Gini

fastica算法的matlab实现,比较简单，容易看懂(fastica algorithm matlab realize, is relatively simple, easy to understand)

波达方向（Direction Of Arrival）估计，又称为谱估计（spectral estimation）、波达角（Angle Of Arrival）估计。一个信源有很多可能的传播路径和到达角。如果几个发射机同时工作，每个信源在接收机处形成潜在的多径分量。因此，接收天线能估计出这些到达角就显得很重要，目的是破译出哪个发射机在工作以及发射机所处的可能位置。通过测量辐射信号的波达方向（DOA全称Direction Of Arrival）或波达角（AOA）来估测辐射源的位置。理论上这种估计只需要两个接收阵元就可以确定辐射源的位置，但在实际中，由于受到角度分辨率、多径和噪声限制，所需阵元通常要多于两个。DOA定位技术原理是利用接收机处的阵列天线和波达方向（DOA）估计技术来确定一个从接收机到信源的波达方向线，即为方向线（LOB），最后利用多个接收机估计的DOA进行三角测量，方向线的交点就是信源的估计位置。(DOA (Direction Of Arrival) estimation , also known as spectral estimation (spectral estimation), angle of arrival (Angle Of Arrival) estimation. There are many possible sources a propagation path and the angle of arrival . If several transmitters simultaneously, a potential source for each multipath component at the receiver. Therefore, the receiving antenna can estimate the angle of arrival becomes very important, which aims to decipher the location of the transmitter at work as well as possible in which the transmitter . By measuring the radiation signal DOA (DOA stands for Direction Of Arrival) or the angle of arrival (AOA) to estimate the location of the radiation source . Theoretically this estimate requires only two receiving array element can determine the position of the radiation source , but in practice , due to the angular resolution , multipath and noise limits , usually required more than two array elements . DOA positioning technology is the use of an array of antennas at)

FastICA Code. Independent Component Analysis code for signal extraction

说明：  contourlet算法作为二阶小波算法的延续，有很强大的应用空间，上传的文件是contourlet算法的工具箱(Contourlet algorithm as a second-order wavelet algorithm for the continuation of a very powerful application space and upload the document is Contourlet algorithm toolbox)

说明：  进行等时延分段，并对等间隔段的两个时延矩阵分别求取均值，采用类似联合近似对角化，估计出最优化的酉矩阵，(ing second-order statistics (different time-delay correlation matrix) of non-stationary and structural characteristics of the signal timing can roughly estimate that the instantaneous linear mixture of blind source signal. With the time delay increasing, only using a single delay covariance equalization and ignoring the time-varying characteristics of the signal, which difficult to ensure the performance of algorithm. Analysis of the average characteristics of ma)

说明：  能够很好的解决牛顿插值问题，对于构建差商表的程序给出了解释和说明！！！(Newton interpolation can be a good solution to the problem, divided difference table for building the program and instructions are given to explain! ! !)

MATLAB M文件调用SIMULINK完成LDPC误码率仿真(MATLAB M SIMULINK LDPC BER SIMULATION)

Deblurring Images Using a Regularized Filter This example shows how to use regularized deconvolution to deblur images. Regularized deconvolution can be used effectively when constraints are applied on the recovered image (e.g., smoothness) and limited information is known about the additive noise. The blurred and noisy image is restored by a constrained least square restoration algorithm that uses a regularized filter.

基于向量内积的相似性K-means聚类算法(Based on vector similarity of inner product K- means clustering algorithm)

这个方法有两个好处。 （ a ）虽然只是一个半径差逼近，很容易找到最高在一个小蓄电池，因为通过降低 隐平滑做的蓄电池。（ b ）在同一时间处理时间搜索减少 (Pattern Recognition and Image Procesing Group. The size of the accumulator is varied as hierarchical voting is performed. This has two advantages. [a] While the radius is only a poor approximation, it is easier to find the maximum in a small accumulator and because by means of the downsampling implicit smoothing is done to the accumulator. [b] At the same time processing time for the search is reduced.)