利用LPCC在數位調變型態之辨識研究

Study of Automatic Digital Modulation Classification by Using Linear Prediction Derived Cepstral Coefficient

周瑞雄(Jui-Hsiung Chou) ； 彭正文(Cheng-Weu Peng)

《中華科技大學學報》 44期 (2010/07) Pp. 71-88

https://doi.org/10.7095/JCUST.201007.0071

摘要

本文是對事先未知的數位調變系統訊號，利用線性預估倒頻譜係數(Linear Prediction Derived Cepstral Coefficient; LPCC)的演算法來做數位調變系統辨識。本文要辨識的數位調變信號為幅移鍵、頻移鍵、直序展頻及多載波分碼展碼。首先產生各類數位調變訊號之轉置.WAV檔。接著求取線性預估係數(LPC)，再將線性預估參數轉換為線性預估倒頻譜係數(LPCC)，當作調變訊號之特徵參數。最後經過模擬的結果，發現以線性預估倒頻譜參數演算法來做辨識，確實有相當不錯的結果。

關鍵字

線性預估係數 ； 線性預估倒頻譜係數 ； 數位調變

並列摘要

The purpose of this paper is to develop an automatic digital modulation classification algorithm using linear prediction derived cepstral coefficient (LPCC). The digital modulated signals are Amplitude Shift Keying (ASK), Frequency Shift Keying (FSK), Direct Sequence Code Division Spread Spectrum (DS-CDSS) and Multi-Carrier Code Division Spread Spectrum (MC-CDSS). First, generate variety of digital modulation signals transfer into .WAV file. Next, the LPCC of samples of received digital modulation signals are chosen as features. Finally, the features of unknown input data are computed, and then its modulation type is classified into the one of given modulation types. The results of computer simulation show that this classification algorithm is effective for recognizing the digital modulation type of received signals.

並列關鍵字

linear prediction derived coefficients ； linear prediction derived cepstral coefficients ； digital Modulation