結合不同種類的物聯網裝置，進而構成智慧型物件之應用程式，最近已被應用在各種領域之中，像是智慧型玩具、智慧型博物館等。許多的研究講述如何架構一套系統去整合網路中的物聯網裝置，並討論如何改進服務質量與延遲時間等特性。然而，對於不同的實際應用，由於異構設備數量的不斷增加，如何減少嚴重的物聯網網路的擁塞以及雲端伺服器的負載是一個重要議題。在本文中，我們提出了一個通用的機器對機器架構，可以動態地將程式推送到機器對機器架構中的任意一個節點。這些程式可以是預處理程式，它們可以在發送到以太網前對數據進行預處理來提取一些更高級的特征，或者是整個物聯網程式，推送到離被服務的節點最近的地方，處理所有來自被服務節點的物聯網數據。此外，我們還提出使用開放服務網關協議規範的資源重用方法，來減少部署時間，從而盡快提供高質量的服務。此外，我們還制定了程式放置問題，并提出了一個有效的啟發式放置演算法來解決該問題。最後，我們對我們所提出的機器對機器架構以及資源重用方法進行效能測試。結果顯示使用資源重用方法在我們所提出的機器對機器架構中可以有效地減少部署時間，此外，我們所提出的演算法也可以實時地顯示近似最優解的性能。

關鍵詞—物聯網, 機器對機器架構, 模塊部署, 放置優化, 雲端計算

Consolidating a variety of IoT devices to compose the application for smart things has been applied in many fields recently, e.g., smart toy, smart museum, etc. Many works discussed about how to build a system to integrate things on network and how to improve some properties, e.g., service quality and latency. However, for real case IoT application, because of the ever-increasing number of heterogeneous devices, how to reduce severe IoT network congestion and surging cloud servers loads are an important issue. In this thesis, we propose a common M2M framework which can dynamically push the program to the nodes which are in the M2M framework. The programs can be pre-process programs which can pre-process data to extract some higher-level features before transmitting them over the Internet, or be whole IoT applications located at the closest position to the served node, which can process all IoT data from the served node. Furthermore, we propose a resource reused method using the OSGi specification to reduce the deployment time in order to provide the high-quality service as soon as possible. Moreover, we formulate a placement problem of program. We propose an efficient heuristic placement algorithm to solve the problem. Finally ,we evaluate some performance of our proposed M2M framework and resource reused method. The results show that using resource reused method can improve the deployment time effectively in our proposed M2M framework, and the proposed algorithm shows near-optimal performance in real-time.

Index Terms—Internet of Things, Machine-to-Machine (M2M) Framework, Module Deployment, Placement Optimization, Cloud Computing

中文摘要 I
ABSTRACT II
Acknowledgments III
TABLE OF CONTENTS IV
TABLE OF FIGURES V
TABLE OF TABLES VI
Chapter 1 Introduction 1
Chapter 2 Related Works 6
Chapter 3 Design and Implementations 9
3.1 Coordination Framework 9
3.1.1 Deployment Component 11
3.1.2 Deployment Controller 12
3.2 Module Deployment Flow in M2M Framework 13
3.3 Resource Reuse 15
Chapter 4 Module Placement Problem 17
4.1 Problem Formulation 17
4.2 Module Placement Algorithm 18
Chapter 5 Experiments Results 21
5.1 Experimental Environment 21
5.2 Evaluation of Resource Reused Method 23
5.3 Evaluation of Basic Procedures 24
5.4 Evaluation of Overhead of Module Architecture 26
5.5 Evaluation of The Module Placement Algorithm 28
Chapter 6 Conclusion and Future Work 31
Reference 33

[1] “Gartner says 6.4 billion connected” things” will be in use in 2016, up 30 percent from 2015,” Available: http://www.gartner.com/newsroom/id/3165317.
[2] Hong H J, Tsai P H, Hsu C H. Dynamic module deployment in a fog computing platform[C]//Network Operations and Management Symposium (APNOMS), 2016 18th Asia-Pacific. IEEE, 2016: 1-6.
[3] V. Stirbu, "Towards a RESTful Plug and Play Experience in the Web of Things," Semantic Computing, 2008 IEEE International Conference.
[4] T. Taleb, and A. Kunz, "Machine type communications in 3GPP networks: potential, challenges, and solutions," Communications Magazine, IEEE (Volume: 50, Issue: 3), 2012.
[5] oneM2M-TS-0001, “Functional Architecture”, v2.10.0, August, 2016
[6] Yoo-chul, Kim (15 July 2016). "LG Uplus expands international business". The Korea Times. Retrieved 25 August 2016.
[7] "oneM2M welcomes GlobalPlatform, TSDSI as partners". Telecompaper. Retrieved 25 August 2016.
[8] OM2M-Open Source platform for M2M communication. Available: http://www.eclipse.org/om2m/
[9] Kura - Open Source Java/OSGi-based framework for IoT gateways. Available:
https://www.eclipse.org/kura/
[10] D. Guinard, V. Trifa, and E. Wilde, "A Resource Oriented Architecture for the Web of Things," Internet of Things (IOT), 2010.
[11] Vogler M, Schleicher J M, Inzinger C, et al. DIANE-dynamic IoT application deployment[C]//Mobile Services (MS), 2015 IEEE International Conference on. IEEE, 2015: 298-305.
[12] “OSGi Specifications” Available: https://www.osgi.org/developer/specifications/
[13] Chen M, Wan J, González-Valenzuela S, et al. A Survey of Recent Developments in Home M2M Networks[J]. IEEE Communications Surveys and Tutorials, 2014, 16(1): 98-114.
[14] G. Wu, S. Talwar, K. Johnsson, N. Himayat, and K. D. Johnson, "M2M: From Mobile to Embedded Internet," Communications Magazine, IEEE (Volume: 49, Issue: 4), 2011.
[15] Alaya M B, Banouar Y, Monteil T, et al. OM2M: Extensible ETSI-compliant M2M service platform with self-configuration capability[J]. Procedia Computer Science, 2014, 32: 1079-1086.
[16] Wu C L, Liao C F, Fu L C. Service-oriented smart-home architecture based on OSGi and mobile-agent technology[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 2007, 37(2): 193-205.
[17] Satyanarayanan M, Bahl P, Caceres R, et al. The case for vm-based cloudlets in mobile computing[J]. IEEE pervasive Computing, 2009, 8(4).
[18] Willis D, Dasgupta A, Banerjee S. Paradrop: a multi-tenant platform to dynamically install third party services on wireless gateways[C]//Proceedings of the 9th ACM workshop on Mobility in the evolving internet architecture. ACM, 2014: 43-48.
[19] “IBM CPLEX optimizer,” Available: http://www-01.ibm.com/software/commerce/optimization/cplex- optimizer/.
[20] “GLPK,” Available: https://www.gnu.org/software/glpk/.
[21] “CBC,” Available: http://www.coin-or.org/projects/Cbc.xml.