A Homogeneous Distributed Computing Framework for Multi-objective Evolutionary Algorithm

This paper proposes a homogeneous distributed computing (HDC) framework for multi-objective evolutionary algorithm (MOEA). In this framework, multiple processors divide a work into several pieces and carry them out in parallel. Every processor does its task in a homogeneous way so that the overall procedure becomes not only faster but also fault-tolerant and independent to the number of processors. To implement this framework into an evolutionary algorithm, the evolutionary process of multi-objective particle swarm optimization (MOPSO) is employed. The effectiveness of the proposed framework is demonstrated by empirical comparisons between the results with the different numbers of processors, one and four. Seven DTLZ functions are used as benchmark functions and hypervolume, diversity, and evaluation time are used as comparison metrics. The results indicate that the evaluation time is significantly reduced by the proposed framework without any loss of overall solution quality and diversity.