The Dynamics involved in the 3N+1-problem

Abstract: The 3N+1-problem, also known as The Collatz Conjecture, concerns the behaviour of natural numbers N when the following rule is used repeatedly: If N is odd then multiply N by three and add one (a type Odd operation/ iteration), if N is even divide N by two (a type Even operation/ iteration). The Collatz Conjecture states that for any N as start-value the iteration-series must eventually reach the known Loop -4-2-1-4-... In the present work is shown that all iteration-series reach a value lower than the start-value, N > 1, at some point in the iteration-series, which is equivalent to showing that all iteration-series also reaches the known Loop. This is done by analysing the behaviour of subsets of the natural numbers. The subsets are called classes: (AX-B), B < A, where the independent variable X is a positive integer, the modulo A is a positive integer and the constant B is a non-negative integer. It is shown that there exist a countable infinity of "reducing" classes e.g. all elements in the class (2X) = {2,4,6,8,10,...} are reduced in one type Even operation and all elements in the class (4X-3) = {1,5,9,13,17,...} are reduced in one type Odd operation followed by two type Even operations. A formula is deduced that for any number of type Odd operations, #O = s, delivers the exact number of type Even operations, #E = r, required to obtain a "reducing combination" of operations: r = ceiling(sZ), s > 0, Z = log3/log2 = 1,5849625007... Another formula is deduced that for any "reducing combination" (s,ceiling(sZ)), s > 1, delivers the exact number of "reducing" classes having modulo A = 2^ceiling(sZ). By the help of a graph in form of an infinite complete binary tree it is then shown that all natural numbers N must belong to exactly one "reducing" class, thus proving that The Collatz Conjecture is true.