| 摘要: | 窄頻物聯網
(Narrowband Internet-of-Things, NB-IoT) 為第三代合作夥伴計畫 (3rd Generation Partnership Project, 3GPP)在低功率廣域網路 (Low-Power Wide-Area Network, LPWAN)訂定的標準，是一種新型蜂巢式網路技術 (cellular network) 亦為 低功耗廣域網
路技術種類之一，能夠滿足用戶與企業相互連接的需求以及擁有更遠的行動設備服務範
圍，具有的優勢包含可以大量連接物聯網裝置、覆蓋範圍廣泛、低功耗，可以延長電池
使用壽命、成本低廉等並能支援各式不同的應用，例如用於智慧城市、環境監測等。
本論文研究了一種的技術，用於有效檢測通過窄頻物聯網實體隨機存取通道
(Physical Random Access Channels, NPRACH)傳輸的重疊前導序列。由於通道的準靜態性
在幾個符元期間保持本地相位一致性，因此本文採用符號級匹配濾波器 (Symbol-level Matched Filter, SLMF) 以有效降低帶外雜訊和干擾，而為了減少由不同載波頻率偏移
(Carrier Frequency Offsets, CFOs) 引起的載波間干擾 (Intercarrier Interference, ICI)，本文
研究了一種基於干擾重建和消除 (Interference Regeneration and Cancellation, IRnC) 的前
向錯誤糾正方法來有效降低遺漏 機 率 通過將量化的正規化 CFO (Normalized CFO, NCFO) 估計 值反饋給相應的 UE 並減少 CFO，最後模擬驗證了理論分析並展示了所提
出技術實現低誤報和遺漏 機 率 的能力。;Narrowband Internet-of-Things (NB-IoT) is defined by the 3rd Generation Partnership Project (3GPP) in Low-Power Wide-Area Network (LPWAN). NB-IoT can meet the connectivity needs of users and businesses, providing extended service coverage for mobile devices. Its advantages include the ability to connect a large number of IoT devices, wide coverage, low power consumption, extended battery life, and cost-effectiveness. It supports various applications, such as smart cities and environmental monitoring.
This thesis explores a novel technique for effectively detecting overlapping preamble sequences transmitted through the Narrowband IoT Physical Random Access Channels (NPRACH). Due to the quasi-static nature of the channel, which maintains local phase coherence over multiple symbol periods, the thesis adopts a symbol-level matched filter (SLMF) to efficiently reduce out-of-band noise and interference. In order to mitigate intercarrier interference (ICI) caused by carrier frequency offsets (CFOs) from different carriers, the paper investigates a forward error correction method based on Interference Regeneration and Cancellation (IRnC) to effectively reduce the probability of missing detections. By feeding back quantized normalized CFO (NCFO) estimates to the corresponding user equipment (UE) and reducing CFO, the proposed technique demonstrates the capability to achieve low false alarm and missing detection probabilities, as validated through theoretical analysis and simulation. |
