Load balancing strategies to solve flowshop scheduling on parallel computing

This paper first presents a parallel solution for the Flowshop Scheduling Problem on parallel computing environment, and then proposes a load balancing strategy. The proposed Proportional Fairness Strategy takes computational performance of computing process sets into account and assigns additional load to computing nodes proportionally to their evaluated performance. In order to efficiently utilize the power of parallel resource, we also discuss the data structure used in communications among computational nodes and design an optimized data transferring strategy. This data transferring strategy as well as the proposed load balancing strategy have been implemented and tested on a super computer consisted of 86 CPUs using MPI [12] as the middleware. The results show that the proposed Proportional Fairness Strategy can achieve better performance in computing time than the existing Adaptive Contracting Within Neighborhood Strategy (ACWN). We also show that with the combination of both the Proportional Fairness Strategy and the proposed data transferring strategy the efficiency of parallelism can obtain a 13 similar to 15% improvement.