深圳大学学报理工版

为优化云服务系统的资源分配,提高不同资源类型的服务质量,提出基于简谐振动的云资源分配模型,设计一种求解模型的迭代算法.根据谐振子运动特性进行能级划分,加强对邻域内最优解的精细搜索,降低云资源被局部分配的概率,依据能级差构造解空间,使用简谐系统能量转换规律自适应调整解向量的搜索步长.通过实验验证分配模型的求解算法以及解的质量,相比分支定界法和遗传算法相比,该算法在较大规模问题上执行效率高且资源分配成本低.

In order to optimize the resource allocation of cloud service system and improve the quality of different resource types of services, we propose a cloud resource allocation model based on harmonic vibration and design an iterative algorithm to solve the developed model. Based on the harmonic oscillator movement characteristics, our model carries out the division of energy level, strengthens the fine search of the optimal solution in neighborhood, and reduces the possibility of the local distribution of cloud resources. Then, the solution space is reorganized based on the energy level difference. Meanwhile, the search step length of solution vector is adaptively adjusted by considering the energy conversion rule of harmonic system. Finally, the experiment validates the solution quality of proposed allocation model and solving algorithm by comparison with the branch/bound method and genetic algorithm. Especially, our method performs more efficiently and needs lower cost of resource allocation when dealing with the large-scale problems.

引言
1 模拟谐振子简述
2 云资源分配模型
3 最优解的计算
4 算法设计
5 实验仿真
6 结语

图1 简谐振动模型<br/>Fig.1 Harmonic vibration model

图1 简谐振动模型
Fig.1 Harmonic vibration model

图2 能级划分<br/>Fig.2 Energy level classification

图2 能级划分
Fig.2 Energy level classification

图3 能级差变化<br/>Fig.3 Change of energy level difference

图3 能级差变化
Fig.3 Change of energy level difference

图4 可用资源状态
Fig.4 Available resources

图5 资源划分示例<br/>Fig.5 An example of resource partition

图5 资源划分示例
Fig.5 An example of resource partition

图6 利用率和负载情况比较<br/>Fig.6 (Color online)Comparision of VM utilization and CPU load

图6 利用率和负载情况比较
Fig.6 (Color online)Comparision of VM utilization and CPU load

图7 最小资源成本比较<br/>Fig.7 Minimal resource cost comparison

图7 最小资源成本比较
Fig.7 Minimal resource cost comparison

[1] Durao F, Carvalho J F S, Fonseka A, et al. A systematic review on cloud computing[J]. The Journal of Supercomputing, 2014, 68(3): 1321-1346.
[2] 陈康,郑维民.云计算:系统实例与研究现状[J].软件学报,2009,20(5):1337-1348.
[3] Lin Weiyu, Lin Guanyu, Wei Hungyu. Dynamic auction mechanism for cloud resource allocation[C]// Proceeding of the 10th IEEE/ACM Conference on Cluster, Cloud, and Grid Computing. Piscataway, USA: IEEE Computer Society, 2010: 591-592.
[4] Shang Shifeng, Jiang Jinlei, Wu Yongwei, et al. DABGPM: a double auction Bayesian game-based pricing model in cloud market[C]// Proceeding of the IFIP International Conference on Network and Parallel Computing. Heidelberg, Germany: Springer-Verlag, 2010: 155-164.
[5] Sun Dawei, Chang Guiran, Wang Chuan, et al. Efficient Nash equilibrium based cloud resource allocation by using a continuous double auction[C]// Proceeding of the International Conference on Computer Design and Applications. Piscataway, USA: IEEE Computer Society, 2010: 94-99.
[6] Zaman S, Grosu D. Combinatorial auction-based allocation of virtual machine instances in clouds[J]. Journal of Parallel and Distributed Computing, 2013, 73(4): 495-508.
[7] 胡志刚,刘艳.云环境下基于组合双向拍卖的动态资源定价[J].计算机工程,2012,38(8):19-21.
[8] 丁丁,罗四维,艾丽华.基于双向拍卖的适应性云计算资源分配机制[J].通信学报,2012,33(Z1):132-140.
[9] Zhu Q, Agrawal G. Resource provisioning with budget constraints for adaptive applications in cloud environments[J]. IEEE Transactions on Services Computing, 2012, 5(4): 497-511.
[10] Harman M, Mansouri S A, Zhang Y. Search-based software engineering: trends, techniques and applications[J]. ACM Computing Surveys, 2012, 45(1): 11-61.
[11] 王鹏.云计算的关键技术与应用实例[M].北京:人民邮电出版社,2010:168-174.
[12] Shi Xuelin, Xu Ke. Utility maximization model of virtual machine scheduling in cloud environment[J]. Chinese Journal of Computers, 2013, 36(2): 252-262.
[13] 赵秀涛,张斌,张长胜.一种基于服务选取的SBS云资源优化分配方法[J].软件学报,2015,26(4):867-885.
[14] Shi Y, Eberhart R C. A modified particle swarm optimizer[C]// IEEE International Conference on Evolutionary Computation Proceedings. Piscataway, USA: IEEE, 1998: 69-73.
[15] 王赞,樊向宇,邹雨果,等.一种基于遗传算法的多缺陷定位方法[J].软件学报,2016,27(4):879-900.
[16] 刘全,王晓燕,傅启明,等.双精英协调进化遗传算法[J].软件学报,2012,23(4):765-775.

备注

引言

1 模拟谐振子简述

2 云资源分配模型

3 最优解的计算

4 算法设计

5 实验仿真

6 结语

期刊信息

备注

引言