Re: How long can we wait before we absolutely must take steps to

[Ian Parker]

On 20 Jan, 18:50, Chalky <chalkys...@bleachboys.co.uk> wrote:
> I have not examined your references, but there is a well documented
> method of providing an apparently higher level of protection than
> normal, known as the Vernam algorithm, which, iirc, dates from around
> the 1940s. Basically what you need to do is generate a completely
> different (and pseudo random) encryption algorithm for every
> communication (or every part of a communication). Provided the sending
> and receiving machines both know what that PRBSG algorithm is, =A0they
> can continue to communicate securely. There are many ways that this
> basic idea can potentially be developed for still higher security, but
> I leave that to your imagination.
>
This is basically what I was saying. If you have a one time pad you
are absolutely secure. Mind you have to deliver your one time pad and
this may well be a dificulty. It was a difficulty forthe Germans in
WW2. The wheels for Enigma had to be set up using plain German. This
was how the code was cracked, it would not have been cracked
otherwise.

You can get a pseudo one time pad quite easily. What you need is a
function that generates integers in a pseudorandom way. The normal
metof of pseudorandon generation is to find. A one time binary pad
works by taking your message and transmitting m^a (m^a)^a is m.

a =3D |a*q|p

where a is your sequence, and pand q are two fairly large numbers
which are relatively prime. The modulus to p of a*q.

We can use large integer numbers if we wish.

An alternative way is to just have 32 bit numbers and to pick q,p out
of an array in a pseudorandom way. This I have described as a
"Generating Function".

Although this is an extremely secure (and fast) method of transmitting
long messsages we cannot set up a protocol remotely.

I can do this with RSA. Computers A & B each work out a code. A & B
then send the public part to each other and then start communicating.
This cannot be done with XOR codes.

Another task of cryptography is authentication. I want to do something
on a remote computer. Before the process can start I have to establish
the authenticity of the code. To do this I have to establish secure
communication and ask the sending computer a question that only it
knows the answer to. Only RSA can do this.

- Ian Parker