LMSadaptivefiltering北京理工大学统计信号处理.doc

统计信号处理（全英文） 结 课 报 告 题目： Least-Mean-Square Adaptive Filters and it’s Application 姓 名：-------- 任课教师： Contents 1 Introduction 2 2 Overview and Structure of Operation of LMS Algorithm 4 2.1 Structure of Operation of LMS Algorithm 4 2.2 Operation Procedures 5 3 Derivation of LMS Algorithm 6 3.1 The Idea of the Steepest-Descent Algorithm 6 3.2 Least-Mean-Square Adaptation Algorithm 9 3.3 Summary of the LMS Algorithm 11 4 The application of the LMS Algorithm 12 4.1 Instantaneous Frequency Measurement 12 4.2 Adaptive Deconvolution for Processing of Time-Varying 14 References 17  Introduction Along with the rapid development of the mobile communications industry, the scope of application of adaptive filtering techniques is also growing. Early in the 1940s, it was established on the stationary random signal Wiener filtering theory. According to the statistical properties of the useful signal and interference noise (the autocorrelation function or power spectrum), the optimum filter with a linear minimum mean square error estimation criterion design, called Wiener filter. This filter can maximize filter out interference noise and extract useful signal. However, when the statistical characteristics of the input signal deviates from the design condition, it is not the best, which is limited in practical applications. By the early 1960s, due to the development of space technology, the emergence of Kalman filtering theory, namely the use of state-variable model of non-stationary random sequence of multi-input multi-output for optimal estimation. Now, the Kalman filter has been successfully applied to many areas, it both on the stationary and non-stationary random signal as a linear optimal filtering, but also for non-linear filtering. In essence, the Wiener filter is a special case of the Kalman filter. In the design of the Kalman filter, you must know the state equation and measurement equation produces the input process of the system, whi