An evolutionary approach to load balancing parallel computations

Abstract

We present a new approach to balancing the workload in a multicomputer. It is based on a genetic algorithm that combines a number of design choices in order to ameliorate the problem of premature convergence. The genetic algorithm is further hybridized by including a hill climbing procedure which significantly improves the efficiency of the evolution. Moreover, it makes use of problem specific information to evade computational costs and to reinforce favorable aspects of the genetic search. The experimental results show that the hybrid genetic algorithm can find solutions that are very close to the optimum.