第四讲匹配滤波匹配滤波是从白噪声背景下检测已知信号是否存在的有效手段(当噪声的频带远宽于信号带宽时可以近似当作白噪声对待)。结合图4.1说明它的特点。设观察值x(k)由波形已知的信号s_i(k)和白噪声n_i(k)相加组成。它通过匹配滤波器后得输出y(k)=s_0(k)+n_0(k)(s_0,n_0分别是s_i,n_i引起的输出分量)。在某一时刻k=N时,s_0(N)对n_0(N)的信噪比最大,因此把这一时 Lecture 4: Matched Filtering Matched filtering is an effective means of detecting the presence or absence of a known signal from a white noise background (which can be approximated as white noise when the noise band is much wider than the signal bandwidth). Combine Figure 4.1 to illustrate its features. Let the observation value x (k) consist of adding the signal s_i (k) of known waveform and white noise n_i (k). After it passes the matched filter, it outputs y (k) = s_0 (k) + n_0 (k) (s_0, n_0 are the output components caused by s_i and n_i, respectively). At a certain moment k = N, s_0 (N) has the highest signal-to-noise ratio for n_0 (N)