臺灣博碩士論文加值系統

現今嵌入式裝置性能越來越好，價格越來越低，基礎網路設施日趨完整，許多需要廣泛部屬的應用也隨之出現。以往必須以少量高效能的定點機器蒐集數據並推估其他位置的方式目前主要的佈建模式為大量定點小型的聯網嵌入式裝置蒐集數據，因此，是否能增加大範圍移動的聯網嵌入式裝置使得蒐集方式更具彈性？本篇論文以增加大範圍移動的聯網嵌入式裝置出發，使得嵌入裝置可以自由的移動的同時維持資料鍵結的完整。

在本篇論文中，引入了IoT中介層，悟空，讓使用相異網路協議的嵌入式裝置可以互相溝通，並且為了消去移動式嵌入裝置移動至另一個位置時所需的重新鍵結和減少的嵌入式裝置的網路使用。引入了新的網路架構，資料命名網路，於嵌入式裝置上，使得嵌入式裝置與其他的嵌入式裝置的鍵結為非直接的。

在實驗的部份，我們藉由模擬的方式，模擬一家賣場，當中的路由器為格子點分布，嵌入式裝置依照使用者的移動模型移動，預估當使用者切換到另一個路由器時，此區的定點嵌入式裝置需要多久時間可以得到資料。在移動式嵌入式裝置啟動時，依據亂數、最高機率的到達地點、路由器負荷三種演算法使用一固定的路由器，或依據位置切換路由器，並與使用IP網路比較。結果可以看出在導入資料命名網路後，主要的時間花費在更新用戶的所在位置，用戶資料可從附近的資料命名網路的路由器得到。相對於IP會受限於流量限制而限制服務人數，使用資料命名網路可以負擔更多的人數和更快的反應。

There are more and more applications are springing due to cheaper and more powerful devices. In the past, data collecting is that gather from a single powerful static device, and then using those data speculates the nearby situation. Now, we can deploy numerous IoT embedded devices, so the speculations are more accuracy. In the meanwhile, we consider using mobile IoT devices which are more flexible when deploying and adjusting. Therefore, this paper is aiming to add numerous wide range of mobile IoT devices and keep the links integrity at the same time.
In addition, we include IoT middleware, WuKong, to help us handle the numerous devices, and Named Data Networking to help us avoid re-config the routing table. NDN let links between devices is indirectly, and data are provided and requested by name. Besides, NDN routers cache data in content store, so the IoT devices can get data from NDN routers to reduced network traffic on mobile devices.

In the experiment, we consider the scenario of a shopping mall. Routers are installed as a grid-point distribution in the area. We compare three algorithms, random, router''s loading, and the highest probabilty, for assigning a home router and estimating the time interval between IoT device updating user''s location and the display receiving data by simulation. Also, we compare the results with changing home routers according user''s location and using IP network.

The result shows that updating location Interest is main delay because the data can be got from neighbor NDN routers.
Comparing to IP, when the numbers of customers are growing, data packets fill up bandwidth and cause high end-end delay.

1 Introduction 1
1.1 Motivation.................................. 1
1.2 Thesis Organization............................. 3
2 Background and Related Works 4
2.1 Background................................. 4
2.1.1 Wukong............................... 4
2.1.2 NDN ................................ 6
3 System Architecture and Problem Definition 8
3.1 System Architecture............................. 8
3.1.1 NDN client............................. 8
3.1.2 Virtual NDN Wuclass ....................... 9
3.2 Formal Model and Problem Definition................... 10
3.2.1 Formal Model............................ 10
3.2.2 Problem Definition......................... 10
4 Design and Implementation 13
4.1 Design.................................... 13
4.1.1 FBP in Wukong........................... 13
4.1.2 Name Policy ............................ 15
4.1.3 Lifetime of Interest packet and Data packet . . . . . . . . . . . . 16
4.2 Home Router ................................ 17
4.3 Implementation ............................... 18
5 Performance 19
5.1 Environment................................. 19
5.2 Experimental Settings............................ 19
5.3 Experiment Results............................. 20
6 Conclusion 29
Bibliography 30

[1] “Intel Edison Specfication,” http://www.intel.com/content/www/us/en/do-it- yourself/edison.html.
[2] “Intel Edison Price,” http://www.newegg.com/Product/Product.aspx?Item=13-121- 814.
[3] “Raspberry,” https://www.raspberrypi.org/.
[4] “WuKong Release 0.4,” https://www.gitbook.com/book/wukongsun/wukong- release-0-4/details.
[5] “WuKong Github,” https://github.com/wukong-m2m/wukong-darjeeling.
[6] L. Zhang, A. Afanasyev, J. Burke, V. Jacobson, k. claffy, P. Crowley, C. Papadopoulos, L. Wang, and B. Zhang, “Named data networking,” SIGCOMM Comput. Commun. Rev., vol. 44, no. 3, pp. 66–73, Jul. 2014. [Online]. Available: http://doi.acm.org/10.1145/2656877.2656887
[7] V. Jacobson, D. K. Smetters, J. D. Thornton, M. F. Plass, N. H. Briggs, and R. L. Braynard, “Networking named content,” in Proceedings of the 5th International Conference on Emerging Networking Experiments and Technologies, ser. CoNEXT ’09. New York, NY, USA: ACM, 2009, pp. 1–12. [Online]. Available: http://doi.acm.org/10.1145/1658939.1658941
[8] S. seok Seo, J.-M. Kang, Y. Han, and J. W.-K. Hong, “Analysis and performance evaluation of data transport methods in content-centric networking,” in Network Operations and Management Symposium (APNOMS), 2013 15th Asia-Pacific, Sept 2013, pp. 1–3.
[9] A. Afanasyev, C. Yi, L. Wang, B. Zhang, and L. Zhang, “Map-and-encap for scaling ndn routing,” in Technical Report NDN-0005, 2015.
[10] “OMNeT++,” https://omnetpp.org/.
[11] “CAT-5e,” http://www.tecdatawire.com/catspec.html.