时间序列分析，分析序列是否是 白色噪声，分析相关性，可以进一步判断序列的相关和自相关-time series analysis, whether the sequence is white noise, correlation analysis, further sequence of judgment and autocorrelation

把辨识分成两步进行：第一步：利用相关分析法获得对象的非参数模型（脉冲响应或相关函数）；第二步：利用最小二乘法、辅助变量法或增广最小二乘法等，进一步求的对象的参数模型。如果模型噪声与输入无关，则Cor-ls相关最小二乘法（二步法）可以得到较好的辨识结果。Cor-ls相关最小二乘法（二步法）实质上是先对数据进行一次相关分析，滤除了有色噪声的影响，再利用最小二乘法必然就会改善辨识结果。能适应较宽广的噪声范围，计算量不大，初始值对辨识结果影响较小。但要求输入信号与噪声不相关-The identifica

ML法能够很好的估计信号的信噪比,包括邓氏关联度、绝对关联度、斜率关联度、改进绝对关联度,保证准确无误，是学习通信的好帮手。- ML estimation method can be a good signal to noise ratio, Including Deng s correlation, absolute correlation, correlation of slope, improved absolute correlation, Ensure accurate communi

实现了图像的加水印，去噪，加噪声等功能，详细画出了时域和频域的相关图，ofdm系统仿真 含16qam调制 fft 加窗 加cp等模块。- Realize image watermarking, de-noising, plus noise and other functions, Correlation diagram shown in detail the time domain and frequency domain, ofdm system simulation including 16

Suppressed carrier type differential phase modulation, Independent component analysis algorithm reduces the raw data noise, For feature reduction, feature fusion, correlation analysis.