Parallelization of ant colony optimization algorithm using Kaira

Abstract

Parallel computing can be very useful for meta-heuristic algorithms, because it can reduce time to find a solution with given properties. On the other hand parallel programming is generally more difficult than the sequential one and we need more development time to get a working parallel solution. We are developing a tool named Kaira which is intended for modelling, debugging, and automated generation of parallel applications. The modelling is based on our own variant of Coloured Petri Nets. In this paper we present how to create parallel programs from an existing sequential implementation of Ant Colony Optimization Algorithm using Kaira. This approach simplifies the development process and thus saves the development time and costs.