针对通信信号的调制识别问题,首先根据通信信号的循环平稳性,提出一种基于循环自相关的OFDM信号和单载波信号的调制识别算法,然后将小波多分辨分析理论与调制信号的瞬时特征以及高阶累积量相结合,提出一种基于小波分解的单载波信号识别方法,在此基础上采用分层结构的神经网络分类器对OFDM,2ASK,4ASK,2PSK,4PSK,8PSK,16QAM这7种调制信号进行识别。仿真结果表明该方法具有良好的分类性能,且对噪声不敏感。
For the problem of modulation recognition of communication signals, an algorithm based on cyclic autocorrelation is proposed to recognize OFDM signals and single-carrier signals according to the cycle-stationarity of communication signals. Then a single-cartier signals recognition method is proposed based on wavelet decomposition whicb combines wavelet theory of muhiresolution analysis with modulated signals' instantaneous characteristics and high-order cumulants. What' s more, a hierarchical neural network classifier is used to identify seven kinds of modulation signals as OFDM, 2ASK, 4ASK, 2PSK, 4PSK, 8PSK, 16QAM. The simulation results show that the method is high in performance and not sensitive to noise.