DSP Lecture 16: FIR filter design using least-squares скачать в хорошем качестве

DSP Lecture 16: FIR filter design using least-squares

ECSE-4530 Digital Signal Processing Rich Radke, Rensselaer Polytechnic Institute Lecture 16: FIR filter design using least-squares (10/27/14) 0:00:08 Introduction to filter design 0:01:14 Why can't we use an ideal low-pass filter? 0:04:46 Pure delays and linear phase 0:08:08 An ideal low-pass filter with linear phase 0:09:59 Filter terminology: group delay, magnitude response, amplitude response, angle, argument 0:15:28 The filter design process 0:19:21 Finite impulse response (FIR) and infinite impulse response (IIR) filters 0:21:19 Cost functions for filter design 0:22:21 Linear phase FIR filters 0:24:22 The form of the frequency response for an FIR filter 0:28:20 When is the frequency response linear phase? 0:31:21 Symmetries implied by linear phase 0:34:21 Type I, II, II, and IV linear phase FIR filters 0:37:37 Completed chart of the 4 filter types 0:40:01 Filter design terminology (passband, stopband, transition band, don't care) 0:43:39 Possible cost functions for filter design 0:47:35 Least-squares design techniques for length N, Type I FIR filters 0:47:38 Frequency sampling with N equally-spaced samples 0:49:54 Obtaining the answer with the inverse DFT 0:52:44 Matlab example with N=15 1:00:27 Extensions of the idea 1:02:32 Frequency sampling with L greater than N equally-spaced samples 1:04:39 Result: design a longer filter and truncate 1:08:13 Matlab example 1:12:57 Matrix formulation for designing filters with non-equally-spaced samples (e.g., to allow don't care regions) Follows Sections 5.4.1-5.4.2, 10.1-10.2 of the textbook (Proakis and Manolakis, 4th ed.).