FoxLi

psk调制解调及相应误码率计算。随机产生一正弦信号并进行相应的调制解调。(psk modem and the corresponding bit error rate. Randomly generate a sinusoidal signal and the corresponding modem.)

fbp 三维重建 基于压缩感知的计算机成像(san wei c jian)

BP神经网络控制的一个源程序，实现对非线性时变系统的PID自整定控制(BP neural network control of a source, to achieve self-tuning PID control for nonlinear time-varying systems)

本程序针对高通公司开发的IS-95通信系统，是一个相对完整的IS-95前向信道的仿真源程序，包括Viterbi编码，数据调制，基于RAKE接收机的解调，实现软判决输入的Viterbi译码等模块，并包括有误码率的计算和性能仿真显示。 (The procedure developed for Qualcomm IS-95 communication system, is a relatively complete IS-95 forward channel simulation source code, including Viterbi coding, data modulation, demodulation based RAKE receiver to achieve soft decision input Viterbi decoding modules, and includes a bit error rate performance computing and simulation show.)

自动化程序生成包，对于进行的操作编码和分析操作指令，生成可以重复使用的程序代码(Automatic program generation package for the operation code and the analysis operation instruction to generate the program code can be reused)

聚焦波束形成 线阵波束形成算法实现：包括非FFT算法和采用FFT算法的程序(Focused beamforming line array beamforming algorithm: including non-FFT algorithm and uses FFT algorithm procedures)

Dominant-set聚类算法，主要用于图像分割，检索等(Dominant-set clustering algorithm, mainly for image segmentation, search)

LMS算法去除加噪声音乐程序，效果比较好，有图为证。(LMS algorithm to remove noise adding music program, the effect is better, has a license to the graph. )

说明：  倒立摆经典控制方法LQR PID介绍和matlab仿真(lqr pid)

说明：  基于压缩感知思想的图像分块压缩与重构方法 考虑到大多数图像信号信息分布有差异, 编码端, 在对图像分块的基础上, 融合熵估计 和边缘检测方法计算各图像块的信息含量, 再从两个不同的角度进行分类采样: 依据信息量多少将图像块分为平滑、过渡和纹理3 类, 使用不同的采样率采样 依据信息量的分布特征, 采用不同的采样率分配策略进行采样. 在解码端, 根据不同类型的图像块构造不同的线性算子进行重构, 再运用改 进的迭代阈值算法去除块效应和噪声. 实验证明, 算法在提升图像重构质量的同时缩短了重构时间,并且对纹理边缘多的图像的重构效果较其他方法理想.(Considering that most information of the image signal difference distribution, the encoding side, in the right image sub-blocks based on the integration entropy estimation And edge detection method for calculating the information content of each image block, and two different angles classification sampling: The number of image blocks based on the amount of information into a smooth, transition and texture three classes, using different sampling rates based on the amount of information distribution characteristics, using different sampling sampling rate allocation strategy. in the decoding side, depending on the type of image blocks to construct different linear operator reconstructed, re-use change Iterative thresholding algorithm into blockiness and noise removal experiments show image reconstruction algorithm to enhance the quality while reducing reconfiguration time, and much of the texture edge image reconstruction results are satisfactory compared to other methods.)