# The Elegance of the Humble Cross-Product in Vector Cryptography.

*From*: adacrypt <austin.obyrne@xxxxxxxxxxx>*Date*: Sat, 15 Oct 2011 02:13:08 -0700 (PDT)

Running through the algorithm quickly, Alice ‘guesses’ (virtually) a

normal vector that she uses to define a plane in space and also to

define a line in that plane that she then commissions as a directed

number line. Her eventual encryption transformation of a plaintext is

that it evolves as a number on this number line that is represented

now by a position vector called Pn i.e the position of ‘n’ on the

line. In collusion with Bob she claps a large change-of-origin onto

to this Pn and sends it as ciphertext to Bob who removes the

obfuscating change-of-origin to get the Pn in its true from and

decrypts it back to ‘n’ before then decoding n back to its plaintext

equivalent.

The Overview in detail.

The geometry of these lines is such that Alice’s encryption line is

inextricably related to the direction of the normal vector that she

guessed at the beginning. It is apart from that line however in the

sense that a) they don’t physically intersect and b) by the same token

is not parallel to it either – in fact it is at right angles to it.

So, the lines are skew by a) and secondly, they are orthogonal by

b).

That they are “orthogonal skew lines” is a valid description – this is

all incidental to what I want to say about the artistic beauty of the

algorithm.

The Elegance of Bob’s Decryption.

The model being used in this cryptography boils down to a pair of

orthogonal skew lines as explained. Let us extend the model now by

drawing a longer line along i.e. coincident with, the normal vector

(extending it by using the normal as the parallel direction of the

line in other words) and we will now call this Bob’s number line as a

compliment to Alice’s number line.

We can now talk unambiguously about Alice’s number line and Bob’s

number line as a pair of orthogonal skew lines. Alice encrypts on her

line and Bob decrypts on his line.

The way it works.

Saying it as quickly as possible, the vector cross product methodology

enables any number on Alice’s number line to be transformed by

rotation onto Bob’s number line where he is able to proceed with its

decryption back into plaintext. Bob performs this ‘transformation of

n by rotation’ using Pn x V0 on Alice’s line to define n’s position

exactly on his own line.

This means that Bob is able to transform a number from a line that he

doesn’t need to know anything about (although he probably does on most

occasions) onto to his own line that he knows everything about and

then decode it back to its plaintext value.

The salient property of vectors that makes this possible is the

perpendicularity of the *cross product methodology (i.e. the cross

product of any two vectors is another, third vector, that is mutually

at right angles to each of the pair being multiplied).

This perpendicularity of the cross product is used widely in mechanics

but its use here in cryptography to decrypt numbers is to my mind

quite beautiful.

As a famous American golfer once said “we should stop to smell the

flowers more often”.

I’m sure readers will agree that the saying above is figuratively

pertinent to cryptography also. It is appropriate to highlight this

elegant application of the cross-product in cryptography by that same

token I think.

Appendix.

– the algebraic corollary of Bob’s “decryption by rotation” is

Pn x V 0 / N = n (dividing corresponding coefficients on the left

enables this equation to be solved although division is not generally

defined in vector arithmetic).

Euphemism.

The model of Alice and Bob’s encryption lines being a pair of

orthogonal skew lines is a bit like a pair of chop-sticks being held

at right angles to each other - Chop stick Cryptography?

Cheers - adacrypt

.

**Follow-Ups**:

- Prev by Date:
**Re: Block Cipher Endedness** - Next by Date:
**Re: The Elegance of the Humble Cross-Product in Vector Cryptography.** - Previous by thread:
**Initializatin Vector schedule** - Next by thread:
**Re: The Elegance of the Humble Cross-Product in Vector Cryptography.** - Index(es):