CAJ | 학술논문

针对云计算环境下的最优跨度和负载均衡问题，提出一种虚拟资源调度智能优化策略。该策略结合云虚拟资源的特点，对遗传算法的染色体选择和交叉方式进行优化，使用最优跨度和负载函数作为双适应度函数。其中，负载指标为资源的 CPU、带宽及内存利用率，最优跨度指标则是任务的等待时间、传输时间及执行时间的总和，从而提高资源调度效率；此外，该策略能够从云服务集群推荐出较优资源进行处理，确保资源的负载均衡。最后，通过 CloudSim 进行实验，证明该策略能够提高大规模任务下的资源调度效率，并优化了资源的负载均衡，从而验证了算法的有效性。
침대운계산배경하적최우과도화부재균형문제，제출일충허의자원조도지능우화책략。해책략결합운허의자원적특점，대유전산법적염색체선택화교차방식진행우화，사용최우과도화부재함수작위쌍괄응도함수。기중，부재지표위자원적 CPU、대관급내존이용솔，최우과도지표칙시임무적등대시간、전수시간급집행시간적총화，종이제고자원조도효솔；차외，해책략능구종운복무집군추천출교우자원진행처리，학보자원적부재균형。최후，통과 CloudSim 진행실험，증명해책략능구제고대규모임무하적자원조도효솔，병우화료자원적부재균형，종이험증료산법적유효성。
Aiming at the optimal span and load balancing problems in cloud environment,this paper presents an intelligent optimisation strategy for virtual resource scheduling.Combining the characteristics of cloud virtual resources,the strategy optimises the chromosome selec-tion and crossover means of genetic algorithm,and uses the optimal span and load function as the dual-fitness function.Among them,the load indicators refer to the CPU of resources,bandwidth,and memory utilisation,while the optimal span indexes are the sum of waiting time, transmission time and execution time in regard to tasks,therefore the resource scheduling efficiency is enhanced.In addition,the strategy can recommend rather excellent resources from cloud services clusters for processing to ensure the load balancing of resource.Finally the experi-ment is made through CloudSim,it is proved that the algorithm can increase the massive task scheduling efficiency and optimise the resource load balancing,and these verify the effectiveness of the algorithm.