基于小波变换的K-L变换去噪应用研究

在通信对抗电台指纹特征识别研究中,需要多个不同的传感器来采集数据.经多路传感器系统所采集到的原始信号数据,各传感器通道的有效信号之间具有较强的相关性,而随机噪声之间不具有相关性,K-L变换对不同通道间不相关随机噪声的去除具有比较大的优势.原始信号中可能含有多个频率,实现对各个频率成分分别去噪具有重要意义.用基于小波变换的K-L变换去噪方法进行电台指纹特征识别的应用.仿真结果显示对各频率分量基于小波变换的K-L变换去噪能够较好的实现信噪分离.

Application of K-L transform based on wavelet transform in reduction of noise

In view of the fact that there are correlations among the waveforms of effective signals and uncorrelated random noise of channels in the multi-sensor system, and there are multi-signals in the channel, the method of reduction of noise by using K-L transform based on the wavelet transform is brought forward in the radio RF object identification. The smulation results show that the method of reduction can effectively separate the noise from the effective signals with different frequency.