saolei

自己写的 归一化函数 平台 matlab(Write your own platform for matlab normalized function)

现代通信系统第二版的源代码（非常全面）学习通信编程的必备。(The source code for the second edition of the modern communication systems)

pattern recognition part two toolbox for matlab

非线性对象的线性化，cstr模型的建立，利用非线性求解方法实现其线性化(nonlinear translate to linear，model of cstr)

平面共轭分度凸轮机构设计计算和运动分析计算，利用matlab软件来计算凸轮主要几何尺寸。( U5E73 u7279 u8B1 u8B1 u8B09 u5A U4EF6 u5F3 u8F3 )

利用凸优化方法求解估计信道的预编码，具体需要参考stphen boy 的凸优化工具包(solve the precoder by convex method)

完成DSSS调制解调功能 mode[1，2] 1进行调制，2进行解调，未指定时自动完成这两个功能(Completion of DSSS modulation and demodulation functionality mode [1,2] 1 modulation, 2 to carry out demodulation, when the auto-complete does not specify that these two functions)

最完整的OFDM matlab程序，适合OFDM入门者。(the whole OFDM matlab program code, which is suitable for the rudimental user.)

提出了一种求解电磁场有限元 边界元混合法所生成的线性方程组的有效方法 ———内观 法结合多波前法.由于该线性方程组的系数是一个部分稀疏部分满填充的矩阵,为了加速求解,应用内观法将系数矩阵分为 2块,一块是有限元法形成的稀疏矩阵,另一块是边界元法生成的满阵,然后用多波前法求解稀疏矩阵方程,用高斯 约当消去法解满阵方程. 采用该方法,计算了二维多层介质柱体的雷达散射截面.计算结果表明,该方法的计算效率远远高于传统的高斯法.(Proposed for solving mixed finite element boundary element linear equations generated by an effective method of Vipassana--- a combination of multi wave front method. Since the coefficients of linear equations is a part of the sparsely populated part of the full matrix, in order to speed up the solution, application of Vipassana coefficient matrix is ​ ​ divided into two, one is the finite element method for the formation of the sparse matrix, the other is the boundary element method generates the full array, and then use the multifrontal method for solving sparse matrix equation Gauss Jordan elimination method with full matrix equation solution. use this method to calculate the two-dimensional multi-layer dielectric cylinder of the radar cross section. The results show the computational efficiency of this method is much higher than the traditional Gaussian law.)

em algorithm - Find approximate solution to Sf = conv(s,f) = d using EM iteration. EM seeks to minimize the Poisson negative log likelihood function J(f) = sum_i {[Sf]_i - (d_i + sigma^2)*log([Sf]_i + sigma^2)}.