ganmi

DEMO MSER WITH DATA OF IDCAR 2015

支持向量机，用于分类，含训练集与测试集，用SVM进行多分类(Support Vector Machine (SVM), used for classification, including training set and test set, and multi-support vector machine (SVM))

Least Mean Square (LMS) equalizer used in coherent receivers

利用差分进化算法（Differential Evolution）实现对经典函数foxhole的验证 (The use of differential evolution algorithm (Differential Evolution) to achieve the verification of classical function foxhole)

K近邻 KNN分类器，读取水质处理数据，并分好类，一共四类(K neighbor KNN classifier, read the water quality data processing, and a good class, a total of four class)

接收端采用的算法和程序流程与发送端发送的OFDM符号的帧结构有关系。具体的帧结构，以及定时估计，频偏估计，剩余误差跟踪的算法可参考算法说明文档。这里对程序的流程进行说明。 首先根据短训练字的特性进行相关运算，进行信号到达检测，当检测到相关值大于门限一定次数后，认为有信号到达。然后根据长训练字的特性，进行相关运算，进行OFDM符号FFT窗口起始位置的估计。估计出FFT窗口的位置后，先在时域进行小频偏的估计，将两个长训练字进行小频偏补偿后，进行FFT运算，根据FFT运算的结果进行整数倍频偏的估计。这些参数估计完成后，就可以进行数据解调了。先对数据部分进行完整的频偏补偿，然后根据估计的FFT窗口位置进行FFT运算得到频域的数据，进行解调。然后在对应于导频的子载波位置上提取出导频信息，根据导频信息估计出剩余定时误差以及剩余的信道响应误差，将误差量送入环路进行跟踪。当收到所有数据后，重新回到信号到达检测状态，进行下一次信号到达的检测和信号接收。 (The receiving end uses the algorithm and the program flow and the frame structure of the OFDM symbol transmitted by the sending end. Specific frame structure, and timing estimation, frequency offset estimation, residual error tracking algorithm can refer to the documentation. Here to illustrate the process of the program. First, according to the characteristics of the short training word correlation operation, the signal reaches the detection, when the correlation value is greater than the threshold of a certain number of times, that there is a signal to reach. Then, according to the characteristics of the long training word, the correlation operation is carried out, and the initial position of the OFDM symbol FFT is estimated. The estimated position of the FFT window, first in the time domain for small frequency offset estimation, two long training word for small frequency offset compensation, the FFT operation. According to the result of FFT for integer frequency bias estimation. Af)

Sparse Subspace Clustering基于人脸分割的子空间聚类的原始代码，已验证没有错误，大家可以放下下载。(Sparse Subspace Clustering segmentation based on face subspace clustering source code, no errors have been verified, you can put down the download.)

物理层网络编码，也称为双向协同通信，采用的是BPSK加差分调制，主要分析系统的BER曲线(Physical layer network coding, also known as two-way collaborative communication, using BPSK differential modulation increases, the BER curve of the main analysis system)

The attached file is a tutorial on matlab that will provide the necesssary guidance to those programming in matlab.

Gaussian matlab extractor