|
个人信息Personal Information
教授
博士生导师
硕士生导师
主要任职：Director of Academic Committee at Kaifa District
其他任职：开发区校区学术分委员会主任（Director of Academic Committee at Kaifa Campus）
性别：男
毕业院校：多伦多大学
学位：博士
所在单位：软件学院、国际信息与软件学院
学科：软件工程. 运筹学与控制论
办公地点：开发区（Kaifa District Campus）
联系方式：mingchul@dlut.edu.cn
电子邮箱：mingchul@dlut.edu.cn
Effective hybrid load scheduling of online and offline clusters for e-health service
点击次数：
论文类型：期刊论文
发表时间：2017-01-12
发表刊物：NEUROCOMPUTING
收录刊物：SCIE、EI、Scopus
卷号：220
页面范围：60-66
ISSN号：0925-2312
关键字：Hybrid load; Online e-health service; Compression index; Storm platform
摘要：Hybrid load in e-health services is composed of online e-health service applications and offline jobs. Previous methods overlooked the impact of system performance for the fine-grained service components. In this paper, a hybrid load scheduling scheme is proposed in which scheduling is performed not only at the level of the component, but also within components. To improve both execution efficiency and searching accuracy, the proposed algorithm searches the compressing method of the Lucene index and then filters that index. Simulations are conducted on a Storm platform to evaluate the performance of the proposed scheme. Simulation results demonstrate that the proposed scheme can increase the response speed by 67.79% with an accuracy of 95.94%, and the response speed decreases by 11.6-53.2%. (C) 2016 Elsevier B.V. All rights reserved.