zad2

Matlab程序，用于比较线性均衡器、判决反馈均衡器(DFE)、盲最大似然序列估计均衡器(MLSE)等误码率性能。(Matlab procedures used to compare the linear equalizer, decision feedback equalizer (DFE), Blind maximum likelihood sequence estimation equalizer (MLSE), such as bit error rate performance.)

此程序为用matlab编写的基于最小二乘法通过求解广义矩阵特征值及特征向量进行椭圆曲线拟合程序。 程序中x,y分别为采样点的横纵坐标。生成的结果矩阵为椭圆曲线aX^2+bXY+cY^2+dX+eY+f=0中的各项系数。(This program was written with matlab by solving the generalized least square method based on matrix values and eigen vectors of the elliptic curve fitting procedure. Process x, y, respectively transverse and longitudinal coordinates of sampling points. The results generated matrix of elliptic curve aX ^ 2+ bXY+ cY ^ 2+ dX+ eY+ f = 0 in the coefficients.)

说明：  小波，功能分析，主要是图形分析方面，还可以图形压缩(matlab)

support vector classification machine ％ soft margin ％ uses "kernel.m" ％ ％ xtrain: (Ltrain,N) with Ltrain: number of points N: dimension ％ ytrain: (Ltrain,1) containing class labels (-1 or +1) ％ xrun: (Lrun,N) with Lrun: number of points N: dimension ％ atrain: alpha coefficients (from svcm_train on xtrain and ytrain) ％ btrain: offest coefficient (from svcm_train on xtrain and ytrain) ％ ％ ypred: predicted y (Lrun,1) containing class labels (-1 or +1) ％ margin: (signed) separation from the separating hyperplane (Lrun,1(support vector machine classification% s oft margin% uses "kernel.m"%% xtrain : (Ltrain, N) with Ltrain : number of points N : % ytrain dimension : (Ltrain. 1) containing class labels (-1 or 1)% xrun : (Lrun, N) with Lrun : number of points N : % atrain dimension : alpha coefficients (from svcm_train on xtrain and ytrain)% btrain : offest coefficient (from svcm_train on xtrain and ytrain)%% ypred : predicted y (Lrun. 1) containing class labels (-1 or 1)% margin : (signed) separation from the separating hyper plane (Lrun, 1)

intra prediction image compression using nine mode prediction.

自适应滤波算法（最小均方误差算法）matlab源码程序。(Adaptive filtering algorithm matlab source program。)

fit_ML_normal - Maximum Likelihood fit of the log-normal distribution of i.i.d. samples!. Given the samples of a log-normal distribution, the PDF parameter is found fits data to the probability of the form: p(x) = sqrt(1/(2*pi))/(s*x)*exp(- (log(x-m)^2)/(2*s^2)) with parameters: m,s format: result = fit_ML_log_normal( x,hAx ) input: x - vector, samples with log-normal distribution to be parameterized hAx - handle of an axis, on which the fitted distribution is plotted if h is given empty, a figure is created. output: result - structure with the fields m,s - fitted parameters CRB_m,CRB_s - Cram?r-Rao Bound for the estimator value RMS - RMS error of the estimation type - ML

OFDM implementation using 64QAM

用矩阵的奇异值分解法确定ARMA模型的阶次，估计模型参数(ARMA model to determine the matrix of the singular value decomposition of the order, the estimated model parameters)

ADSB广播式自动相关监视信号的生成，采样频率80MHZ(Generate ADSB automatic dependent surveillance-broadcast signals, the sampling frequency 80MHZ)