Page 35 - index

input and output. For special functions used, for example linear functions, this problem may
be solved efficiently. On the other hand, that can not be the case for general functions,
because then we deal with much complex mathematical functions and operations.

It is clear that no finite number of experiments will suffice for the cryptanalyst to separate the
given functions from each other, as the arbitrary number may have been set to any value.
That’s why we must invoke the following theorem:

Theorem: There can not exist an algorithm that can identify a general computational
process based upon the input/output relation.

We conclude that if we use a cipher that includes a general computational process, and
keep all construction parameters of that process secret, the cryptanalyst will face a problem
which he will be unable to solve. We must however carefully get familiar with the
inconvenience which should occur if the system falls into the hands of the enemy. We see
specifically that simply using an optimal encryption algorithm, that is kept secret, will not be a
solution.

For that reasons, we invoke the concept of the specific universal machine. The specific
universal machine that we will make use of here must have a few specific properties. It
should be designed into accepting any binary string as valid input, i.e. no input string shall be
rejected as having wrong syntax. This requirement is equivalent to that the set of operations,
of the universal machine, is devised such that an operation will be selected in response to
any possible input information stream. This modification is of no difficulty, and can be
implemented without restricting the set of possible computations.

The input stream must further be kept secret, as knowledge of this would essentially be
equivalent to knowing the key of the system. This choice will not pose any difficulties, as the
universal machine may use any binary string as input. We see that the secret input stream
and the internal memory of the universal machine, may easily be protected during encryption
or decryption, and can be erased afterwards.

Why Dynamic Encryption Matters?

The modern world is a very dynamic place. We are going very digital and the information are
getting a normal part of our lives. Everything changes very fast and sometimes it’s quite
challenging to follow all those changes. If our world is going dynamic, the logical question
should be as follows: “Do we need the protection that will go dynamic as well?” The answer
is simple – yes.

What do we use to secure our so valuable information? An encryption, indeed. So, what we
need at this stage is a dynamic form of an encryption. It is well-known that modern
encryption systems are based on very strong mathematics and can appear in a form of both
– hardware and software. What is typical for many dynamic cryptographic systems is that
they are based not only on logical circuits, but include memory elements as well. Digital
science classifies these systems as sequential. The characteristic of sequential circuits is
that they go from one logical state to another. Basically, they make a cycle. Imagine how
these could be useful in terms of binary information permutation. For example, a logical part

35 Cyber Warnings E-Magazine – July 2014 Edition
Copyright © Cyber Defense Magazine, All rights reserved worldwide