A full process algebraic representation of Ant Colony Optimization

TL;DR

This paper introduces PA^2CO, a process algebra framework that formally models various parallel implementations of Ant Colony Optimization algorithms, enabling detailed specification and analysis of their parallel schemes.

Contribution

It presents a novel process algebra, PA^2CO, for formalizing and representing diverse parallelized ACO algorithms and their implementations.

Findings

Formal definitions of three ACO algorithms

Representation of fine-grained and coarse-grained parallel schemes

Enabling detailed analysis of parallelization strategies

Abstract

We present a process algebra capable of specifying parallelized Ant Colony Optimization algorithms in full detail: PA$^2$CO. After explaining the basis of three different ACO algorithms (Ant System, MAX-MIN Ant System, and Ant Colony System), we formally define PA$^2$CO and use it for representing several types of implementations with different parallel schemes. In particular fine-grained and coarse-grained specifications, each one taking advantage of parallel executions at different levels of system granularity, are formalized.