Bài giảng Digital Signal Processing - Introduction - Nguyễn Thanh Tuấn
Classification of Signals
Signals which are generated by multiple sources or multiple sensors
can be represented in a vector form. Such a vector of signals is
referred to as a multichannel signals
Ex: 3-lead and 12-lead electrocardiograms (ECG) are often used in practice,
which results in 3-channel and 12-channel signals.
A signal is called M-dimensional if its value is a function of M
independent variable
Picture: the intensity or brightness I(x,y) at each point is a function of 2
independent variables
TV picture is 3-dimensional signal I(x,y,t)
Continuous-time versus discrete-time signal
2. Classification of Signals
Signals can be classified into four different categories depending on
the characteristics of the time variable and the values they take.
Tóm tắt nội dung tài liệu: Bài giảng Digital Signal Processing - Introduction - Nguyễn Thanh Tuấn
Click to edit Master subtitle style Nguyen Thanh Tuan, M.Eng. Department of Telecommunications (113B3) Ho Chi Minh City University of Technology Email: nttbk97@yahoo.com Introduction Chapter 0 Digital Signal Processing A signal is defined as any physical quantity that varies with time, space, or any other independent variable(s). 1. Signal and System 2 Speech, image, video and electrocardiogram signals are information-bearing signals. Mathematically, we describe a signal as a function of one or more independent variables. Examples: ( ) 110sin(2 50 )x t t 2( , ) 3 2 10I x y x xy y A system is defined as a physical device that performs any operation on a signal. A filter is used to reduce noise and interference corrupting a desired information-bearing signal. Introduction Digital Signal Processing Signal processing is to pass a signal through a system. 1. Signal and System 3 A digital system can be implemented as a combination of hardware and software (program, algorithm). Introduction Digital Signal Processing Multichannel and Multidimensional signals 2. Classification of Signals 4 Signals which are generated by multiple sources or multiple sensors can be represented in a vector form. Such a vector of signals is referred to as a multichannel signals Ex: 3-lead and 12-lead electrocardiograms (ECG) are often used in practice, which results in 3-channel and 12-channel signals. A signal is called M-dimensional if its value is a function of M independent variable Picture: the intensity or brightness I(x,y) at each point is a function of 2 independent variables TV picture is 3-dimensional signal I(x,y,t) Introduction Digital Signal Processing Continuous-time versus discrete-time signal 2. Classification of Signals 5 Signals can be classified into four different categories depending on the characteristics of the time variable and the values they take. Introduction Time Amplitude Continuous Discrete Continuous Analog signal Discrete signal Discrete Quantized signal Digital signal t x(t) n x(n) n xQ(n) 000 001 010 011 100 101 110 111 t xQ(t) Digital Signal Processing 3. Basic elements of a DSP system 6 Most of the signals encountered in science and engineering are analog in nature. To perform the processing digitally, there is a need for an interface between the analog signal and the digital processor. Fig 0.1: Analog signal processing Fig 0.2: Digital signal processing Introduction Xử lý tín hiệu số Xử lý số tín hiệu Digital Signal Processing Telephony: transmission of information in digital form via telephone lines, modem technology, mobile phone. 4. DSP applications-Communications 7 Introduction Encoding and decoding of the information sent over physical channels (to optimize transmission, to detect or correct errors in transmission) Digital Signal Processing 4. DSP applications-Radar and Sonar 8 Introduction Target detection: position and velocity estimation Tracking Digital Signal Processing Analysis of biomedical signals, diagnosis, patient monitoring, preventive health care, artificial organs. 4. DSP applications-Biomedical 9 Introduction Examples: Electrocardiogram (ECG) signal provides information about the condition of the patient’s heart. Electroencephalogram (EEG) signal provides information about the activity of the brain. Digital Signal Processing Noise reduction: reducing background noise in the sequence produced by a sensing device (a microphone). 4. DSP applications-Speech 10 Introduction Speech recognition: differentiating between various speech sounds. Synthesis of artificial speech: text to speech systems. Digital Signal Processing Content based image retrieval: browsing, searching and retrieving images from database. 4. DSP applications-Image Processing 11 Introduction Image enhancement Compression: reducing the redundancy in the image data to optimize transmission/storage Digital Signal Processing Generation, storage and transmission of sound, still images, motion pictures. 4. DSP applications-Multimedia 12 Introduction Digital TV Video conference Digital Signal Processing The Journey 13 Introduction “Learning digital signal processing is not something you accomplish; it’s a journey you take”. R.G. Lyons, Understanding Digital Signal Processing Digital Signal Processing 5. Advantages of digital over analog signal processing 14 A digital programmable system allows flexibility in reconfiguring the DSP operations simply by changing the program. A digital system provides much better control of accuracy requirements. Digital signals are easily stored. DSP methods allow for implementation of more sophisticated signal processing algorithms. Limitation: Practical limitations of DSP are the quantization errors and the speed of A/D converters and digital signal processors -> not suitable for analog signals with large bandwidths. Introduction Digital Signal Processing Course overview 15 Introduction Chapter 0: Introduction to Digital Signal Processing (3 periods) Chapter 7: Fourier transform and FFT algorithm (6 periods) Chapter 1: Sampling and Reconstruction (6 periods) Chapter 3: Analysis of linear time invariant systems (LTI) (6 periods) Chapter 4: Finite Impulse Response and convolution (3 periods) Chapter 5: Z-transform and its applications (6 periods) Chapter 6: Transfer function and filter realization (3 periods) Chapter 8: FIR and IIR filter designs (6 periods) Chapter 2: Quantization (3 periods) Review and mid-term exam: 3 periods Digital Signal Processing Text books: [1] S. J. Orfanidis, Introduction to Signal Processing, Prentice- Hall Publisher 2010. [2] J. Proakis, D. Manolakis, Digital Signal Processing, Macmillan Publishing Company, 1989. References 16 Introduction Reference books: [3] V. K. Ingle, J. Proakis, Digital Signal Processing Using Matlab, Cengage Learning, 3 Edt, 2011. Digital Signal Processing Learning outcomes 17 Introduction Understand how to convert the analog to digital signal Be able to design and implement FIR and IIR filters. Have a thorough grasp of signal processing in linear time-invariant systems. Understand the z-transform and Fourier transforms in analyzing the signal and systems. Digital Signal Processing Assessment 18 Introduction Mid-term test: 20% Homework: 20% Final exam: 60% Bonus: added to Test and Homework Test and Homework (40%) Final exam (60%) Final Mark (100%) 0.0 7.5 4.50 4.5 2.5 6.0 4.60 4.5 3.0 6.0 4.80 5.0 4.0 5.5 4.90 5.0 5.5 4.5 4.90 5.0 6.0 4.0 4.80 5.0 7.0 3.5 4.90 5.0 7.5 3.0 4.80 5.0 7.0 3.0 4.60 4.5 10.0 2.5 5.50 2.5 10.0 4.00 Absent Digital Signal Processing Assessment 19 Introduction Điểm ghi trên Bảng điểm kiểm tra, Bảng điểm thi và Bảng điểm tổng kết được làm tròn đến 0,5. (từ 0 đến dưới 0,25 làm tròn thành 0; từ 0,25 đến dưới 0,75 làm tròn thành 0,5; từ 0,75 đến dưới 1,0 làm tròn thành 1,0) Nếu điểm thi nhỏ hơn 3 và nhỏ hơn điểm tổng kết tính từ các điểm thành phẩn (kể cả điểm thi) thì lấy điểm thi làm điểm tổng kết. Digital Signal Processing Timetable 20 Introduction Time Class Monday (T1-3) DD13BK01-A02 314B1 Tuesday (T7-9) DD13KSTD 206B1 Wednesday (T10-12) DD13LT04-A04 303B1 Digital Signal Processing Review of complex number 21 Introduction cosx r siny r Rectangular form: Real part: Imaginary part: Polar form: Absolute value (modulus, magnitude): Argument (angle): cos sinie i Euler’s formula: ire z r x iy z 2 2| |r x y z 1arg( ) tan y x z Cartesian coordinates Polar coordinates Argand diagram (−π , π] Digital Signal Processing Review of periodic signals 22 Introduction Definition: x(t) = x(t + T) t Fundamental period (cycle duration): smallest T Ordinary frequency: f = 1/T (cps or Hz) --> F Radial (angular) frequency: = 2 f (rad/s) --> Digital Signal Processing Review of special functions 23 Introduction Rectangular (rect) Unnormalized: Normalized: Cardinal sine (sinc) Digital Signal Processing Review of special functions 24 Introduction Dirac delta: Properties: Digital Signal Processing Review of special functions 25 Introduction Dirac comb (impulse train, sampling function): Properties: Digital Signal Processing Review of spectral analysis 26 Introduction Periodic signal: Fourier series (line spectrum) Aperiodic signal: Fourier transform Digital Signal Processing Review of Fourier transforms 27 Introduction 0 0 0 1 cos(2 ) [ ( ) ( )] 2 FTF t F F F F 0 0 0 1 sin(2 ) [ ( ) ( )] 2 FTF t j F F F F Digital Signal Processing Review of Fourier transform properties 28 Introduction Linear (superposition): Delay: Convolution: Digital Signal Processing Review of trigonometric formulas 29 Introduction 1 cos( )cos( ) [cos( ) cos( )] 2 a b a b a b 1 sin( )sin( ) [cos( ) cos( )] 2 a b a b a b 1 sin( )cos( ) [sin( ) sin( )] 2 a b a b a b Digital Signal Processing Review of Poisson summation formula 30 Introduction Statement: Condition: Digital Signal Processing Review of convolution and correlation 31 Introduction Convolution: Correlation: Auto-correlation: Digital Signal Processing Review of analog linear time-invariant system 32 Introduction 0( ) cos(2 )x t A F t ( )x t Analog LTI system h(t) H(F) ( ) ( ) ( )y t x t h t ( )X F ( ) ( ) ( )Y F X F H F Linear: Time-invariant: Impulse response: Frequency response: Amplitude (magnitude): |H(F)| Phase: arg{H(F)} 0 0 0( ) | ( ) | cos(2 arg{ ( )})y t A H F F t H F Digital Signal Processing Review of analog filters 33 Introduction Decibel: |A|dB = 20log10|A| Logarithmic scales: Decade: decades = log10(F2/F1) Octave: octaves = log2(F2/F1) Cut-off (-3dB) frequency Bandwidth Digital Signal Processing Example of octave scale 34 Introduction An 88-key piano in twelve-tone equal temperament, with the octaves numbered and Middle C (cyan) and A440 (yellow) highlighted. C D E F G A B Digital Signal Processing Bonus 1 35 Introduction Write a program generating tones of an 88-key piano in twelve-tone equal temperament with A440 standard. Digital Signal Processing Bonus 2 36 Introduction Write a program generating tones of a guitar with standard below. Digital Signal Processing Bonus 3 37 Introduction Write a program plotting the waveform of signal below. Digital Signal Processing Bonus 4 38 Introduction Write a program plotting the spectrum of signal below. Digital Signal Processing Greek alphabet 39 Introduction Digital Signal Processing Portraits of Scientists and Inventors 40 Introduction René Descartes (1596-1650): French philosopher, mathematician and scientist. “Cogito, ergo sum” (“Tôi tư duy, vậy tôi tồn tại”). Jean-Robert Argand (1768-1822): French amateur mathematician. Jean-Baptiste Joseph Fourier (1768-1830): French mathematician and physicist. Siméon Denis Poisson (1781-1840): French mathematician, geometer, and physicist. Digital Signal Processing Portraits of Scientists and Inventors 41 Introduction Heinrich Rudolf Hertz (1857-1894) was a German physicist who first conclusively proved the existence of electromagnetic waves. Alexander Graham Bell (1847-1922) was an eminent Scottish- born scientist, inventor, engineer and innovator who is credited with inventing the first practical telephone. Digital Signal Processing Homework 1 42 Introduction For each case below, find the modulus and argument (both in radian and degree): 1) –2 2) –3i 3) –2 – 3i 4) –2 + 3i 5) 2 – 3i 6) 1/(2 – 3i) 7) (2 – 3i)/i 8) (2 – 3i)^2 9) (2 – 3i) + 1/(2 – 3i) 10) (2 – 3i).(–2 – 3i) 11) (2 – 3i)/(–2 – 3i) 12) (2 – 3i)/( 2 + 3i) Digital Signal Processing Homework 2 43 Introduction For each case below, find the modulus and argument (both in radian and degree): 1) e^(i ) 2) e^(i /2) 3) e^(–i /2) 4) e^(i /4) 5) e^(i /2) + e^(i /4) 6) 1/e^(i /4) 7) e^(i /4) / e^(–i /4) 8) e^(i /4) + e^(–i /4) 9) e^(i /4) – e^(–i /4) 10) 1 + e^(i /2) 11) 1 – e^(i /2) 12) (2 – 3i). e^(i /4) Digital Signal Processing Homework 3 44 Introduction For each case below, sketch the locus of z on the complex plane: 1) |z| = 1 2) |z – 2| = 1 3) |z – 1| = 2 4) |z – 1 – 2i| = 3 5) |z| < 3 6) |z| > 2 7) 2 < |z| < 3 8) |z -1| < 4 9) |z -1| > 2 10) 2 < |z -1| < 4 11) z + z -1 ≠ ∞ 12) 1 + z -2 ≠ ∞ Digital Signal Processing Homework 4 45 Introduction For each case below, sketch the waveform of the signal: 1) x(t) = 4sin(2t) (t:s) 2) x(t) = 4sin(2 t) (t:s) 3) x(t) = 4cos(2 t) (t:s) 4) x(t) = 4cos(10 t) (t:s) 5) x(t) = 4cos(10 t) (t:ms) 6) x(t) = 1 + 4cos(10 t) (t:s) 7) x(t) = 4cos(2 t) + 4cos(10 t) (t:s) 8) x(t) = 4sin2(2 t) (t:s) 9) x(t) = 4sinc(2t) (t:s) 10) x(t) = 4{(t – 3)/2} 11) x(t) = k{4{(t – k5 – 3)/2}} 12) x(t) = 4(t – 3) – 3(t + 4) Digital Signal Processing Homework 5 46 Introduction For each case below, plot the magnitude spectrum of the signal: 1) A 2) A.cos(2 Ft+) 3) A.cos(2 Ft+) + B 4) A.cos(2 F1t+1) + B.cos(2 F2t+2) 5) A.cos(2 Ft+1) + B.cos(2 Ft+2) 6) A.cos(2 Ft+1) + A.cos(2 Ft+2) 7) A.cos(2 Ft+) + A.sin(2 Ft+) 8) x(t) = 10 – 4cos6 t (t: ms) 9) x(t) = 1 – 2cos6 t + 3sin14 t (t: ms) 10) x(t) = 3cos103πt – 4sin104πt (t: s) 11) x(t) = 14sin23 t + 3sin14 t (t: ms) 12) x(t) = 4cos22πt – 10sin10πt (t: ms) Digital Signal Processing Homework 6 47 Introduction Suppose a filter has magnitude response as shown in figure below. Determine the expression (ignoring the phase) of the output signal and plot it’s magnitude response for each case of the input signal: 1) x(t) = 2 2) x(t) = 2cos(2 t) (t:ms) 3) x(t) = 2cos(20 t) (t:ms) 4) x(t) = 2cos(200 t) (t:ms) 5) x(t) = 2cos(400 t) (t:ms) 6) x(t) = 2cos2(400 t) (t:ms) 7) x(t) = 2cos(200 t).sin(400 t) (t:ms) 8) x(t) = 2cos(200 t) – 2cos(400 t) (t:ms) 9) x(t) = 2cos(200 t) + 2sin(400 t) (t:ms) 10) x(t) = 2cos(200 t) + 2sin(200 t) (t:ms) Digital Signal Processing Homework 7 48 Introduction Cho hệ thống tuyến tính bất biến có hàm truyền H(f) như hình: a) Xác định biểu thức đầy đủ của tín hiệu ngõ ra y(t) khi tín hiệu ngõ vào x(t) = 10cos2@πt – 30sin40πt (t:s). b) Xác định biểu thức đầy đủ của tín hiệu ngõ vào x(t) để tín hiệu ngõ ra y(t) = 10cos2@πt (t:s). Digital Signal Processing Homework 8 49 Introduction Cho các tín hiệu tương tự x1(t) = 2cos 22πt (t: s) và x2(t) = 6sin6πt + 7cos7πt + 8sin8πt (t:s) lần lượt đi qua hệ thống tuyến tính bất biến có hàm truyền H(f) như hình: a) Xác định biểu thức (theo thời gian) của tín hiệu ngõ ra y1(t). b) Tính giá trị của tín hiệu ngõ ra y2(t = 0.125s). Digital Signal Processing Homework 9 50 Introduction Tìm giá trị đáp ứng biên độ |H(f)| tại các tần số sau: a) 1KHz. b) 3KHz. c) 4KHz. d) 5KHz. e) 8KHz. Digital Signal Processing Homework 10 51 Introduction Cho bộ lọc thông thấp có đáp ứng biên độ phẳng 0dB trong khoảng [0 4]KHz, suy giảm với độ dốc 12dB/octave trong khoảng [4 8]KHz và suy giảm với độ dốc 20dB/decade ngoài 8KHz. Tìm giá trị đáp ứng biên độ của bộ lọc tại các tần số sau: a) 2KHz. b) 3KHz. c) 5KHz. d) 6KHz. e) 7KHz. f) 8KHz. g) 10KHz. h) 12KHz. i) 16KHz. j) 20KHz.
File đính kèm:
- bai_giang_digital_signal_processing_introduction_nguyen_than.pdf