. Fix-rate Signal Processing Fix rate filters - same number of input and output samples Filter x(n) 8 samples y(n) 8 samples y(n) = h(n) * x(n) Figure.

Slides:



Advertisements
Similar presentations
Digital Filters. A/DComputerD/A x(t)x[n]y[n]y(t) Example:
Advertisements

CEN352, Dr. Ghulam Muhammad King Saud University
DFT Filter Banks Steven Liddell Prof. Justin Jonas.
Qassim University College of Engineering Electrical Engineering Department Course: EE301: Signals and Systems Analysis The sampling Process Instructor:
1.The Concept and Representation of Periodic Sampling of a CT Signal 2.Analysis of Sampling in the Frequency Domain 3.The Sampling Theorem — the Nyquist.
1 Copyright © 2001, S. K. Mitra Polyphase Decomposition The Decomposition Consider an arbitrary sequence {x[n]} with a z-transform X(z) given by We can.
Multirate Digital Signal Processing
Name: Dr. Peter Tsang Room: G6505 Ext: 7763
APPLICATIONS OF FOURIER REPRESENTATIONS TO
Image (and Video) Coding and Processing Lecture 2: Basic Filtering Wade Trappe.
SAMPLING & ALIASING. OVERVIEW Periodic sampling, the process of representing a continuous signal with a sequence of discrete data values, pervades the.
The Nyquist–Shannon Sampling Theorem. Impulse Train  Impulse Train (also known as "Dirac comb") is an infinite series of delta functions with a period.
Analysis of Discrete Linear Time Invariant Systems
Chapter 4: Sampling of Continuous-Time Signals
EE513 Audio Signals and Systems Digital Signal Processing (Systems) Kevin D. Donohue Electrical and Computer Engineering University of Kentucky.
… Representation of a CT Signal Using Impulse Functions
Lecture 41 Practical sampling and reconstruction.
Discrete-Time and System (A Review)
Fourier Analysis of Systems Ch.5 Kamen and Heck. 5.1 Fourier Analysis of Continuous- Time Systems Consider a linear time-invariant continuous-time system.
DTFT And Fourier Transform
H 0 (z) x(n)x(n) o(n)o(n) M G 0 (z) M + vo(n)vo(n) yo(n)yo(n) H 1 (z) 1(n)1(n) M G 1 (z) M v1(n)v1(n) y1(n)y1(n) fo(n)fo(n) f1(n)f1(n) y(n)y(n) Figure.
Sampling Theorems. Periodic Sampling Most signals are continuous in time. Example: voice, music, images ADC and DAC is needed to convert from continuous-time.
Applications of Fourier Transform. Outline Sampling Bandwidth Energy density Power spectral density.
Z Transform When the transform is identical to DFT (11)
6.2 - The power Spectrum of a Digital PAM Signal A digtal PAM signal at the input to a communication channl scale factor (where 2d is the “Euclidean.
8.1 representation of periodic sequences:the discrete fourier series 8.2 the fourier transform of periodic signals 8.3 properties of the discrete fourier.
111 Lecture 2 Signals and Systems (II) Principles of Communications Fall 2008 NCTU EE Tzu-Hsien Sang.
1 Lab. 4 Sampling and Rate Conversion  Sampling:  The Fourier transform of an impulse train is still an impulse train.  Then, x x(t) x s (t)x(nT) *
1 Chapter 5 Ideal Filters, Sampling, and Reconstruction Sections Wed. June 26, 2013.
README Lecture notes will be animated by clicks. Each click will indicate pause for audience to observe slide. On further click, the lecturer will explain.
Professor A G Constantinides 1 The Fourier Transform & the DFT Fourier transform Take N samples of from 0 to.(N-1)T Can be estimated from these? Estimate.
Notice  HW problems for Z-transform at available on the course website  due this Friday (9/26/2014) 
Zhongguo Liu_Biomedical Engineering_Shandong Univ. Chapter 8 The Discrete Fourier Transform Zhongguo Liu Biomedical Engineering School of Control.
Unit-V DSP APPLICATIONS. UNIT V -SYLLABUS DSP APPLICATIONS Multirate signal processing: Decimation Interpolation Sampling rate conversion by a rational.
Digital Signal Processing Chapter 3 Discrete transforms.
Discrete-Time Signals and Systems
Linear filtering based on the DFT
Discrete-time Random Signals
By Dr. Rajeev Srivastava CSE, IIT(BHU)
Lecture 09b Finite Impulse Response (FIR) Filters
2. Multirate Signals.
Decimation & Interpolation (M=4) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 M M Bandwidth -  /4 Figure 12 USBLSB.
Sampling Rate Conversion by a Rational Factor, I/D
Prepared by:D K Rout DSP-Chapter 2 Prepared by  Deepak Kumar Rout.
In summary If x[n] is a finite-length sequence (n  0 only when |n|
بسم الله الرحمن الرحيم Digital Signal Processing Lecture 14 FFT-Radix-2 Decimation in Frequency And Radix -4 Algorithm University of Khartoum Department.
Prof. Brian L. Evans Dept. of Electrical and Computer Engineering The University of Texas at Austin EE445S Real-Time Digital Signal Processing Lab Spring.
Chapter 4 Discrete-Time Signals and transform
MULTI-RATE DIGITAL SIGNAL PROCESSING
Chapter4. Sampling of Continuous-Time Signals
FFT-based filtering and the
SAMPLING & ALIASING.
7.0 Sampling 7.1 The Sampling Theorem
How Signals are Sampled: ADC
Zhongguo Liu Biomedical Engineering
EE Audio Signals and Systems
Multi-Resolution Analysis
Outline Linear Shift-invariant system Linear filters
Changing the Sampling Rate
Lecture 4 Sampling & Aliasing
Reconstruction of Bandlimited Signal From Samples
DTFT from DFT samples by interpolation.
CS3291: "Interrogation Surprise" on Section /10/04
Interpolation and Pulse Shaping
Coherence spectrum (coherency squared)
Chapter 8 The Discrete Fourier Transform
Signals and Systems Revision Lecture 2
CEN352, Dr. Ghulam Muhammad King Saud University
Chapter 8 The Discrete Fourier Transform
Chapter 3 Sampling.
Presentation transcript:

. Fix-rate Signal Processing Fix rate filters - same number of input and output samples Filter x(n) 8 samples y(n) 8 samples y(n) = h(n) * x(n) Figure 1

. Multirate Signal Processing Multirate filters - different numbers of input and output samples Filter x(n) 8 samples y(n) 4 samples Figure 2

. Multirate Signal Processing Filter Decimation - M>NM samplesN samples Filter Interpolation - M<NM samplesN samples Figure 3

Decimator x(n)x(n) M y(n)y(n) Basic application - reduce bit-rate by discarding samples Consequence - Distortion and Aliasing error Figure 4

Interpolator x(n)x(n) M y(n)y(n) Basic application - Insert samples between missing gaps Consequence - Restore the number of samples before decimation Figure 4

Decimation x(n)x(n) M y(n)y(n) e.g. M = 2x(n)x(n)x’(n) y(n)y(n) Figure 5

Decimation x(n)x(n)x’(n) x’(n) = x(n) n = 0, +M, +2M, +3M, +4M,... 0 otherwise (1)

Decimation (1) i(n) is a periodic impulse train that can be expressed as (2) i(n)i(n) 0M2M-2M-M Figure 6

Decimation (1) (2) (3)

Decimation x(n)x(n) M y(n)y(n) e.g. M = 2x(n)x(n)x’(n) y(n)y(n)

Decimation x(n)x(n)x’(n)y(n)y(n) y(n) = x’(Mn)(4) According to equation (3) hence (5)

Decimation Distortion due to Decimation can be seen in the frequency domain Consider the z transform of x(n) and x’(n) (6) According to equation (3)

Decimation According to equation (3) (7)

Decimation According to equation (4),y(n) = x’(Mn) (8)where p = Mn (9)

Key equations of Decimation x(n)x(n)x’(n)y(n)y(n) y(n) = x’(Mn)

Key equations of Decimation x(n)x(n)x’(n)y(n)y(n) Convert to DFT with z = e j  z transform

Spectral Changes in Decimation x(n)x(n)x’(n)y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  Figure 7a

Spectral Changes in Decimation x(n)x(n)x’(n)y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  images M=4 Figure 7a Figure 7b

Spectral Changes in Decimation x(n)x(n)x’(n)y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 Figure 7a Figure 7b Figure 7c

Spectral Changes in Decimation x(n)x(n)x’(n)y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 Figure 7a Figure 7b Figure 7c

Spectral Changes in Decimation x(n)x(n)x’(n) 1. Retaining one out of M samples in x(n) generates M replicated images of the original spectrum. 2. The spacing of images is 2  /M 3. Decimation by a factor of M stretches the width of the spectrum by M times x’(n)y(n)y(n)

Interpolation x(n)x(n) M y(n)y(n) e.g. M = 2x’(n) y(n)y(n) e.g. M = 3x’(n) y(n)y(n) Figure 8

Interpolation x(n)x(n) M y(n)y(n) (10) y(n) = x(n/M) n = 0, +M, +2M, +3M,... 0 otherwise (11) (12)

Spectral Changes in Interpolation x(n)x(n) M y(n)y(n) (13)

Spectral Changes in Interpolation 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  Figure 9a x(n)x(n) M y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 Figure 9b

Spectral Changes in Interpolation 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  Figure 9a x(n)x(n) M y(n)y(n) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 Figure 9b

Spectral Changes in Interpolation x(n)x(n)y(n)y(n) 2. M images separating from each other by a spacing of 2  /M are generated 1. Interpolation by a factor of M compresses the width of the spectrum by M times

Decimation & Interpolation (M=4) M M Input Sequence Output Sequence Figure

Decimation & Interpolation (M=4) 0  /4  /4  /2  /4  /4  /2  /4  /4  /2  /4  M=4 M M Bandwidth -  /8 Figure 11

Decimation & Interpolation It seems that the bitrate can be reduced simply by decimation

Decimation & Interpolation But something is wrong,what’s the problem?