中文摘要：

针对多径干扰下调制信号识别难的问题,提出了一种基于循环谱的调制信号识别方法.根据信号循环谱的性质,提取频率谱截面和循环谱截面的相关系数作为特征参数,并证明了该特征参数能够消除多径信道参数的影响,克服了利用循环谱或循环相关系数识别信号时受多径干扰影响大的缺点.仿真结果表明：在信噪比为0dB的多径瑞利衰落信道下,2PSK,4QAM和ODFM信号的识别率达到90%以上,与基于循环谱的方法相比,该方法识别率有显著的提高,且具有更强的抗多径能力.

英文摘要：

In order to solve the difficulty in classifying a variety of digital signal types for multipath fading channel,using cyclic spectral analysis,a classification method was proposed,according to the characteristic of cyclic spectral analysis,the correlation coefficients between the frequency spectrum′s section and cyclic spectrum′s were calculated,as the feature parameter of the modulation identification.Through theoretical analysis,it is certified that the extracted feature parameter can eliminate the influence of multipath channel parameter.Thus,this approach overcomes the disadvantage that the probability of correct classification is strongly influenced by multipath fading channel in traditional methods based on cyclic spectrum or cyclic correlation coefficient.Monte Carlo simulation results show that in identification of 2PSK（2-ary phase shift keying）,4QAM（4-ary quadrature amplitude modulation） and OFDM（orthogonal frequency division multiplexing） signals,the proposed method yields the probability of correct classification of 90% above in multipath fading channel when the signal-to-noise ratio is 0 dB.Compared with previous approaches based cyclic spectrum,the method proposed by this paper achieves better identification performance.Besides,this method has better anti-multipath interference capability.