脈衝雜訊為一種出現時間短但能量很強的干擾訊號，其來源可能是開關電器產品或者使用過程中所發出。在實際的通訊系統中脈衝雜訊通常會造成整個收到的訊號失真，且如果接收到的訊號進行解調處理時需要使用之前時間的訊號，脈衝雜訊會連帶影響到後來收到訊號的品質。 在歐規數位電視廣播（DVB）中，針對脈衝雜訊定義六種不同測試訊號[2]。本篇論文研究內容在不同規格參數的歐規數位電視廣播環境下，分別加入六種不同測試訊號進行脈衝雜訊的偵測與消除。 本論文先針對偵測脈衝雜訊部分：一般偵測方法與本論文提出偵測方法進行說明；接著針對脈衝雜訊消除：一般消除方式、S. V. Zhidkov於2003年於IEEE提出之方法、本論文提出之最小平方法（LS, Least Square）與尋找量化階取代（SOQL, Search Over Quantization Level）逐一進行說明。最後將各種偵測與消除脈衝雜訊方法進行模擬，由模擬結果可知使用本論文提出之偵測與消除脈衝雜訊方法可以有效降低脈衝雜訊對系統的影響。 Impulse noise is a kind of noise with short duration and strong interference. It comes from switching on/off electric appliances or the duration of using them. In practical communication system, impulse noise usually causes received signal to distort. If received signal needs the former signal to demodulate the instant signal, impulse noise will influence the later signal. DVB specification defines six different test patterns[2] for impulse noise. This thesis research content is simulation on different DVB-T specification with these test patterns. This thesis first talks about impulse noise detection including convention algorithm, proposed algorithm and then talks about impulse noise cancellation including convention algorithm, S. V. Zhidkov proposed algorithm, this thesis proposed algorithms Least Square (LS) and Search Over Quantization Level (SOQL). By simulating all these algorithm can find that proposed algorithm can effective reduce influence of impulse noise.
