CAJ | 학술논문

在云计算环境下的多分簇目标的任务调度是提高软件嵌入式系统稳定性的关键，由于云计算中能耗扩散，在多处理器集群多目标任务调度中出现任务调度优先级列表混乱，调度性能不好的问题。传统方法采用节点行为信息流特征分析方法进行任务调度，随着分簇能耗扩散，收敛性不好。提出一种基于分簇能耗聚集的云计算多目标任务调度算法。待分析任务调度信息流的分层能耗聚集轨迹峰值往往含有重要的信息，提取待分析任务调度信息的能耗特征，构建云计算多目标任务调度模型，实现对云计算多目标任务调度算法的改进。仿真结果表明：采用该算法能有效提高云计算多目标任务调度效率，任务执行时间较短，系统稳定性高。
재운계산배경하적다분족목표적임무조도시제고연건감입식계통은정성적관건，유우운계산중능모확산，재다처리기집군다목표임무조도중출현임무조도우선급렬표혼란，조도성능불호적문제。전통방법채용절점행위신식류특정분석방법진행임무조도，수착분족능모확산，수렴성불호。제출일충기우분족능모취집적운계산다목표임무조도산법。대분석임무조도신식류적분층능모취집궤적봉치왕왕함유중요적신식，제취대분석임무조도신식적능모특정，구건운계산다목표임무조도모형，실현대운계산다목표임무조도산법적개진。방진결과표명：채용해산법능유효제고운계산다목표임무조도효솔，임무집행시간교단，계통은정성고。
AbstractComputing task scheduling multicluster target environment in the cloud is the key to improve the stability of the software of the embedded system. Owing to the energy diffusion in cloud computing,task scheduling priority list appears chaos in the multi processor cluster multi-objective task scheduling,and scheduling performance is not good. The traditional method uses the node behavior information flow characteristic to analyze task scheduling,with the cluster energy diffusion,convergence is not good. This pa-per presents a method for calculating the multi-objective task scheduling algorithm based on energy con-sumption of the cloud cluster aggregation. The simulation results show that,the algorithm can effectively im-prove the efficiency of cloud computing multi-objective task scheduling,task execution time is short,and it has high stability of the system.