臺灣博碩士論文加值系統

隨著4G邁入5G的時代，然而因為5G得到來，將為物聯網帶來極大的衝擊，能夠有更多的支援無線設備的連入和支援龐大的數據流量，並且以低於現今成本的聯網服務。而在5G正式上路之前，低功耗廣域網路(Low Power Wide Area Network,LPWAN)提供產業持續前進的動力，其中較受注目的包括使用授權頻段的NB-IoT(NarrowBand-IoT) 。近來，NB-IoT (FDD)的應用可能會受到在LTE (TDD)的架構下影響，當NB-IOT小基地台利用LTE的頻段進行上行資料傳送，在使用LTE的頻段的時候，會受到附近的LTE大基地台(Marco cell)的下行功率溢漏干擾。
為此，本篇論文即是研究在NB-IoT (FDD)在LTE (TDD)的環境下，如何有效且精準的配置功率。
最後模擬的結果可以看出本論文提出的KM-Q演算法在任何溢漏環境下都能夠配置出較佳的功率參數，並建立溢漏迴歸模型作為防止溢漏干擾之最終目標。

The 5G is coming soon after the 4G, the 5G will bring a great impact on the Internet of Things.It will be able to support the connection of wireless devices and support huge data traffic. It will provide IoT services at lower cost than today. Before the 5G is opened officially,The LPWAN (Low Power Wide Area Network,LPWAN) provides the power which the industry goes forward continually. It is worth noting that the use of licensed band NB-IoT (NarrowBand-IoT). Recently, the NB-IoT (FDD) application possibly can be affected by the LTE (TDD) construction. When the NB-IoT small base station uses the LTE frequency band for the uplink transmission.It can be interference by downlink power to leak from nearby the LTE eNB (Marco cell).
For this reason, this paper proposed the NB-IoT (FDD) how to effectively and accurately allocate the power under the LTE (TDD) environment.
Finally,it simulates the result indication the KM-Q algorithm which this paper proposed in any leakage environment all to be able to show the better power parameter, and establishes the leakage regression model achievement to prevent leakage the interference for the ultimate goal.

目錄
中文摘要...................................I
英文摘要...................................II
誌謝.......................................III
目錄.......................................IV
圖目錄.....................................VI
表目錄.....................................IX
第一章 序論..................................1
1-1 前言....................................2
1-2 研究動機.................................2
1-3 論文架構.................................3
第二章 相關研究背景...........................4
2-1 長期演進技術(LTE)介紹.....................4
2-2 窄頻物聯網(NB-IoT)介紹....................7
2-3 機器學習(Machine Learning)介紹...........13
2-4 K-Means演算法............................21
2-5 貪婪演算法...............................25
2-6 Q-Learnging演算法........................25
第三章 系統環境架構與演算法計..................28
3-1 系統架構…................................28
3-2 環境問題.................................30
3-3 溢漏規格與環境問題........................31
3-4 KM-Q演算法...............................34
3-5 NB-IoT NPUSCH 功率公式...................40
3-6 溢漏迴歸模型流程圖........................41
第四章 模擬與分析.............................43
4-1 模擬環境與參數定..........................43
4-2 模擬結果與析.............................44
4-3 溢漏迴歸型...............................63
第五章 結論與未來研究向.......................69
參考文獻.....................................70

參考文獻 :
[1].廖專崇,”LPWAN戰國六雄逐鹿廣域物聯網賽局鳴槍起跑”,新通訊 2017 年 2 月號 192 期《 封面故事 》

[2].3GPP TR 125 942 V11.00.00, July 26-29, 1999 ; Simulation results on FDD/TDD co-existence including real receive filter and C/I based power control.

[3].Zainatul Yushaniza Mohamed Yusoff; Aduwati Sali Fazirulhisyam Hashim ,“Energy efficiency of Coordinated Multipoint transmission (CoMP) over LTE-A”, IEEE 13th Malaysia International Conference on Communications (MICC),pp 305 – 310, 2017

[4].郭昱賢,林盈達,”LTE 架構、協定與效能”,國立交通大學資訊工程系, 2011年9月

[5].許亨仰,”邁向全IP/扁平化架構　LTE/SAE掀行動網路新革命”, 新通訊 2009 年 9 月號 103 期《 技術前瞻 》

[6].恆朗科技, ”NB-Iot技術介紹”, 2017

[7]. u-blox ,”Narrowband IoT (NB-IoT)” , 2018

[8].林家男, ”架構/傳輸/呼叫流程有訣竅NB-IoT核心網路細部分解”,新通訊 2017 年 6 月號 196 期《 技術前瞻 》

[9]. Huawei, “NB-IOT White Paper”, 2015.

[10].高通,”優化窄頻技術LTE IoT為5G物聯網打底”,新通訊 2016 年 11 月號 189 期《 技術前瞻 》

[11]. Mr.Opengate,”AI - Ch13 機器學習(1)，機器學習簡介與監督式學習Introduction to Machine Learning Supervised Learning”, 2016

[12]. KEVIN,”機器學習(Machine Learning)介紹”, 2016

[13].林惠玲 陳正倉, ”應用統計學,第十四章簡單迴歸分析與相關分析”,2006
https://www.cyut.edu.tw/~fmchang/docs/Chapter%2014.pdf
[14].蔡聰明,”最小平方法與迴歸分析”, 2017
http://web.math.sinica.edu.tw/mathmedia/HTMLarticle18.jsp?mID=41303

[15]. Mr.Opengate,”AI-Ch18 機器學習(6)，分群/聚類：K平均演算法 Clustering:K-means Algorithm”, 2015

[16].陳鍾誠,”K-Means 分群演算法”, 2013

[17].陳鍾誠,”貪婪演算法”, 2013

[18]. Mr.Opengate,”Q-learning與類神經網路：用OpenAI gym 模擬木棒台車平衡問題 (cart-pole system)” , 2016

[19].Supratim Deb; Pantelis Monogioudis, “Learning-Based Uplink Interference Management in 4G LTE Cellular Systems”, IEEE/ACM Transactions on Networking, pp: 398 – 411, 2015

[20]. Zhi Yu, Ke Wang, Hong Ji, Xi Li, Heli Zhang,” Dynamic Frame Configuration in TD-LTE Heterogeneous Cellular Networks ”, IEEE Globecom Workshops (GC Wkshps),pp:1-6, 2015

[21].Yuqing Wang; Zhanji Wu,” A Coexistence Analysis Method to Apply ACLR and ACS between NB-IoT and LTE for Stand-Alone Case”, International Conference on Instrumentation & Measurement, Computer, Communication and Control (IMCCC),
pp: 375 – 379, 2016

[22]. R4-158207,” Way forward on coexistence simulation assumptions”, Intel,Ericsson, Huawei, Nokia, Qualcomm.

[23].3GPP TR 25.942 (V13.0.0). Radio Frequency (RF) system
scenarios(Release 13), 2016.