基于多层次用户属性的移动云调度研究
发布时间:2018-12-16 05:26
【摘要】:移动云计算服务是技术融合在当前环境下的新发展,它是在云计算以及移动互联网基础上发展而来的,它结合了两者的优点,但是也带来了新的难题。本论文研究课题主要是移动云计算服务中的移动云资源调度问题,主要研究解决当前移动云调度中的两个基本问题:1)如何在调度时减少数据检索规模,加快云调度速度;2)如何在移动云调度中结合用户需求以及多层次属性,为移动用户带来更加个性化的服务。为了解决上述问题,本文研究了移动云调度算法,提出了一个基于多层次用户属性的移动云调度系统。模糊聚类可以对事物进行分类,从而使问题在更精确的范围内解决,因此,本文在模糊聚类算法的基础上提出了相关改进算法(IGAFCM),以解决移动云调度中数据量过大的问题。同时,本文还利用了推荐系统中的协同过滤算法(FCMCF),对移动用户多层次属性进行系统抽象,提出一个基于多层次用户属性的评分模型,让移动云调度系统在充分考虑移动用户需求和相关属性的基础上进行系统调度,从而解决为用户提供更加个性化服务的问题。最后,本文还利用了 Matlab以及CloudSim等仿真工具,对研究的算法以及设计的系统进行实验验证。通过实验证明,本文提出的基于多层次用户属性的移动云调度算法是有效可行的。
[Abstract]:Mobile cloud computing service is a new development of technology amalgamation in the current environment. It is based on cloud computing and mobile Internet. It combines the advantages of both, but also brings new problems. This thesis focuses on mobile cloud resource scheduling in mobile cloud computing services. It mainly studies two basic problems in mobile cloud scheduling: 1) how to reduce the scale of data retrieval and speed up cloud scheduling; 2) how to combine user requirements and multi-level attributes in mobile cloud scheduling to bring more personalized services to mobile users. In order to solve the above problems, this paper studies the mobile cloud scheduling algorithm, and proposes a mobile cloud scheduling system based on multi-level user attributes. Fuzzy clustering can classify things so that the problem can be solved in a more accurate range. Therefore, based on the fuzzy clustering algorithm, this paper proposes an improved algorithm, (IGAFCM), to solve the problem of too large amount of data in mobile cloud scheduling. At the same time, the collaborative filtering algorithm (FCMCF), in recommendation system is used to abstract the multi-level attributes of mobile users, and a scoring model based on multi-level user attributes is proposed. In order to solve the problem of providing more personalized service to users, the mobile cloud scheduling system can take full account of the needs of mobile users and related attributes to carry out system scheduling. Finally, the simulation tools such as Matlab and CloudSim are used to verify the algorithm and the designed system. Experimental results show that the proposed mobile cloud scheduling algorithm based on multi-level user attributes is effective and feasible.
【学位授予单位】:北京邮电大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:TP393.09
本文编号:2381876
[Abstract]:Mobile cloud computing service is a new development of technology amalgamation in the current environment. It is based on cloud computing and mobile Internet. It combines the advantages of both, but also brings new problems. This thesis focuses on mobile cloud resource scheduling in mobile cloud computing services. It mainly studies two basic problems in mobile cloud scheduling: 1) how to reduce the scale of data retrieval and speed up cloud scheduling; 2) how to combine user requirements and multi-level attributes in mobile cloud scheduling to bring more personalized services to mobile users. In order to solve the above problems, this paper studies the mobile cloud scheduling algorithm, and proposes a mobile cloud scheduling system based on multi-level user attributes. Fuzzy clustering can classify things so that the problem can be solved in a more accurate range. Therefore, based on the fuzzy clustering algorithm, this paper proposes an improved algorithm, (IGAFCM), to solve the problem of too large amount of data in mobile cloud scheduling. At the same time, the collaborative filtering algorithm (FCMCF), in recommendation system is used to abstract the multi-level attributes of mobile users, and a scoring model based on multi-level user attributes is proposed. In order to solve the problem of providing more personalized service to users, the mobile cloud scheduling system can take full account of the needs of mobile users and related attributes to carry out system scheduling. Finally, the simulation tools such as Matlab and CloudSim are used to verify the algorithm and the designed system. Experimental results show that the proposed mobile cloud scheduling algorithm based on multi-level user attributes is effective and feasible.
【学位授予单位】:北京邮电大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:TP393.09
【参考文献】
相关期刊论文 前4条
1 汪星荷;刘绍华;俞俊生;;移动云计算中基于LBS的个性化服务推荐模型[J];数学的实践与认识;2013年02期
2 李文娟;张启飞;平玲娣;潘雪增;;基于模糊聚类的云任务调度算法[J];通信学报;2012年03期
3 梁茹冰;刘琼;;公路网移动终端的KNN查询技术[J];华南理工大学学报(自然科学版);2012年01期
4 姚婧;何聚厚;;基于模糊聚类分析的云计算负载平衡策略[J];计算机应用;2012年01期
相关硕士学位论文 前2条
1 郑思远;基于二分图的混合推荐系统的研究与实现[D];北京邮电大学;2015年
2 董世龙;基于模糊聚类的云任务调度优化策略研究[D];广西大学;2014年
,本文编号:2381876
本文链接:https://www.wllwen.com/guanlilunwen/ydhl/2381876.html