隨著網路技術和應用服務的需求越來越多元，帶動了全球移動通訊網路的蓬勃發展。其中適地性服務(Location Based Service, LBS)已成為廣泛探討之應用。而定位技術是推動LBS發展的重要關鍵。近年來由於低功耗廣域網路(Low-Power Wide-Area Network, LPWAN)的崛起，室外定位相關應用不再僅以GPS技術獨樹一格。且無線通訊技術同時兼具定位及通訊兩大功能，大幅提升LBS發展的彈性。在本研究中將透過機器學習法解決雜訊問題以提升定位模型的準確率。
本研究提出一整合型機器學習架構，透過資料預處理以及機器學習演算法來完成LPWAN於室外定位之分類模型建立與評估。傳統的定位演算法受限於必須取得距離以進行幾何定位，因此本研究提出Fingerprint定位機制，並結合機器學習演算法訓練出適用於特定場域之定位模型。在資料預處理方面，經實際量測發現各類別的訓練資料集中除以了有明顯的離群值之外，還有零星的偏差值，本研究使用非監督式學習中的分群演算法統一濾除上述雜訊，提升模型的穩定度。考量到濾除雜訊之後各類別被濾除的資料量不均等問題，為了避免訓練階段時某類別被過度訓練，本研究提出處理數據不平衡機制(Imbalance Learning)生成新的有效資料，除了解決數據不平衡問題，同時增加模型的多樣性。另外在進行上述所提出之架構訓練前，透過ANOVA進行初步的數據分析，將不具關聯性或帶有雜訊之特徵提前濾除，有效提高模型之穩定性及可靠性。
本研究中使用SMOTE方法生成新的有效數據，相當於增加數據採集的位置，準確率從95.36％提高到98.38％。另外為了避免SMOTE處理後連同生成雜訊數據以影響模型的穩定性，本研究事先透過分群演算法中的DBSCAN方法對數據群集化，進而濾除雜訊資料。其準確率從98.38％提高到99.17％。

As the increasing demand for network technology and application services, the global mobile communications network is booming. The most well-known service is Location-Based Service (LBS), and the localization technology is an important key to the advancement of LBSs. In recent years, since the rise of the Low-Power Wide-Area Network (LPWAN), GPS is no longer the only option for outdoor localization. In addition, wireless communication technology has both two functions of localization and communication. Greatly improve the flexibility of LBS application development. This study solves the noise problem through machine learning to improve the accuracy of the localization model.
This study proposes an integrated machine learning architecture that implements the classification model of LPWAN for outdoor localization through data preprocessing and machine learning algorithms. Traditional localization algorithms are limited by needing to obtain distance for geometric localization. Therefore, this study proposes the Fingerprint mechanism and combines machine learning algorithms to train a localization model suitable for a specific field. In terms of data preprocessing, through actual measurement can found that in addition to the obvious outliers, there are s some offset values in the training data sets of each class. This study uses the clustering algorithm in unsupervised learning to simultaneously filter out the above noises and improve the stability of the model. Considering the problem of different amounts of data be filtered out by each class after filtering out noise. To avoid overtraining of a certain class during the training phase, this study proposes the data imbalance mechanism to generate new valid data. This method is solving the problem of data imbalance, while increasing the diversity of the model. Before conducting the above-mentioned framework training, the first step will perform the feature evaluation, such as ANOVA mechanisms, and filter out the unrelated or with noise features as much as possible to improve the stability and reliability of the model.
In this study, SMOTE method be used to generate new valid data, which is equivalent to increasing the location of data collection, and the accuracy is increased from 95.36% to 98.38%. To avoid the noise data being expand by SMOTE and generating invalid data to affect the stability of the model, the DBSCAB method cluster data to filter out noise. The accuracy is increased from 98.38% to 99.17%.

[1] B. Islam, M. T. Islam and S. Nirjon, “Feasibility of LoRa for Indoor Localization,” Technical Reports produced by the Department of Computer Science at the University of North Carolina at Chapel Hill, 2017.
[2] E. Goldoni, L. Prando, A. Vizziello, P. Savazzi and P. Gamba, “Experimental Data Set Analysis of RSSI‐based Indoor and Outdoor Localization in LoRa Networks,” Internet Technology Letters, Vol.2, No.1, pp. 75-80, 2018.
[3] B. Mukhopadhyay, S. Sarangi and S. Kar, “Novel RSSI Evaluation Models for Accurate Indoor Localization with Sensor Networks,” Proceedings of the Twentieth National Conference on Communications (NCC), pp. 1-6, 2014.
[4] M. Yassin, E. Rachid and R. Nasrallah, “Performance Comparison of Positioning Techniques in Wi-Fi Networks,” Proceedings of the International Conference on Innovations in Information Technology (INNOVATIONS), pp. 75-79, 2014.
[5] A. Yassin et al., “Recent Advances in Indoor Localization: A Survey on Theoretical Approaches and Applications,” IEEE Communications Surveys & Tutorials, Vol. 19, No. 2, pp. 1327-1346, 2017.
[6] Y. Geng, J. He, H. Deng and K. Pahlavan, “Modeling the Effect of Human Body on TOA Ranging for Indoor Human Tracking with Wrist Mounted Sensor,” Proceedings of the International Symposium on Wireless Personal Multimedia Communications (WPMC), pp. 1-6, 2013.
[7] R. Kaune, “Accuracy Studies for TDOA and TOA Localization,” Proceedings of the International Conference on Information Fusion (FUSION), pp. 408-415, 2012.
[8] A.A. Wahab, A. Khattab and Y.A. Fahmy, “Two-Way TOA with Limited Dead Reckoning for GPS-Free Vehicle Localization Using Single RSU,” Proceedings of the International Conference on ITS Telecommunications (ITST), pp. 244-249, 2013.
[9] J.Y Yoon, J.W. Kim, W.H Lee and D.S. Eom, “A TDOA-based Localization Using Precise Time-Synchronization,” Proceedings of the International Conference on Advanced Communication Technology (ICACT), pp. 1266-1271, 2012.
[10] S. Uebayashi, M. Shimizu and T. Fujiwara, “A Study of TDOA Positioning Using UWB Reflected Waves,” Proceedings of the IEEE Vehicular Technology Conference (VTC Fall), pp. 1-5, 2013.
[11] S. Shuo, S. Hao and S. Yang, “Design of an Experimental Indoor Position System Based on RSSI,” Proceedings of the International Conference on Information Science and Engineering (ICISE), pp. 1989-1992, 2010.
[12] R. Carvalho, Shan-Ho Yang, Yao-Hua Ho and Ling-Jyh Chen, “Indoor Localization Using FM and DVB-T Signals,” Proceedings of the IEEE Annual Consumer Communications & Networking Conference, pp. 862-867, 2016.
[13] M. E. Rida, F. Liu, Y. Jadi, A. A. A. Algawhari and A. Askourih, “Indoor Location Position Based on Bluetooth Signal Strength,” Proceedings of the International Conference on Information Science and Control Engineering, pp. 769-773, 2015.
[14] F. Palumbo, P. Barsocchi, S. Chessa and J. C. Augusto, “A Stigmergic Approach to Indoor Localization using Bluetooth Low Energy Beacons," Proceedings of the IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), pp. 1-6, 2015.
[15] N. Pirzada, M. Y. Nayan, M. F. Hassan, F. Subhan and H. Sakidin, “WLAN Location Fingerprinting Technique for Device-Free Indoor Localization System,” Proceedings of the International Conference on Computer and Information Sciences, pp. 650-655, 2016.
[16] M.M. Atia, A. Noureldin and M.J. Korenberg, “Dynamic Propagation Modeling for Mobile Users' Position and Heading Estimation in Wireless Local Area Networks,” IEEE Wireless Communications Letters, Vol. 1, No. 2, pp. 101-104, 2012.
[17] S. Alamri, D. Taniar and M. Safar, “Indexing Moving Objects in Indoor Cellular Space,” Proceedings of the International Conference on Network-Based Information Systems (NBiS), pp. 38-44, 2012.
[18] T. Janssen, M. Weyn and R. Berkvens, “Localization in Low Power Wide Area Networks Using Wi-Fi Fingerprints,” Applied Sciences, Vol. 7, No. 9, 2017.
[19] A. Thaljaoui, T. Val, N. Nasri and D. Brulin,” BLE Localization Using RSSI Measurements and iRingLA,” Proceedings of the IEEE International Conference on Industrial Technology (ICIT), pp. 2178-2183,2015.
[20] N. Kuxdorf-Alkirata and D. Brückmann, “Reliable and Low-cost Indoor Localization Based on Bluetooth Low Energy,” Proceedings of the International Symposium on Wireless Systems within the Conferences on Intelligent Data Acquisition and Advanced Computing Systems, pp.92-96, 2016.
[21] S. He and S. -. G. Chan, “INTRI: Contour-Based Trilateration for Indoor Fingerprint-Based Localization,” IEEE Transactions on Mobile Computing, Vol. 16, No. 6, pp. 1676-1690, 2017.
[22] T. I. Chowdhury et al., “A Multi-step Approach for RSSi-based Distance Estimation using Smartphones,” Proceedings of the International Conference on Networking Systems and Security (NSysS), pp. 1-5, 2015.
[23] Y. Shu, Y. Huang, J. Zhang, P. Coue, P. Cheng, J. Chen and K. G. Shin, ”Gradient-Based Fingerprinting for Indoor Localization and Tracking,” IEEE Transactions on Industrial Electronics, Vol. 63, No. 4, pp. 2424-2433, 2016.
[24] R. Faragher and R. Harle, “Location Fingerprinting with Bluetooth Low Energy Beacons,” IEEE Journal on Selected Areas in Communications, Vol. 33, No. 11, pp. 2418-2428, 2015.
[25] E. S. Lohan, J. Talvitie, P. Figueiredo e Silva, H. Nurminen, S. Ali-Löytty and R. Piché, “Received Signal Strength Models for WLAN and BLE-based Indoor Positioning in Multi-floor Buildings,” Proceedings of the International Conference on Location and GNSS (ICL-GNSS), pp. 1-6, 2015.
[26] C. Dong, J. Ding, J. Lin, “Segmented Polynomial RSSI-LQI Ranging Modelling for ZigBee-based Positioning Systems,” Proceedings of the Chinese Control Conference(CCC), pp. 8387- 8390, 2016.
[27] M. Centenaro, L. Vangelista, A. Zanella and M. Zorzi, “Long-range Communications in Unlicensed Bands: the Rising Stars in the IoT and Smart City Scenarios,” IEEE Wireless Communications, Vol. 23, No. 5, pp. 60-67, 2016.
[28] U. Raza, P. Kulkarni and M. Sooriyabandara, “Low Power Wide Area Networks: An Overview,” IEEE Communications Surveys & Tutorials, Vol. 19, No. 2, pp. 855-873, 2017.
[29] E. Kail, A. Banati, E. Lászlo and M. Kozlovszky, “Security Survey of Dedicated IoT Networks in the Unlicensed ISM Bands,” IEEE 12th International Symposium on Applied Computational Intelligence and Informatics (SACI), pp. 449-454, 2018.
[30] X. Lin et al., “Positioning for the Internet of Things: A 3GPP Perspective,” IEEE Communications Magazine, Vol. 55, No. 12, pp. 179-185, 2017.
[31] S. Hu, A. Berg, X. Li and F. Rusek, “Improving the Performance of OTDOA Based Positioning in NB-IoT Systems,” Proceedings of the GLOBECOM 2017 - 2017 IEEE Global Communications Conference, pp. 1-7, 2017.
[32] K. Radnosrati, G. Hendeby, C. Fritsche, F. Gunnarsson and F. Gustafsson, “Performance of OTDOA Positioning in Narrowband IoT Systems,” IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), pp. 1-7, 2017.
[33] Q. Song, S. Guo, X. Liu and Y. Yang, “CSI Amplitude Fingerprinting-Based NB-IoT Indoor Localization,” IEEE Internet of Things Journal, Vol. 5, No. 3, pp. 1494-1504, 2018.
[34] S. Gezici et al., “Localization via Ultra-wideband Radios: a Look at Positioning Aspects for Future Sensor Networks,” IEEE Signal Processing Magazine, Vol. 22, No. 4, pp. 70-84, 2005.
[35] H. Sallouha, A. Chiumento and S. Pollin, “Localization in Long-range Ultra Narrow Band IoT Networks using RSSI,” Proceedings of the IEEE International Conference on Communications (ICC), pp. 1-6, 2017.
[36] T. Janssen, M. Aernouts, R. Berkvens and M. Weyn, “Outdoor Fingerprinting Localization Using Sigfox,” Proceeding of the International Conference on Indoor Positioning and Indoor Navigation (IPIN), pp. 1-6, 2018.
[37] M. Aernouts, R. Berkvens, K. Van Vlaenderen and M. Weyn, “Sigfox and LoRaWAN Datasets for Fingerprint Localization in Large Urban and Rural Areas,” Mathematics & Computer Science, Vol. 1, 2018.
[38] L. Rokach and O. Maimon, “Top-down Induction of Decision Trees Classifiers - a Survey,” IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), Vol. 35, No. 4, pp. 476-487, 2005.
[39] A. S. More and D. P. Rana, “Review of Random Forest Classification Techniques to Resolve Data Imbalance,” Proceedings of the 1st International Conference on Intelligent Systems and Information Management (ICISIM), pp. 72-78, 2017.
[40] D. Misha , M. David and D.F. Nando, “Narrowing the Gap: Random Forests In Theory and In Practice,” Proceedings of the International Conference on Machine Learning, pp. 1-9, 2014.
[41] V. Barnett and T. Lewis, “Outliers in Statistical Data,” Wiley Series in Probability and Mathematical Statistics, 1994.
[42] B. C. Fargas and M. N. Petersen, “GPS-free Geolocation using LoRa in Low-Power WANs," Proceedings of the Global Internet of Things Summit, pp. 1-6, 2017.
[43] A. S. More and D. P. Rana, “Review of Random Forest Classification Techniques to Resolve Data Imbalance,” Proceedings of the 1st International Conference on Intelligent Systems and Information Management (ICISIM), pp. 72-78, 2017.
[44] Z. Zheng, Y. Cai and Y. Li, “Oversampling Method for Imbalanced Classification,” Computing and informatics, Vol. 34, No. 5, pp. 1-21, 2015.
[45] H. Han, W. Y. Wang and B. H. Mao, “Borderline-SMOTE: A New Over-sampling Method in Imbalanced Data Sets Learning,” Proceedings of the 2005 international conference on Advances in Intelligent Computing, pp. 278-887, 2005.
[46] K. Lam, C. Cheung and W. Lee, “LoRa-based Localization Systems for Noisy Outdoor Environment,” Proceedings of the IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), pp. 278-284, 2017.
[47] K. Lam, C. Cheung and W. Lee, “New RSSI-Based LoRa Localization Algorithms for Very Noisy Outdoor Environment,” Proceedings of the IEEE Annual Computer Software and Applications Conference (COMPSAC), pp. 794-799, 2018.
[48] LoRa-Alliance, A Technical Overview of LoRa and LoRaWAN; LoRa-Alliance, San Ramon, 2015.
[49] M. Aernouts, B. Bellekens, R. Berkvens and M. Weyn, “A Comparison of Signal Strength Localization Methods with Sigfox,” Proceeding of the Workshop on Positioning, Navigation and Communications (WPNC), pp. 1-6, 2018.
[50] G. G. L. Ribeiro, L. F. d. Lima, L. Oliveira, J. J. P. C. Rodrigues, C. N. M. Marins and G. A. B. Marcondes, “An Outdoor Localization System Based on SigFox,” Proceedings of the Vehicular Technology Conference, pp. 1-5, 2017.
[51] N. Poursafar, M. E. E. Alahi and S. Mukhopadhyay, “Long-range Wireless Technologies for IoT Applications: A Review,” Proceedings of the Eleventh International Conference on Sensing Technology (ICST), pp. 1-6, 2017.