中文摘要：

资源调度是云计算的核心问题，传统遗传算法（GA）、Sufferage算法等都可以用于云计算环境中的资源调度，但传统遗传算法存在收敛慢、易早熟等缺点，Sufferage算法则不适用于多聚类环境的密集型任务调度．本文在充分考虑云计算环境的动态异构性和大规模任务处理特性的基础上，提出了一种基于染色体编码方式和适应度函数的改进遗传算法（IGA），并在云仿真器CloudSim上对3种算法进行了仿真．仿真结果表明，该算法在性能和服务质量QoS（Qualityof Service）方面都优于传统遗传算法和Sufferage，能更好地适用于大规模任务下的云计算环境资源调度．

英文摘要：

Cloud computing is an emerging distributed computing method which integrates heterogeneous, distributed resources on the internet into a supercomputer to provide services for users by virtualization technology. The basic scheme is that complex and large computing tasks are divided into smaller sub-tasks, which will be first executed by cloud resources and then the executed results will be send back to users, so resources scheduling is the core problem in cloud computing environment. Traditional generic algorithm （GA）, sufferage algorithm can both be used for resources scheduling in a cloud computing environment, traditional generic algorithm has the disadvantage of slow convergence and prematurity, sufferage performs worse in case of data-intensive applications in multiple cluster environments. Since characteristics of dynamic, heterogeneous and large-scale tasks need to be processed in cloud computing environment, we propose here an improved generic algorithm （IGA） based on chromosome encoded mode and fitness function, to emulate the three algorithms on CloudSim. Simulation data showed that the improved algorithm performed better than GA and sufferage method in regard to performance and QoS （Quality of Service）, which would be better applicable for resource scheduling in a cloud computing environment.