| 摘要: | 窄頻物聯網(narrowband Internet of Things, NB-IoT)是第三代合作夥伴計畫(the third generation partnership project, 3GPP)在規範中新指定的應用範例，旨在同時支持大量用戶設備(user equipment, UE)在大規模機器通信(massive machine-type communication, mMTC)的場景中運作。上行鏈路中，與移動通信網路中UE啟動隨機接入(random access, RA)過程以請求上行資源相似，NB-IoT中的UE透過窄頻物理隨機存取通道(narrowband physical random access channel, NPRACH)發送前導碼，並藉其作為檢測UE的依據。RA作為處理UE及基地台間的首要程序，目標是讓UE能被基地台識別並進行同步，在此過程，除了要檢測出所有活躍的UE，還必須對其同步參數進行估計才能及時回傳給UE做修正，包含載波頻率偏移(carrier frequency offset, CFO)與時序誤差(timing error, TE)。本文主要對於NB-IoT上行鏈路的同步方案，在檢測方法中採用了基於奈曼-皮爾森準則(Neyman-Pearson criterion)的符元平均功率檢測；對於參數的估計則採用最大聯合概似估計(joint maximum-likelihood estimation, JMLE)方法。基於上述方案，考量了加性白高斯雜訊(additive white Gaussian noise, AWGN)通道以及3GPP協定所定義的仿真多路徑衰落通道模型，擴展步行模型(extended pedestrian A model, EPA)通道進行模擬，並且討論了在有CFO以及TE之下對於檢測出的誤報機率(false alarm probability)、漏檢機率(miss detection probability)所造成的影響。
;Narrowband Internet-of-Things (NB-IoT) is a new application specified in the specifications of the third generation partnership project (3GPP), aiming to support a large number of user equipment devices (UEs) operating in the scenario of massive machine-type communication (mMTC). In the uplink, similar to the random access (RA) process initiated by UE in mobile communication networks to request uplink resources, a UE in NB-IoT sends preambles via the narrowband physical random access channel (NPRACH) and the serving base station detects the preamble for the active UE. RA serves as a primary procedure for handling UE and base station interactions, with the goal of enabling base stations to identify and synchronize with UE. In this process, besides detecting all active UEs, it is also necessary to estimate their synchronization parameters to provide timely feedback to UE for correction, including carrier frequency offset (CFO) and timing error (TE). This thesis mainly focuses on the synchronization scheme for the NB-IoT uplink. In the detection method, a symbol-average power detection based on the Neyman-Pearson criterion is employed; for parameter estimation, the joint maximum-likelihood estimation (JMLE) method is used. Based on the above scheme, considering the additive white Gaussian noise (AWGN) channel and the simulation multipath fading channel model defined by the 3GPP protocol, the extended pedestrian A model (EPA) channel is simulated, and the impacts caused by false alarm probability and miss detection probability with the presence of CFO and TE are investigated. |
