傅立成Fu, Li-Chen臺灣大學:資訊工程學研究所吳兆麟Wu, Chao-LinChao-LinWu2010-06-092018-07-052010-06-092018-07-052009U0001-2706200917173100http://ntur.lib.ntu.edu.tw//handle/246246/185365科技的進步讓生活環境進入了普及運算的時代，異質行動裝置及動態服務是普及運算環境的兩大特色，並會造成環境中的動態變化。智慧家庭系統的傳統架構通常是集中式的，因此會在應用於普及運算環境時造成許多問題，環境中的動態變化增加了系統元件互動的困難性，也導致系統資源無法充分運用。為了解決上述問題，這篇論文提出了一個以服務導向架構為基礎的智慧家庭系統，並應用了OSGi及行動代理人技術以實現此架構。此架構是一個包含數個OSGi平台的點對點模型，服務導向的機制在其中扮演著讓各平台元件互動的角色，行動代理人技術則用來加強互動機制並充分運用系統資源。在設計此架構的系統元件時，這篇論文應用了情境感知、混合主導、智慧型機器等設計概念，讓智慧家庭系統可以提供適當的服務、具備適切的人機互動機制、並達到彈性且有效的智慧型控制。基於此架構，智慧家庭是一個包含數個以代理人為基礎的自主互動空間的普及運算環境，在提供服務給使用者時可以在空間、軟硬體、服務品質上有多種選擇，也帶來了如何從中尋找最好選擇的挑戰，以能夠在科技方面展現智慧之外，也能夠在與使用者互動時滿足家的感覺。因此，這篇論文也同時從人機互動的觀點出發，分析智慧家庭系統中服務、空間、使用者之間的互動關係，提出一個人機互動框架及其相對應的演算法，以讓智慧家庭可以在與使用者互動時同時展現出智慧與家的感覺。The architecture of a conventional smart home system is usually server-centric and thus causes many problems, and it is also not suitable for ubiquitous computing environment. Applying a traditional server-centric architecture in a ubiquitous computing environment, the dynamically changing environment, which is caused by mobile devices and dynamic services, can result in very difficult interaction, and also makes system resources not fully utilized. In addition, how to provide services efficiently and appropriately is always an important issue for a smart home. To solve the problems caused by traditional architectures, to deal with the dynamic environment, and to provide appropriate service, in this thesis, a service-oriented architecture for smart home systems is proposed based on OSGi and mobile agent technology. This architecture is a Peer-to-Peer model based on multiple OSGi platforms, in which service-oriented mechanisms are used for system components to interact with one another, and mobile agent technology is applied to augment the interaction mechanisms and to fully utilize system resources. And when designing system components in this proposed architecture, the concepts of context-aware, mixed-initiative, and intelligent machine are applied, for smart home systems to provide appropriate services, to possess suitable human-computer interaction mechanism, and to achieve dynamic and efficient intelligent control. Based on this proposed architecture, current smart home is a ubiquitous computing environment consisting of multiple agent-based autonomous spaces, and its advantage is that a service interacting with home users can be set with different configurations in space, hardware, software, and quality. Not only being smart technologically speaking, a smart home should also never forget to retain the “home nature,” including warmth and/or relaxation, when it is serving its users. Therefore, from the viewpoint of human-computer interaction, the relationship among services, spaces, and users, is analyzed. And based on this analysis, a framework and its corresponding algorithm to model their interaction are proposed, so that a smart home can behave like a smart system while maintaining all kinds of “home” functions.誌謝 i文摘要 iiBSTRACT iiiONTENTS vIST OF FIGURES ixIST OF TABLES xiiihapter 1 Introduction 1.1 Basic Components and Interaction Flowchart of a Smart Home System 3.2 Challenges 5.3 Contribution 7.4 Thesis Organization 8hapter 2 Preliminaries 9.1 Service-Oriented Architecture 9.2 Context-Aware 13.3 OSGi 14.4 Mobile Agent Technology 16.5 Intelligent Machine and Agent-Based Control System 18.6 Mixed Initiative System 20.7 Human-Computer Interaction 21hapter 3 System Architecture 23.1 Context-Aware Smart Home 24.1.1 User Interface Agent (UI Agent) 25.1.2 Device Agent 26.1.3 Application Agent 27.1.4 Agent Directory 28.1.5 Mobile Agent Generator (MAG) 28.1.6 Context Agent 29.1.7 Database Agent 30.1.8 Inference Agent 31.1.9 Agent Coordinator 32.1.10 Design Issues 33.2 Interoperable Mobile Agent Migration Mechanism 40.2.1 Inter-OSGi Interaction with Web Services 41.2.2 Mobile Agent Architecture for OSGi Platforms 42.2.3 MASML 45.2.4 ASIA (Agent Specification Interpretation Algorithm) 49.3 Analysis and Discussion 51.3.1 Performance 51.3.2 Fault Tolerance 55.4 Prototype Implementation 58.5 Extended System Architecture for Multiple Spaces 59hapter 4 Human-System Interaction 65.1 Framework for Human-System Interaction 68.1.1 The properties of users 68.1.2 The requirements of services 69.1.3 The status of environment 70.2 Algorithm for Human-System Interaction 79.2.1 System Initiative Services 80.2.2 Find QS and CS from the Environment 80.2.3 List of CS 81.2.4 Configuration of Services 82.2.5 Notification 83.2.6 Manage Spaces 84.3 Analysis and Discussion 85.3.1 Comfort 85.3.2 Convenience 86.3.3 Security 87.4 System Design and Implementation 88.4.1 Locator Agent with Identity Recognition 89.4.2 Space Agent 89.4.3 Application Database (AppDB) 91.4.4 User Interface Agent (UI Agent) 92.4.5 Privacy Manager 93.4.6 Interaction Manager 94.4.7 Find_QS_and_CS Agent 95.4.8 Notification Agent 96.5 Application Examples and Scenarios 97.5.1 Media_Follow_Me Application 97.5.2 Ubiquitous Skype Application 99hapter 5 Experiments and Evaluations 103.1 System Performance 103.1.1 Analysis of HSI models 103.1.2 Experiments 107.2 Verifications and Evaluations for Framework 115.2.1 Experimental Verifications 115.2.2 Framework Evaluations 117hapter 6 Conclusion 130hapter 7 Future Works 133EFERENCE 136ublication List 1446696196 bytesapplication/pdfen-US智慧家庭服務導向架構智慧型機器多代理人系統人機互動互動框架Smart HomeSOAIntelligent MachineMulti-Agent SystemHuman-Computer InteractionHuman-System InteractionInteraction Frameworkthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/185365/1/ntu-98-F89922042-1.pdf