基于云计算的流媒体任务调度算法  

Task scheduling algorithm based on cloud computing for streaming media

作　　者：杨戈[1,2] 吴俊言 Yang Ge;Wu Junyan(Key Laboratory of Intelligent Multimedia Technology,Beijing Normal University(Zhuhai Campus),Zhuhai 519087,China;Engineering Lab on Intelligent Perception for Internet of Things(ELIP),Shenzhen Graduate School,Peking University,Shenzhen 518055,China)

机构地区：[1]北京师范大学珠海分校智能多媒体技术重点实验室,广东珠海519087 [2]北京大学深圳研究生院深圳物联网智能感知技术工程实验室,广东深圳518055

出　　处：《电子技术应用》2021年第8期97-100,105,共5页Application of Electronic Technique

基　　金：广东高校省级重大科研项目(2018KTSCX288,2019KZDXM015,2020ZDZX3058);广东省学科建设专项资金(2013WYXM0122);北京师范大学珠海分校科研项目(201762005);北京师范大学珠海分校2019校级“质量工程”项目(201932)。

摘　　要：针对当前流媒体的大量视频资源从而带来的云计算的负载均衡和任务分配问题,在Cloudsim云环境下实现了任务调度的GAAC算法(Greedy And Ant Colony Algorithm,GAAC)。GAAC算法具有迭代学习机制、局部最优和负载均衡的特点。并在Cloudsim的环境下,完成了对GAAC算法、轮转算法(Round Roll Algorithm,RR)、贪心算法和蚁群算法的仿真比较。实验验证,GAAC算法从总体上而言,任务调度所用的时间明显较低于贪心算法和传统的轮转算法和蚁群算法,即其任务执行的时间更短,效率更高。Aiming at the problem of cloud computing load balancing and task allocation brought about by a large number of video resources in the current streaming media,the task scheduling GAAC algorithm(Greedy And Ant Colony Algorithm,GAAC)is implemented in the Cloudsim cloud environment.GAAC algorithm has the characteristics of iterative learning mechanism,local optimization and load balancing.In the context of cloudsim,simulations of GAAC algorithm,Round Roll Algorithm(RR),greedy algorithm and ant colony algorithm were completed.The experimental verification shows that GAAC algorithm is generally lower in the time spent on task scheduling than greedy algorithm,traditional rotation algorithm and ant colony algorithm.

关 键 词：云计算 任务调度 贪心算法 

分 类 号：TN949.2[电子电信—信号与信息处理]