FULL PDF
Abstract
As Big Data applications continue to grow, energy efficiency and cost management have become critical concerns in heterogeneous computing environments. Traditional resource allocation and job scheduling algorithms often fail to balance trade-offs between performance, energy consumption, and operational cost. Multi-objective optimization (MOO) algorithms offer a solution by simultaneously optimizing multiple conflicting objectives, such as minimizing energy consumption while maximizing processing efficiency. This paper provides a comprehensive study of MOO algorithms for cost-efficient and energy-aware data processing in heterogeneous Big Data frameworks. We analyze traditional and emerging optimization approaches and propose a hybrid model integrating evolutionary algorithms and dynamic workload scheduling techniques. Experimental results demonstrate that our model significantly reduces energy consumption while maintaining high computational efficiency.
View more >>