easyspin-2.6.0

pe to snr for equalizer zfl and mmse for channel [1 .55]/norm([1 .55]) and tap=5

设计一维纳滤波器 (1)产生三组观测数据：首先根据s(n)=a s(n-1)+w(n)产生信号s(n)，将其加噪（信噪比分别为20dB,10 dB，6 dB），得到观测数据x1(n), x2(n), x31(n). (2)估计xi(n)，i=1,2,3的AR模型参数。假设信号长度为L，AR模型阶数为N，分析实验结果，并讨论改变L,N对实验结果的影响。 (A Wiener filter design (1) to produce three sets of observations: first generating a signal s (n) according to s (n) = as (n-1)+w (n), to add noise (20dB SNR, respectively , 10 dB, 6 dB), to obtain observational data x1 (n), x2 (n), x31 (n). (2) estimate xi (n), i = AR model parameters 1,2,3. Assuming the signal length L, AR model of order N, analysis of experimental results and discuss the impact changing L, N of the experimental results.)

The zip file contains several matlab files. These files implement the optimal vlassifier (optimal_classifier.m), a sub-optimal classifier (SOS_STBC_Classifier.m) which is based on Second Order Statistic and a low-complexity sub optimal code-parameter classifier (CP_Classifier.m). See reference [1] for the theoretical background. Example: To create the fig 3,4 and 5 of the paper, call the routine: Blind_recognition_STBC2_simulation_results(512,4, PSK ,4) Reference: [1] V. Choqueuse, M. Marazin, L. Collin, K. Yao and G. Burel "Blind recognition of STBC: A Likelihood-based Approach", IEEE Transactions on Signal Processing (IF: 2.33), vol. 58(3), pp : 1290-1299, March 2010.

fsk simulation in Matlab

四阶龙格库塔公式代码 控制水箱的程序，方便简单(Fourth-order Runge-Kutta formula code control tank program, convenient and simple)

　K-means算法是最为经典的基于划分的聚类方法，是十大经典数据挖掘算法之一。K-means算法的基本思想是：以空间中k个点为中心进行聚类，对最靠近他们的对象归类。通过迭代的方法，逐次更新各聚类中心的值，直至得到最好的聚类结果。(K-means algorithm is based on the division of the classic clustering method, is ten classic one of data mining algorithm. K-means the basic idea of the algorithm is: to the space K point as the center of the cluster analysis, near their object classification. Through the iterative method, each successive update clustering center value, until get the best clustering results. )

matlab code for The INface toolbox v2.0 for illumination invariant face recognition

Algorithm for frequency analysis usign FFT

用二次插值求f(x)最优解。根据输入确定三点的初始值，并求出对应的函数值。(With quadratic interpolation find f (x) the optimal solution. According to the initial value of the input to determine the three points, and find the corresponding function value)

说明：  针对一般的粒子群优化(PSO)学习算法中存在的容易陷入局部最优和搜索精度不高的缺点,对改进型PSO算法进行研究。由于惯性权重系数ω对算法是否会陷入局部最优起到关键的作用,因此,通过改变惯性权重ω的选择,对惯性权重系数采取线性减小的方法,引入改进型的PSO算法。采用改进的PSO算法对PID控制器进行参数优化并把得到的最优参数应用于控制系统中进行仿真。仿真实验结果表明:改进型PSO算法不会陷入局部最优,能得到全局最优的PID控制器的参数,并使得控制系统的性能指标达到最优,控制系统具有较好的鲁棒性。(Aiming at the shortcomings of general PSO learning algorithm, which is easy to fall into local optimum and low search precision, the improved PSO algorithm is studied)