目次

1              Introduction       1.1          Cloud Computing             1.2          Cloud Service Models    1.3          Provider Centered Cloud Service Computing       1.4          Use Cases: Cloud Service Composition 1.5          Key Research Challenges              1.4          Research Contributions 1.5          Organization 2              Cloud Service Composition: The State of the Art                2.1          Cloud Service Composition from an End User’s Perspective 2.2          Cloud Service Composition from a Provider’s Perspective             2.3          Economic Models            2.4          Prediction Modeling in Service Composition        2.5          Optimization Approaches in Service Composition              3              Long-term IaaS Composition for Deterministic Requests 3.1          Introduction 3.2          The Heuristics on Consumer Behavior 3.3          The Long-term Composition Framework for Deterministic Requests 3.4          Predicting the Dynamic Behavior of Consumer Requests               3.5          An ILP Modeling for Request Optimization 3.6          Experiments and Results 4              Long-term IaaS Composition for Stochastic Requests 4.1          Introduction       4.2          Long-term Dynamic IaaS Composition Framework 4.3          Long-term Economic Model of IaaS Provider 4.4          Genetic Optimization using IaaS Economic Model             4.5          Hybrid Adaptive Genetic Algorithm (HAGA) based Composition                 4.6          Experiments and Results                              5              Long-term Qualitative IaaS Composition                5.1          Introduction       5.2          Motivation:  A Qualitative IaaS Economic Model with Decision Variables 5.3          The Temporal CP-Net based Qualitative Economic Model             5.4          Optimization Algorithms for Qualitative IaaS Composition             5.5          Reinforcement Learning for Long-term IaaS   Requests Composition                        5.6          Experiments and Results                                              6              Service Providers’ Long-term QoS Prediction Model        6.1          Introduction       6.2          The Multivariate QoS Forecasting Framework     6.3          Multivariate QoS Prediction Model (MQPM)       6.4          Forecasting from the MQPM 6.5          Experiments and Results 7              Conclusion 7.1          Future Work