摘要
在云计算中,系统规模和虚拟机迁移数量都是十分庞大的,需要高效的调度策略对其进行优化。将云计算的任务分配抽象为背包求解问题,可通过遗传算法进行求解。传统的遗传算法具有局部搜索能力差以及早熟现象的缺点,采用遗传和贪婪相结合的混合遗传算法。针对混合遗传算法在资源利用率与能源消耗的收敛速度较慢问题,通过改进适应度函数,改变了适应度函数在不同染色体间的差异度,从而提高了染色体在选择算子中的择优性能。仿真结果表明,该方法能够有效提高混合遗传算法在云计算资源优化中的收敛速度。
The size of system and the number of virtual machine migration in cloud computing are very, large, for which the efficient scheduling strategy is essential. The task alloeation for cloud eomputing can be abstracted to knapsack problem, and then is solved by genetic algorithm. The traditional genetie algorithm has the shortcoming of poor local searching ahility and precocious phenomenon, which can adopt the combination of genetic and greed hybrid genetic algorithm to solve. For hybrid genetic algorithm convergence speed problem in resource utilization and energy consumption, in this paper, the fitness function is changed to increase the difference of chromosomes and improve the performance of chromosome preferred in sclection operator. The simulation results show that this method can effectively improve the hybrid genetic algorithm convergence speed in cloud computing resources optimization.
出处
《电视技术》
北大核心
2015年第18期36-41,共6页
Video Engineering
基金
国家自然科学基金项目(61172054
61362006)
广西自然科学基金项目(2014GXNSFAA118387
2013GXNSFAA019334)
桂林电子科技大学研究生创新项目(GDYCS201409)
关键词
云计算
资源调度
混合遗传算法
cloud computing
resource scheduling
hybrid genetic algorithm