歐規數位電視系統所使用的傳輸架構為正交分頻多工(Orthogonal Frequency Division Multiplexing, OFDM),OFDM系統可以支援大量的資料傳輸,但是它對於同步(Synchronization)的要求卻更加嚴格。 本篇論文描述了三種同步的架構,包含了最大相似估測法(Maximum Likelihood Detection, ML)、整體相關性估測法(Ensemble Correlation Detection, ECD)和部分相似估測法(Part Likelihood Detection, PLD),將這三種演算法應用於DVB-T系統上評估其效能,改善接收機接收訊號時因為時間偏移(Timing Offset)和頻率偏移(Frequency Offset)所造成的誤差。 最後將傳統數位電視接收機的同步與通道估測演算法換成部分相似估測法與時域下自相關估測法(Frequency Domain Pilot Time Domain Correlation Method, FPTC)後,在固定接收的情況下效能增加了36.7%,時速40km/h接收的情況下效能增加了29.3%。
DVB-T employs Orthogonal Frequency Division Multiplexing (OFDM) as its modulation method. OFDM is known for its better transmission rate, however, it is quite strict with requirements of synchronization. This paper describes, applies, and evaluates three different synchronization architectures for DVB-T system, including Maximum Likelihood Detection (ML), Ensemble Correlation Detection (ECD), and Part Likelihood Detection (PLD), which are used to reduce the BER performance at the receiving side due to timing offset and frequency offset. Finally, after using PLD and Frequency Domain Pilot Time Domain Correlation Method(FPTC)to replace the conventional synchronization and channel estimation scheme of DVB-T receiver. The reception performance increases 30.7% and 29.3%, respectively, in fixed and 40km/hr cases.