Tao Yu的论文之三 "Service Selection Algorithms for Composing Complex Services with Multiple QoS Constraints(ICSOC2005)"
2009-03-19 20:12
459 查看
T. Yu and K.J. Lin, "Service Selection Algorithms for Composing Complex Services with Multiple QoS Constraints"(gs: 52), ICSOC2005
ISeBM05那篇讨论了单约束条件下的服务选择, 本文则进一步讨论多QoS约束下的服务选择.
combinatorial model
将问题建模为multi-dimension multi-choice 0-1 knapsack problem (MMKP)
基于Khan, S的heuristic algorithm(HEU), 开发了WE_HEU算法来求解.
graph model
将问题建模为multi-constraint optimal path problem (MCOP)
基于针对单限制最优路径的CSP算法, 开发了MCSP算法. 这个算法与CSP并没有本质的区别, 无非是原先只要比较一个QoS, 现在要比较多个而已.
MCSP算法可能会生成过多的备选路径, 因而作者由提出了改进算法MCSP-K, 在RELAX时只保留K个路径.
定义了nonlinear cost function(引用于Multi-constrained optimal path selection, INFOCOM 2001)作为筛选路径的标志. 对于这个nonlinear cost function函数, 有两点要提一下
(1) 作者特意讨论了λ->∞时的情况, 并在算法中有相应体现. 没看出强调这种情况的必要性.
(2) QoS参数分为成本型和效益型, 该函数应该是针对效益型参数而言, 而对于成本型的QoS参数应先进行转换.
使用了Inet3.0来生成模拟的实验环境.
实验参数
sservice candidate nodes: 25~2500
service candidate graph: random graph topology
仅考虑one process plan with the sequential composition model
number of service class: 5~50
candidates in each service class: 5~50
5个QoS参数, 采用[1, 100]间的uniform distribution
utility value of each link: [1, 200]间的uniform distribution
network impact factor: 按照大小分为两种情况. combinatorial approach中无法建模network impact factor.
与ISeBM05的比较
要解决的问题: 单属性 vs. 多属性
解决方法: combinatorial approach的算法没有细看; graph approach中的算法没有本质区别
本文对问题的简化
(1) 不关心服务之间接口的兼容性问题.
(2) 只考虑sequential composition model (S3)
ISeBM05那篇讨论了单约束条件下的服务选择, 本文则进一步讨论多QoS约束下的服务选择.
combinatorial model
将问题建模为multi-dimension multi-choice 0-1 knapsack problem (MMKP)
基于Khan, S的heuristic algorithm(HEU), 开发了WE_HEU算法来求解.
graph model
将问题建模为multi-constraint optimal path problem (MCOP)
基于针对单限制最优路径的CSP算法, 开发了MCSP算法. 这个算法与CSP并没有本质的区别, 无非是原先只要比较一个QoS, 现在要比较多个而已.
MCSP算法可能会生成过多的备选路径, 因而作者由提出了改进算法MCSP-K, 在RELAX时只保留K个路径.
定义了nonlinear cost function(引用于Multi-constrained optimal path selection, INFOCOM 2001)作为筛选路径的标志. 对于这个nonlinear cost function函数, 有两点要提一下
(1) 作者特意讨论了λ->∞时的情况, 并在算法中有相应体现. 没看出强调这种情况的必要性.
(2) QoS参数分为成本型和效益型, 该函数应该是针对效益型参数而言, 而对于成本型的QoS参数应先进行转换.
使用了Inet3.0来生成模拟的实验环境.
实验参数
sservice candidate nodes: 25~2500
service candidate graph: random graph topology
仅考虑one process plan with the sequential composition model
number of service class: 5~50
candidates in each service class: 5~50
5个QoS参数, 采用[1, 100]间的uniform distribution
utility value of each link: [1, 200]间的uniform distribution
network impact factor: 按照大小分为两种情况. combinatorial approach中无法建模network impact factor.
与ISeBM05的比较
要解决的问题: 单属性 vs. 多属性
解决方法: combinatorial approach的算法没有细看; graph approach中的算法没有本质区别
本文对问题的简化
(1) 不关心服务之间接口的兼容性问题.
(2) 只考虑sequential composition model (S3)
相关文章推荐
- Tao Yu的论文之一 "Service Selection Algorithms for Web Services with End-to-end QoS Constraints (ISeBM2005)"
- [论文笔记]Combining Global Optimization with Local Selection for Efficient QoS-aware Service Composition (WWW 2009)"
- Efficient algorithms for Web services selection with end-to-end QoS constraints
- [论文笔记] SOA Middleware Support for Service Process Reconfiguration with End-to-End QoS Constraints (ICWS 2009)
- Tao Yu的论文之二 "Adaptive algorithms for Finding Replacement Services in Autonomic Distributed Business Processes(ISADS2005)"
- Zeng Liangzhao的经典论文之二 "QoS-Aware Middleware for Web Services Composition" (TOSE04)
- MC Jaeger论文之一 "QoS aggregation for Web service composition using workflow patterns" (EDOC2004)
- [论文泛读] High performance approach for multi-QoS constrained Web services selection (ICSOC, 2007)
- 2013.4.20 Combining global optimization with local selection for efficient QoS-aware service composition
- [论文笔记] QoS Analysis for Web Service Composition (SCC, 2009)
- Job for nginx.service failed because the control process exited with error code. See "systemctl stat
- [论文笔记] Gradual Removal of QoS Constraint Violations by Employing Recursive Bargaining Strategy for Optimizing Service Composition Execution Path (ICWS, 2009)
- Jenkins 安装启动提示“iJob for jenkins.service failed because the control process exited with error code. See "systemctl status jenkins.service" and "journalctl -xe" for details.”
- Real-time tracking of multiple objects using adaptive correlation filters with complex constraints
- [论文笔记] A Backwards Composition Context Based Service Selection Approach for Service Composition (SCC, 2009)
- 论文读书笔记-Submodular meets Spectral: Greedy Algorithms for Subset Selection, Sparse Approximation and D
- 论文阅读笔记之ICML2012::The Landmark Selection Method for Multiple Output Prediction 基于代表性特征选取的多维度回归方法
- 2013.4.13 Domain-Specific Service Selection for Composite Services
- 论文阅读笔记之ICML2012::The Landmark Selection Method for Multiple Output Prediction 基于代表性特征选取的多维度回归方法
- [论文阅读] Dynamic Web service selection for reliable Web service composition (TSC, 2008)