目前利用無線通訊技術應用在室內的絕大多數是使用WIFI與藍芽，使用移動式通訊技術（2G,3G,4G,5G）來做室內定位的非常少數，大多都有著嚴格的限制，本論文建立一個透過接收多個4G Long Term Evolution（LTE）基地台的深度學習模型來實現室內定位用於辨識地點，本實驗環境在室內建立5個參考點，在每個參考點搜集多個基地台的訊號功率、訊號品質、頻率、基地台編號，4個重要參數來做訓練。在實驗過程中，訊號皆由Android 手機做接收，接收後的訊號經過Deep Neural Network (DNN)模型，最後輸出預測位置，此模型預測的位置分類可以達到99%準確率，藉此可以改善Global Positioning System(GPS)室內收訊不良的缺失，達到穩定的實際應用。

Currently, the vast majority of indoor applications that use wireless communication technology are Wi-Fi or Bluetooth. Very few people use mobile communication technology such as (2G, 3G, 4G, 5G) for indoor positioning, and most of them have strict restrictions. This paper establishes a deep learning model that receives multiple 4G Long Term Evolution (LTE) base stations to achieve indoor positioning for location identification. In this experimental environment, five reference points are established indoors, and data from multiple base stations are collected at each reference point. Signal power, Signal quality, Frequency, Cell ID, four important parameters for training. During the experiment, all the signals are received by the Android mobile phone. The received signals go through the Deep Neural Network (DNN) model, and finally output the predicted position. The predicted position classification of this model can reach 99% accuracy, which can improve the lack of poor indoor reception of the Global Positioning System (GPS), to achieve stable practical application.

[1] Ali Yassin et al, "Recent advances in indoor localization: A survey on theoretical approaches and applications", IEEE Communications Surveys & Tutorials 19(2): 1327-1346, 2016.
[2] Fahad Alhomayani and Mohammad Mahoor, "Deep learning-based symbolic indoor positioning using the serving eNodeB", 19th IEEE International Conference on Machine Learning and Applications (ICMLA): 489-494, 2020.
[3] Maurizio Rea et al, "Smartphone positioning with radio measurements from a single wifi access point", Proceedings of the 15th International Conference on Emerging Networking Experiments And Technologies: 200-206, 2019.
[4] Saideep Tiku, Prathmesh Kale, and Sudeep Pasricha, "QuickLoc: Adaptive Deep-Learning for Fast Indoor Localization with Mobile Devices", ACM Transactions on Cyber-Physical Systems (TCPS) 5(4): 1-30, 2021.
[5] Chung-Ming Own, Jiawang Hou, and Wenyuan Tao, "Signal fuse learning method with dual bands WiFi signal measurements in indoor positioning", IEEE Access 7: 131805-131817, 2019.
[6] Xuyu Wang et al, "CSI-based fingerprinting for indoor localization: A deep learning approach", IEEE Transactions on Vehicular Technology 66(1): 763-776, 2016.
[7] Daniel Halperin et al, "Tool release: Gathering 802.11 n traces with channel state information", ACM SIGCOMM computer communication review 41(1): 53-53, 2011.
[8] Jeongyeup Paek et al, "Energy-efficient positioning for smartphones using cell-id sequence matching", Proceedings of the 9th international conference on Mobile systems, applications, and services: 293-306, 2011.
[9] Salih Ergut, Ramesh R. Rao, and Ozgur Dural, "Localization via multipath strengths in a cdma2000 cellular network using neural networks", IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence): 4066-4069, 2008.
[10] Amel Mohamed et al, "DeepFeat: Robust Large-Scale Multi-Features Outdoor Localization in LTE Networks using Deep Learning", IEEE Access: 3400-3414, 2022.
[11] Luigi Serreli et al, "Fingerprint-based Positioning Method over LTE Advanced Pro Signals with GAN training contribute", IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB): 1-5, 2021.
[12] Giovanni Pecoraro et al, "CSI-based fingerprinting for indoor localization using LTE signals", EURASIP Journal on Advances in Signal Processing: 1-18, 2018.
[13] Xiaokang Ye et al, "Neural-network-assisted UE localization using radio-channel fingerprints in LTE networks", IEEE Access 5: 12071-12087, 2017.
[14] Yu-Siang Feng et al, "An RSSI-Based Device-Free Localization System for Smart Wards", IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW): 1-2, 2021.
[15] Matteo Ridolfi et al, "Experimental evaluation of UWB indoor positioning for sport postures", Sensors 18(1): 168, 2018.
[16] Alberto Belmonte-Hernández et al, "SWiBluX: Multi-Sensor Deep Learning Fingerprint for precise real-time indoor tracking", IEEE Sensors Journal 19(9): 3473-3486, 2019.
[17] Maurizio L. Rocca, "RSRP and RSRQ measurement in lte", laroccasolutions Technology & Services: 9, 2015.
[18] Evolved universal terrestrial radio access (E-UTRA); User equipment (UE) radio transmission and reception, 3GPP TS 36.101 version 12.5.0 Release 12, 2017.
[19] Hamada Rizk, Marwan Torki, and Moustafa Youssef, "CellinDeep: Robust and accurate cellular-based indoor localization via deep learning", IEEE Sensors Journal 19(6): 2305-2312, 2018.
[20] Erik W. Grafarend, Linear and nonlinear models: fixed effects, random effects, and mixed models, Berlin, Germany:Walter de Gruyter, 2006.
[21] Yann LeCun, Yoshua Bengio, and Geoffrey Hinton, "Deep learning", nature 521(7553): 436-444, 2015.
[22] Francois Chollet, Deep learning with Python, 2018.
[23] Johan Bjorck et al, "Understanding batch normalization", Advances in neural information processing systems 31, 2018.
[24] 郭清智,"應用 Wi-Fi 與 GPS 技術於室外定位之研究",碩士論文,國立政治大學地政研究所,2010.