臺灣博碩士論文加值系統

隨著物聯網的興起，透過無線網路傳輸來串連現實世界和網路世界是一件容易的事，物聯網由許多感測器和無線網路組成且運用於非常多的應用，例如：環境感測、工廠配置、健康監測等等。在醫院內，護理師和醫生在意病人的健康狀況，當然，還有他們的位置。在工廠內，管理者不只希望監控機器人的操作行為更希望知道機器人的位置所在，在各式各樣的應用中，感測器和使用者的位置是最首要知道的事情。
雖然全球定位系統 （Global Positioning System, GPS）是一項有名的定位技術，但在大部分的室內環境中無法使用，一個常見的室內定位和追蹤方法是藉由已知位置的錨節點 （anchor）和搭配上射頻訊號 （radio frequency, RF），也就是說，被定位和追蹤的物體會發送信標 （beacon），接著，錨節點會基於信標的能量大小來測量距離，然而，由於訊號的能量大小容易受到多重路徑干擾 （multipath）和遮蔽 （shadowing）影響，使用射頻訊號最常見的問題是它大幅度的浮動，還有其他訊號像是超聲波 （ultrasound）和雷射 （laser）可以達到更好的距離測量結果，但這些方法需要額外的硬體設備以及安裝的費用。
不論使用何種方法來測量距離，為了能夠在現實生活中實作定位和追蹤技術，有兩個的挑戰是我們會面臨到的。首先，由於物件是會移動的，大部分驗量來源依賴於電池。因此，他們的電量是有限的且必須被適當的管理，如何達到省電同時又不失追蹤的精準度是個重要的議題。第二，在許多實際的場景下，夾帶著距離的資訊通常會經過大型的多中繼網路 （multi-hop network）進而透過本地端或是雲端來處理計算。因此，在短時間內，這些資訊的端對端 （end-to-end）可靠度也是一項重要的議題。此篇論文，我們針對這兩個挑戰提出可行的解決辦法。我們採用工作週期 （duty-cycling）機制，並考慮了微控制器的工作週期和最大信標傳送時間。加速度感測儀也被使用，在能量消耗和追蹤的精準度中達到一個平衡。針對端對端的可靠度，我們考慮了網路壅塞情形，藉由方向性和優先轉傳，距離資訊可以準時地被傳送。
我們實作出這些解決辦法來評估他的效益。在我們的測試平台中，共有四十八台錨節點和十二個移動裝置被佈建在台大博理館，端對端的傳送成功率在八個移動裝置的情況下可以達到八十五百分比，在一個移動裝置的情況下更可以達到九十六百分比。此外，平均追蹤誤差在靜止時為一點一公尺，在移動時為五點一公尺。在移動裝置的壽命方面，使用三百毫安培小時的電池狀況下可以長達兩年。實驗結果也顯示了我們提出的解決方案在實際環境中的可行性。

With the rapid rise of Internet of Things (IoT), it is convenient to connect the physical world to the Internet through wireless communication. IoT is composed of many sensors and wireless networks. IoT has been applied to a lot of applications such as environmental sensing, factory automation, healthcare monitoring, etc. In the hospital, nurses and doctors are interested in the health status of patients and, of course, the location of them. In the factory, the managers want to monitor not only the robot operations but also the locations of them. Among these various applications, knowing the locations of the sensors and users is great importance.
Although Global Positioning System (GPS) is a popular localization technology, it does not work in most of indoor environments. A common method for locating and tracking objects in indoor environments is to use known positions of anchors with the radio frequency (RF) signal. In short, objects to be located or tracked transmit beacons and then anchors estimate the distances based on the signal strength of beacons. A well-known problem of using the RF signal is its large variation as the received signal strength is often influenced by the multipath and shadowing. Some other signal signals such as ultrasound and the laser can be used to obtain better distance measurement but doing so incurs additional hardware and installation cost.
Regardless of the signal used to measure the distance, two other challenges need to be resolved in order to enable localization/tracking in a real environment. First, tracked objects are usually battery powered due to mobility. Thus, the power of it is limited and must be well managed. How to save the energy while maintaining tracking accuracy then becomes an important issue. Second, the distance info usually goes through a large and potentially multi-hop networks for process at a local or cloud server in many practical usage scenarios. Therefore, end-to-end reliability of transporting a very large number of distance messages in a short period of time is also a critical design issue. In this thesis, we focus on these two challenges and propose a feasible locating/tracking solution. We adopt a duty-cycling mechanism that takes MCU wake-up interval (MWI) and Maximum Beacon Transmission Interval (MBTI) into consideration. G-sensor is also used to make tradeoff between energy consumption and tracking accuracy. For end-to-end reliability, we take network congestion into account. By using directional and prioritized forwarding, distance messages can be delivered on time.
The proposed solution is implemented to evaluate its performance. In our test bed in the NTU BL building, a total of 48 anchors are installed while up to 12 tags are deployed. An end-to-end delivery rate of 85% can be reached in case with 8 tags. The rates even increase to 96% in case with 1 tag. In addition, the average tracking error is 1.1 meters when the tags are static and is 5.1 meters when the tags are motion. The lifetime of mobile tags is almost two years with a 300 mAH battery. The results show the feasibility of our solution in real-world environments.

Chapter 1: Introduction 1
1.1 RSSI-based localization 1
1.2 Power consumption issue 2
1.3 Network reliability 4
1.4 Conclusions 8
1.5 The organization of the thesis 9
Chapter 2: Related work 10
2.1 Distance detecting techniques 10
2.1.1 Range-based algorithm 10
2.1.2 Range-free algorithm 12
2.2 Power consumption issue 13
2.3 Network reliability 16
2.4 Summary 18
Chapter 3: System design 19
3.1 Mobile tags 20
3.2 Anchors 25
3.3 Gateways 28
3.4 Servers 30
3.5 Environmental setup 31
Chapter 4: Problem statement 32
4.1 Power consumption issue 32
4.2 Network reliability 34
Chapter 5 Proposed solutions and experimental results 39
5.1 Network reliability 39
5.1.1 Aggregation method 39
5.1.2 Randomize transmission time of RSSI reports (ITDI and RDI) 41
5.1.3 Prioritize RSSI reports 44
5.1.4 Restricted Implicit ACK 48
5.1.5 Optimal MRC 50
5.1.6 Optimal MWI 51
5.1.7 Summary 52
5.2 Power consumption issue 53
5.3 Tracking algorithm 61
5.3.1 Static cases 61
5.3.2 Motion cases 64
5.3.3 Results 65
Chapter 6: Conclusions 67
Reference 69

[1] CarlosMedina, J. ́CarlosSegura, and A.ńngelDelaTorre, “Ultrasound indoor po- sitioning system based on a low-power wireless sensor network providing sub- centimeter accuracy,” sensors, 2013.
[2] B. Choi, B. Kim, and E. Kim, “Location estimation and obstacle tracking using laser scanner for indoor mobile robots,” 2012.
[3] L. Zwirello, T. Schipper, M. Harter, and T. Zwick, “Uwb localization system for indoor applications: Concept, realization and analysis,” Journal of Electrical and Computer Engineering, 2012.
[4] B. Kim, B. Choi, and E. Kim, “Indoor localization using laser scanner and vision marker for intelligent robot,” Journal of Electrical and Computer Engineering, 2012.
[5] S. Anandamurugan and C. Venkatesh, “Power saving method for target tracking sensor networks to improve the lifetime,” International Journal of Recent Trends in Engineering, IEEE, 2009.
[6] A. Guidara, F. Derbel, and M. B. Jemaa, “Energy-e cient model for indoor localization process based on wireless sensor networks,” 13th International Multi- Conference on Systems, Signals and Devices (SSD), IEEE, 2016.
[7] Y. Xu, J. Winter, and W.-C. Lee, “Prediction-based strategies for energy saving in object tracking sensor networks,” IEEE International Conference on Mobile Data Management, 2004. Proceedings, IEEE, 2004
[8] B. Jiang, B. Ravindran, and H. Cho, “Probability-based prediction and sleep scheduling for energy-e cient target tracking in sensor networks,” pp. 467–478, IEEE Transactions on Mobile Computing, IEEE, 2013.
[9] F. Xia1, X. H. Liu, D. Zhang, and W. Z. Yang, “Energy-e cient opportunistic localization with indoor wireless sensor networks,” in Dynamic Spectrum Access Networks (DySPAN), 2017 IEEE International Symposium, 2011.
[10] K. A. Denney, M. Hamada, and Y. Nagao, “High-accuracy positioning system based on toa for industrial wireless lan,” 4th NAFOSTED Conference on Infor- mation and Computer Science, 2017.
[11] B. Barua, N. Zaarour, and N. Hakem, “E ect of uwb channel time delay parame- ters on tdoa localization,” Sixth International Conference on Digital Information, Networking, and Wireless Communications (DINWC), 2018.
[12] T. L. N. Nguyen and Y. Shin, “A new approach for positioning based on aoa measurements,” 2013.
[13] D. Munoz, F. Bouchereau, C. Vargas, and R. E. Caldera, “Position location techniques and applications,” 2009.
[14] M. Srbinovska, C. Gavrovski, and V. Dimcev, “Localization estimation system using measurement of rssi based on zigbee standard,” vol. 19, no. 1, 2008.
[15] K. Benkic, M. Malajner, P. Planinsic, and Z. Cucej, “Using rssi value for distance estimation in wireless sensor networks based on zigbee,” IEEE 15th International Conference on Systems, Signals and Image Processing, 2008.
[16] F. Dalkılıç, U. C. Çabuk, and E. Arıkan, “An analysis of the positioning accuracy of ibeacon technology in indoor environments,” 2017.doi:10.6342/NTU201804034REFERENCES 71
[17] P. Fonseka and K. Sandrasegaran, “Indoor localization for iot applications using ngerprinting,” 2018.
[18] N. Bulusu, J. Heidemann, and D. Estrin, “Gps-less low-cost outdoor localization for very small devices,” 2000.
[19] D. N. Nath, “Dv based positioning in ad hoc networks,” 2003.
[20] D. Petrovic and R. Kanan, “Extremely low power indoor localization system,” IEEE Eighth International Conference on Mobile Ad-Hoc and Sensor, IEEE, 2011.
[21] orge Juan Robles, S. Tromer, and M. Quiroga, “Enabling low-power localization for mobile sensor nodes,” International Conference on Indoor Positioning and Indoor Navigation, 2010.
[22] S. Tilak, V. Kolar, and N. Abu-Ghazaleh, “Dynamic localization control for mobile sensor networks,” 24th IEEE International Performance, Computing, and Communications Conference, IEEE, 2005.
[23] C. wen You, Y. chao Chen, and J. rung Chiang, “Sensor-enhanced mobility prediction for energy-e cient localization,” 3rd Annual IEEE Communications Society on Sensor and Ad Hoc Communications and Networks, IEEE, 2006.
[24] T.-H. Lin, P. Huang, and H. hua Chu, “Enabling energy-e cient and quality lo- calization services,” Fourth Annual IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOMW’06), IEEE, 2006.
[25] L. Brás, M. Oliveira, N. B. de Carvalho, and P. Pinho, “Low power localization engine based on wireless sensor networks,” Ibersensor 2010, Lisbon, Portugal, 2010.