自相关分析

摘 要

在实际信号处理过程中，观测信号总是混杂着干扰和噪声，对信号处理的检测与估计结果有很大影响。因此，信号处理的一个基本任务就是将混杂在噪声和干扰中的有用信号准确地检测和估计出来，而信号的可分离性是完成这个任务的关键。通常，对信号的分析与处理都是在某个特定的处理域内进行的，所以就要求信号在该处理域内具有可分离性。常用的信号处理域是时域和频域，但是在实际系统中，信号经常同时在时域和频域内混叠，使得在时域和频域内难以准确地分离出信号。因此，有必要考虑在其它处理域内来实现信号的分离。自相关域是另外一种描述信号基本特征的处理域，由于白噪声信号在自相关域内具有其独特的特性，自然地与其它非白信号具有在该域内的可分离性。因此，本文从分析信号在自相关域的描述出发，研究信号在自相关域内的可分特征，并在此基础上进行相应分析与处理，以期实现所要目标。

关键词：自相关域，信号可分离性，检测与估计，滤波

ABSTRACT

One of the fundamental tasks of signal processing is to detect and estimate the useful signal from the noise and interference,since the observed signal is always mixed with the interference and noise. And to finish this taks, the separability of signal is one of the key problems. Usually, the signal is analysised and processed in time domain or frequency domain, and it is difficult to separate the signal from the interference and noise when the signal is overlapping both in the time domain and in the frequency domain. Hence, it is necessary to find the other domain which can be used to separate the signals. The autocorrelation domain isanother signal processing domain, and the white noise is naturally can be separated from the other non-white signals due to itsunique characteristic in this domain. This paper focus on the separability of signal in autocorrelation domain, and the signal processing techniques based on the characteristic of the singal which can be used to separate the singal from the interference, noise in the aucorrelation domain. The research works of this thesis are as following:

Key ：ords:autocorrelation domain, signal separability, detection and estimation, filter