傅立成臺灣大學:資訊工程學研究所鍾亞文Jong, Ya-WenYa-WenJong2010-06-092018-07-052010-06-092018-07-052009U0001-2207200917253000http://ntur.lib.ntu.edu.tw//handle/246246/185415近年來，市場上的智慧型裝置以及各式各樣的感測器越來越多，我們逐漸可以感受到新一世代的計算環境（computing environment）的到來。我們相信在不久的將來，這些智慧型裝置以及感測器將會逐一進入我們的家居生活中，為居住者帶來便利與舒適。許多研究設計出許多服務管理系統使得這些智慧型裝置以及感測器能夠順利相互溝通並合作，但其中卻很少系統關注於系統的可用性（availability）及穩定性（robustness）。這兩個特點其實對於服務管理系統而言扮演著極重要的角色，特別是在智慧家庭中，因為家中通常會有較多與居住者的健康具有密切關係的服務。然而家中通常沒有一個專業的系統管理者來監控以及修復系統意外的錯誤。因此，本研究的主要目的在於設計一個適用於智慧家庭的系統服務管理架構。這個系統服務管理架構具有自我修復與自我組織的能力，以增加系統的可用性及穩定性。研究的實驗結果證實我們所提出的方法能夠有效率的偵測出軟體錯誤並加以回復，藉此增加系統的可用性以及穩定性。More and more intelligent devices are becoming part of our life. In the near future, these intelligent devices will also be brought to our homes to provide comfort and convenience to habitants. A considerable number of studies have been made on service management systems for pervasive environments in order to coordinate these intelligent devices; however only few or none of them have addressed the system availability and robustness issue. These two aspects are fundamental for service management systems in smart homes, because there are usually pervasive services that are critical to the habitants’ health and life. In this paper, we propose a Pervasive Service Management System (PSMS) with self-organizing ability. Additionally, we have designed a failure detection and recovery protocol to enhance the availability and robustness of the PSMS. Experiment results show that the proposed solution can detect software failure efficiently without causing network instability, hence preserving the availability and robustness of the PSMS.口試委員會審定書 #謝 i文摘要 iiBSTRACT iiiONTENTS ivIST OF FIGURES viIST OF TABLES viiihapter 1 Introduction 1.1 Motivation 2.2 Related Works 3.3 Objective 7.4 Organization 7hapter 2 Preliminaries 9.1 Message-Oriented Middleware 9.2 Pervasive Node Application Model 11.3 Pervasive Communities 14.4 UPnP and SSDP 17hapter 3 Pervasive Service Management System 21.1 Message-Oriented Service Composition Protocol 22.2 Service Nodes Heartbeat Protocol 25.3 Rotating Roll-Call-based Protocol 27.3.1 RRCP Overview 28.3.2 Description of RRCP 31.3.3 Performance and Correctness Analysis 36.4 Message-Oriented Service Recovery Protocol 40hapter 4 Evaluation 42.1 Experiment Setup 42.2 Metrics 44.3 Experiment Description 46.4 Results and Discussions 49.4.1 Service Composition and Activation Latency 49.4.2 Service Nodes Failure Detection and Recovery Latency 50.4.3 RRCP and SSDP Failure Detection with varying failure rate 51.4.4 RRCP and SSDP Failure Detection with varying number of nodes 53.4.5 RRCP Failure Recovery with varying failure rate 55.4.6 RRCP Failure Recovery with varying number of nodes 56hapter 5 Conclusion 59EFERENCES 611105637 bytesapplication/pdfen-US自我組織錯誤偵測自我修復服務管理智慧家庭self-organizingfailure detectionself-healingservice managementsmart homes.thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/185415/1/ntu-98-R96922022-1.pdf