Weighted randomized algorithms for efficient load balancing in distributed computing environments
Maniza Hijab, Avula Damodaram · Materials Today Proceedings · 2020
Randomized algorithms for resource choice make use of information concerning the spread of their keyed-in data by using random samples. This can effectively improve resource utilization but can create a load imbalance naturally due to the randomness of its input space. For specific problems it is helpful to use a helper to direct the solution space in the right direction. In this paper, to enact effective resource utilization combined with optimized load balancing, a weighted randomized resource assignment algorithm is proposed. The simulation results using standard workload format datasets reveal that the proposed algorithm outperforms existing solutions in average resource utilization by 8% to 12% while improving on load balance by 5% to 11%.