to the calculated Plaintext File Randomness, &psi(r), and should not be close to the Mono-alphabetic Plaintext Reference, &psi(p).

where &kappa = .0667 for the English language and &alpha = .00390625 or 1/256 for an ASCII distribution of cipher text (since there are 256 ASCII characters), N is the number of characters in the file, and f(i) is the frequency of occurrence of each ASCII character in the encrypted file.

The calculated Cipher Text File Randomness, &psi(o) lies between &psi(r) and &psi(p) where the value for &psi(o) should be close to &psi(r).