Download presentation

Presentation is loading. Please wait.

1
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Electrical Communications Systems ECE.09.331 Spring 2007 Shreekanth Mandayam ECE Department Rowan University http://engineering.rowan.edu/~shreek/spring07/ecomms/ Lecture 2b January 24, 2007

2
S. Mandayam/ ECOMMS/ECE Dept./Rowan UniversityPlan CFT’s (spectra) of common waveforms Impulse Sinusoid Rectangular Pulse CFT’s for periodic waveforms Sampling Time-limited and Band-limited waveforms Nyquist Sampling Impulse Sampling Dimensionality Theorem Discrete Fourier Transform (DFT) Fast Fourier Transform (FFT)

3
S. Mandayam/ ECOMMS/ECE Dept./Rowan University ECOMMS: Topics

4
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Fourier Series Exponential Representation Periodic Waveform w(t) t |W(n)| f -3f 0 -2f 0 -f 0 f 0 2f 0 3f 0 2-Sided Amplitude Spectrum f 0 = 1/T 0 ; T 0 = period T0T0

5
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Fourier Transform Fourier Series of periodic signals finite amplitudes spectral components separated by discrete frequency intervals of f 0 = 1/T 0 We want a spectral representation for aperiodic signals Model an aperiodic signal as a periodic signal with T 0 ----> infinity Then, f 0 -----> 0 The spectrum is continuous!

6
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Continuous Fourier Transform We want a spectral representation for aperiodic signals Model an aperiodic signal as a periodic signal with T 0 ----> infinity Then, f 0 -----> 0 The spectrum is continuous! t T 0 Infinity w(t) Aperiodic Waveform |W(f)| f f 0 0

7
S. Mandayam/ ECOMMS/ECE Dept./Rowan UniversityDefinitions Continuous Fourier Transform (CFT) Frequency, [Hz] Amplitude Spectrum Phase Spectrum Inverse Fourier Transform (IFT) See p. 45 Dirichlet Conditions

8
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Properties of FT’s If w(t) is real, then W(f) = W*(f) If W(f) is real, then w(t) is even If W(f) is imaginary, then w(t) is odd Linearity Time delay Scaling Duality See p. 50 FT Theorems

9
S. Mandayam/ ECOMMS/ECE Dept./Rowan University CFT’s of Common Waveforms Impulse (Dirac Delta) Sinusoid Rectangular Pulse Matlab Demo: recpulse.m

10
S. Mandayam/ ECOMMS/ECE Dept./Rowan University CFT for Periodic Signals Recall: CFT: Aperiodic Signals FS: Periodic Signals We want to get the CFT for a periodic signal What is ?

11
S. Mandayam/ ECOMMS/ECE Dept./Rowan University CFT for Periodic Signals Sine Wave w(t) = A sin (2 f 0 t) Square Wave A -A T 0 /2 T 0 Instrument Demo

12
S. Mandayam/ ECOMMS/ECE Dept./Rowan UniversitySampling Time-limited waveform w(t) = 0; |t| > T Band-limited waveform W(f)= F {(w(t)}=0; |f| > B -T T w(t) t -B B W(f) f Can a waveform be both time-limited and band-limited?

13
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Nyquist Sampling Theorem Any physical waveform can be represented by where If w ( t ) is band-limited to B Hz and

14
S. Mandayam/ ECOMMS/ECE Dept./Rowan University What does this mean? 1/f s 2/f s 3/f s 4/f s 5/f s w(t) t a 3 = w(3/f s ) If then we can reconstruct w(t) without error by summing weighted, delayed sinc pulses weight = w(n/f s ) delay = n/f s We need to store only “samples” of w(t), i.e., w(n/f s ) The sinc pulses can be generated as needed (How?) Matlab Demo: sampling.m

15
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Impulse Sampling How do we mathematically represent a sampled waveform in the Time Domain? Frequency Domain?

16
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Sampling: Spectral Effect w(t) t w s (t ) t f -B 0 B |W(f)| f |W s (f) | -2f s -f s 0 f s 2 f s (-f s -B) -(f s +B) -B B (f s -B) (f s +B) F F Original Sampled

17
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Spectral Effect of Sampling Spectrum of a “sampled” waveform Spectrum of the “original” waveform replicated every f s Hz =

18
S. Mandayam/ ECOMMS/ECE Dept./Rowan UniversityAliasing If f s < 2B, the waveform is “undersampled” “aliasing” or “spectral folding” How can we avoid aliasing? Increase f s “Pre-filter” the signal so that it is bandlimited to 2B < f s

19
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Dimensionality Theorem A real waveform can be completely specified by N = 2BT 0 independent pieces of information over a time interval T 0 N: Dimension of the waveform B: Bandwidth BT 0 : Time-Bandwidth Product Memory calculation for storing the waveform f s >= 2B At least N numbers must be stored over the time interval T0 = n/f s

20
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Discrete Fourier Transform (DFT) Discrete Domains Discrete Time: k = 0, 1, 2, 3, …………, N-1 Discrete Frequency:n = 0, 1, 2, 3, …………, N-1 Discrete Fourier Transform Inverse DFT Equal time intervals Equal frequency intervals n = 0, 1, 2,….., N-1 k = 0, 1, 2,….., N-1

21
S. Mandayam/ ECOMMS/ECE Dept./Rowan University Importance of the DFT Allows time domain / spectral domain transformations using discrete arithmetic operations Computational Complexity Raw DFT: N 2 complex operations (= 2N 2 real operations) Fast Fourier Transform (FFT): N log 2 N real operations Fast Fourier Transform (FFT) Cooley and Tukey (1965), ‘Butterfly Algorithm”, exploits the periodicity and symmetry of e -j2 kn/N VLSI implementations: FFT chips Modern DSP

22
S. Mandayam/ ECOMMS/ECE Dept./Rowan University How to get the frequency axis in the DFT The DFT operation just converts one set of number, x[k] into another set of numbers X[n] - there is no explicit definition of time or frequency How can we relate the DFT to the CFT and obtain spectral amplitudes for discrete frequencies? (N-point FFT) n=0 1 2 3 4 n=N f=0 f = f s Need to know f s

23
S. Mandayam/ ECOMMS/ECE Dept./Rowan UniversitySummary

Similar presentations

© 2020 SlidePlayer.com Inc.

All rights reserved.

To make this website work, we log user data and share it with processors. To use this website, you must agree to our Privacy Policy, including cookie policy.

Ads by Google