420 likes | 431 Views
This chapter provides a detailed review of signals, systems, Fourier analysis, and their importance in communication engineering research and design. Explore topics like line spectra, Fourier series, convolution, and more. Understand the essential mathematical tools for signal processing and the design of efficient communication systems.
E N D
Chapter 2. Signals and Spectra • This chapter reviews one of the two pre-requisites for communications research. • Signals and Systems • Probability, Random Variables, and Random Processes • We use linear, particularly LTI, systems to develop the theory for communications. • Outline • 2.1 Line Spectra and Fourier Series • 2.2 Fourier Transform and Continuous Spectra • 2.3 Time and Frequency Relations • 2.4 Convolution • 2.5 Impulses and Transforms in the Limit • 2.6 Discrete Time Signals and the Discrete Fourier Transform
Communication Engineering 통신공학 Step 1. Given a communication medium, we first analyze the channel and build a mathematical model. 주어진 통신 매체에 따라 Channel 을 분석하고 모형을 만든다. Step 2. Using the model, we design the pair of a transmitter and a receiver that best exploits the channel characteristic. Channel 에 가장 효과적 신호처리를 할 수 있도록 Transmitter 와 Receiver를 설계한다. ex) Modulation (변조)과 Demodulation (복조) Encoding 과 Decoding Multiplexing 과 Demultiplexing
Mathematical Tool for Signal Processing: Fourier Analysis time domain frequency domain analysis, synthesis, design • 2.1 Line Spectra and Fourier Series • Linear Time-Invariant system
대한민국 1호 라디오 (금성 A-501) 1959년, 금성사 김해수가 설계와 생산을 담당. –대한민국 역사 박물관
Line spectrum of periodic signals • 복소지수 (Complex exponential)에 의한 sinusoidal wave정현파 신호의 표현 복소수? Euler’s theorem/identity Amplitude A phase
Phasor를 이용한 정현파 신호의 표현 Phasor representation is useful when sinusoidal signal is processed by real-in real-out LTI systems. 허수축 실수축
Q1왜 frequency domain 표현이 중요한가? (여러 가지 정현파형이 선형적으로 결합된 신호)
A1 Line Spectrum “왜 Phase는 Amplitude보다 덜 중요한가? (phase time delay ) “모든 주기적 신호는 정현파 신호의 선형적 결합으로 표현될 수 있다.” Phase Amplitude 90 5 40 3 2 10 35 0 0 10 35 Frequency content
Periodic Signals (주기 신호) Rectangular pulse train Figure 2.1-7
Fourier Series 어떠한 periodic signal 정현파 신호의 선형적 집합 Where Phasor표현 two-sided line spectrum
주기함수의 주파수 특성 (Spectrum of periodic signals) 1. harmonics of fundamental frequency . 2. 3. 실함수 는
Spectrum of rectangular pulse train with ƒ0 = 1/4 (a) Amplitude (b) Phase Figure 2.1-8
Fourier-series reconstruction of a rectangular pulse train Figure 2.1-9
Fourier-series reconstruction of a rectangular pulse train Figure 2.1-9c
Gibbs phenomenon at a step discontinuity Figure 2.1-10
2.2 FourierTransforms and Continuous Spectra • Fourier Transform 비주기 신호 or Energy signal called the analysis equation. Definition
Inverse Fourier Transform called the synthesis equation.
Rectangular pulse spectrum V(ƒ) = A sinc ƒ Figure 2.2-2
Rayleigh’s Energy Theorem Generally Also called Parseval’s relation/theorem.
2.3 Time and Frequency Relations • Superposition Property • Time Delay • Time Scale Change useful tool for linear systems linear phase Slow Playback Fast Playback Low Tone High Tone
continued (a) RF pulse (b) Amplitude spectrum Figure 2.3-3
Differentiation and Integration In general Example. Triangular pulse Principle of FM demodulator differentiator
2.4 Convolution • Convolution Integral Graphical interpretation of convolution Figure 2.4-1
Result of the convolution in Fig. 2.4-1 Figure 2.4-2 In general, convolution is a complicated operation in the TD.
2.5 Impulses and Transforms in the Limit • Dirac delta function Thus
Two functions that become impulses as 0 Figure 2.5-2
Fourier Transform of Power Signals infinite energy
2.6 Discrete Time Signals and Discrete Fourier Transform • DT signal • DT periodic signal and DFTS • Analysis equation • Synthesis equation • DFT, IDFT • Periodic extension and Fourier Series • DTFT • Analysis equation • Synthesis equation
Convolution using the DFT • Q. We are given a convolution sum of two finite-length DT signals. Each signal has support N_1, N_2. Find the finite-length (at most N_1+N_2-1) output of the convolution using DFT. • A. Choose N>= N_1+N_2-1. Compute DFT(x) and DFT(h). Perform entry-by-entry multiplication. Apply the inverse DFT. Done.
HW #1 (Due on Next Tuesday 9/22. Please turn in handwritten solutions.) • 2.7 Questions • 3 • 4 • 6 • 2.1-9, 13 • 2.2-7, 10 • 2.3-8, 14 • 2.4-8, 15 • 2.5-10 • 2.6-4, 6