Deterministic Computing Mechanism for Perfect Density Classification

TL;DR

This paper investigates deterministic cellular automata rules capable of perfectly solving the density classification task, analyzing their state transition diagrams and properties to identify exact solutions that outperform previous soft computing methods.

Contribution

It introduces a mathematical framework for identifying deterministic CA rules that achieve perfect density classification, advancing beyond approximate solutions.

Findings

Number conserving CA rules can generate perfect DCT solutions.

State transition diagrams reveal classes of binary strings with equal weight.

Deterministic methods outperform soft computing techniques for DCT.

Abstract

The purpose of the present study is to search one-dimensional Cellular Automata (CA) rules which will solve the density classification task (DCT) perfectly. The mathematical analysis of number conserving functions over binary strings of length n gives an indication of its corresponding number conserving cellular automata rules (either uniform or non-uniform). The state transition diagrams (STDs) of number conserving CA rules have been analyzed where it has been found that these STDs can generate different DCT solutions. While studying the properties of STDs, an interesting classification of binary strings could be made where equal weight strings form a class and the cardinality of each class is same as the binomial coefficient nCk; n is the length and k is the weight of the binary string. Apart from STDs, other deterministic methods have been proposed to obtain the exact solution of DCT. All these exact solutions of DCT using different deterministic methods can be viewed as an improvement over the soft computing techniques used earlier to obtain approximate solutions.