Nachrichtentechnik I WS 2005/2006

Σχετικά έγγραφα
Probability and Random Processes (Part II)

Fundamentals of Signals, Systems and Filtering

Linear System Response to Random Inputs. M. Sami Fadali Professor of Electrical Engineering University of Nevada

Ψηφιακή Επεξεργασία Φωνής

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 4, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 3

CT Correlation (2B) Young Won Lim 8/15/14

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 3, Νέα Κτίρια ΣΗΜΜΥ Ανάλυση Επικοινωνιακών Σημάτων κατά Fourier

ΗΜΥ 220: ΣΗΜΑΤΑ ΚΑΙ ΣΥΣΤΗΜΑΤΑ Ι Ακαδημαϊκό έτος Εαρινό Εξάμηνο Κατ οίκον εργασία αρ. 2

LTI Systems (1A) Young Won Lim 3/21/15

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 4, Νέα Κτίρια ΣΗΜΜΥ Ανάλυση Επικοινωνιακών Σημάτων κατά Fourier

The Spiral of Theodorus, Numerical Analysis, and Special Functions

Introduction to Time Series Analysis. Lecture 16.

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

Stationary Stochastic Processes Table of Formulas, 2017

3 Frequency Domain Representation of Continuous Signals and Systems

Contents Introduction to Filter Concepts All-Pole Approximations

Main source: "Discrete-time systems and computer control" by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1

Biostatistics for Health Sciences Review Sheet

BandPass (4A) Young Won Lim 1/11/14

Statistics 104: Quantitative Methods for Economics Formula and Theorem Review

6.003: Signals and Systems. Modulation

w o = R 1 p. (1) R = p =. = 1

Elements of Information Theory

Math221: HW# 1 solutions

Ανάλυση Θορύβου Σε Γραμμικά Κυκλώματα

Αθανάσιος Σκόδρας /

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Stationary Stochastic Processes Table of Formulas, 2016

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 4, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 4

5.4 The Poisson Distribution.

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 3, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 3

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 3, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 5

Problem Set 9 Solutions. θ + 1. θ 2 + cotθ ( ) sinθ e iφ is an eigenfunction of the ˆ L 2 operator. / θ 2. φ 2. sin 2 θ φ 2. ( ) = e iφ. = e iφ cosθ.

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Aquinas College. Edexcel Mathematical formulae and statistics tables DO NOT WRITE ON THIS BOOKLET

ECE 308 SIGNALS AND SYSTEMS FALL 2017 Answers to selected problems on prior years examinations

Spherical Coordinates

DiracDelta. Notations. Primary definition. Specific values. General characteristics. Traditional name. Traditional notation

Second Order RLC Filters

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

HMY 799 1: Αναγνώριση Συστημάτων

ECE 468: Digital Image Processing. Lecture 8

Example Sheet 3 Solutions

4.6 Autoregressive Moving Average Model ARMA(1,1)

Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3

Ψηφιακές Επικοινωνίες

Parametrized Surfaces

ΣΗΜΑΤΑ ΔΙΑΚΡΙΤΟΥ ΧΡΟΝΟΥ

Other Test Constructions: Likelihood Ratio & Bayes Tests

Survival Analysis: One-Sample Problem /Two-Sample Problem/Regression. Lu Tian and Richard Olshen Stanford University

ΜΕΛΕΤΗ ΚΑΙ ΠΡΟΣΟΜΟΙΩΣΗ ΙΑΜΟΡΦΩΣΕΩΝ ΣΕ ΨΗΦΙΑΚΑ ΣΥΣΤΗΜΑΤΑ ΕΠΙΚΟΙΝΩΝΙΩΝ.

Fundamentals of Wireless Communication

Assignment 1 Solutions Complex Sinusoids

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

Sampling Basics (1B) Young Won Lim 9/21/13

HW 3 Solutions 1. a) I use the auto.arima R function to search over models using AIC and decide on an ARMA(3,1)

2 Composition. Invertible Mappings

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

ECE Spring Prof. David R. Jackson ECE Dept. Notes 2

University of Illinois at Urbana-Champaign ECE 310: Digital Signal Processing

Notations. Primary definition. Specific values. General characteristics. Series representations. Traditional name. Traditional notation

HMY 429: Εισαγωγή στην Επεξεργασία Ψηφιακών. Χρόνου (Ι)

Spectrum Representation (5A) Young Won Lim 11/3/16

Outline Analog Communications. Lecture 05 Angle Modulation. Instantaneous Frequency and Frequency Deviation. Angle Modulation. Pierluigi SALVO ROSSI

Πανεπιστήµιο Κύπρου Πολυτεχνική Σχολή

ExpIntegralE. Notations. Primary definition. Specific values. Traditional name. Traditional notation. Mathematica StandardForm notation

ST5224: Advanced Statistical Theory II

Practice Exam 2. Conceptual Questions. 1. State a Basic identity and then verify it. (a) Identity: Solution: One identity is csc(θ) = 1

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

ΔΙΑΚΡΙΤΟΣ ΜΕΤΑΣΧΗΜΑΤΙΣΜΟΣ FOURIER - Discrete Fourier Transform - DFT -

Reminders: linear functions

Notations. Primary definition. Specific values. General characteristics. Traditional name. Traditional notation. Mathematica StandardForm notation

Space-Time Symmetries

Appendix A. Curvilinear coordinates. A.1 Lamé coefficients. Consider set of equations. ξ i = ξ i (x 1,x 2,x 3 ), i = 1,2,3

SCHOOL OF MATHEMATICAL SCIENCES G11LMA Linear Mathematics Examination Solutions

CHAPTER 101 FOURIER SERIES FOR PERIODIC FUNCTIONS OF PERIOD

ΣΤΗΑ ΨΕΣ /3/ :03 µµ

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

Durbin-Levinson recursive method

ΣΤΟΧΑΣΤΙΚΑ ΣΥΣΤΗΜΑΤΑ & ΕΠΙΚΟΙΝΩΝΙΕΣ 1o Τμήμα (Α - Κ): Αμφιθέατρο 4, Νέα Κτίρια ΣΗΜΜΥ Θεωρία Πιθανοτήτων & Στοχαστικές Ανελίξεις - 2

Introduction to the ML Estimation of ARMA processes

Solution Series 9. i=1 x i and i=1 x i.

HMY 220: Σήματα και Συστήματα Ι

An Automatic Modulation Classifier using a Frequency Discriminator for Intelligent Software Defined Radio

SOLUTIONS TO MATH38181 EXTREME VALUES AND FINANCIAL RISK EXAM

ME 374, System Dynamics Analysis and Design Homework 9: Solution (June 9, 2008) by Jason Frye



6.1. Dirac Equation. Hamiltonian. Dirac Eq.

Takeaki Yamazaki (Toyo Univ.) 山崎丈明 ( 東洋大学 ) Oct. 24, RIMS

What happens when two or more waves overlap in a certain region of space at the same time?

DERIVATION OF MILES EQUATION FOR AN APPLIED FORCE Revision C

Αναγνώριση Προτύπων. Baysian Θεωρία Αποφάσεων (DETECTION)

Εισαγωγή. Διάλεξη 1. Εισαγωγή Σήματα και Συστήματα Διακριτού Χρόνου. Τι είναι σήμα; Παραδείγματα

Module 5. February 14, h 0min

ΕΙΣΑΓΩΓΗ ΣΤΗ ΣΤΑΤΙΣΤΙΚΗ ΑΝΑΛΥΣΗ

2.153 Adaptive Control Lecture 7 Adaptive PID Control

HMY 799 1: Αναγνώριση Συστημάτων

Ψηφιακή Επεξεργασία Φωνής

Transcript:

Nachrichtentechnik I WS 2005/2006 1 Signals & Systems wt 10/2005 1

Overview (Signals & Systems) Signals: definition & classification properties basic signals Signal transformations Fourier transformation signal space representation Stochastic signals properties white Gaussian noise Systems definition & classification linear time-invariant systems wt 10/2005 2

Motivation Signal: Physical means to represent information Necessary for information transmission Here: mathematical concept required to model communications System: Mathematical model for the transmission medium which transports information Examples: wire, optical fiber, radio transmission, filter Results in a transformation of the input signal Input Signal s(t) System g(t) Output Signal Transformation wt 10/2005 3

Overview (Signals & Systems) Signals: definition & classification properties basic signals Signal transformations Fourier transformation signal space representation Stochastic signals properties white Gaussian noise Systems definition & classification linear time-invariant systems wt 10/2005 4

Signals Definition (cf Lüke: Signalübertragung ) Change in a physical quantity in order to attract attention and to transfer meaning Examples: Pressure fluctuations Ink distribution on a white paper Brightness distribution on a screen Voltage / current variations Mathematical description (time signal): s(t) range (Wertebereich) domain (Definitionsbereich) wt 10/2005 5

Signal Classification (1) range domain real-valued real-valued / complex-valued continuous-time & continuous-range integers continuous-time & discrete-range analog signal integers discrete-time & continuous-range discrete-time & discrete-range digital signal wt 10/2005 6

k s(t) k+1 s(t) k+2 s(t) Signal Classification (2) Deterministic signals 6 Described by a closed mathematical expression, eg s( t) = cos(2πf 0t) - t Stochastic signals 6 Characterized by a stochastic process: - Probability distribution function / cumulative distribution function - Power density spectrum Represented by sample functions: Examples: speech signal, noise, 6 - t - t wt 10/2005 7

Signal Properties (1) Scalar product: Orthogonality: Cross-correlation function (CCF): (s(t),g(t)) = Z s(t) g (t) dt (s(t),g(t)) = 0 ϕ E sg(τ) =(s(t),g(t + τ)) Z = s (t) g (t + τ) dt Autocorrelation function (ACF): ϕ E ss(τ) = Z s(t) s (t + τ) ) dt wt 10/2005 8

Signal Properties (2) Distance between signals (norm): d = s(t) g(t) = p (s(t) g(t),s(t) g(t)) Signal energy: E s = s(t) 2 =(s(t),s(t)) = ϕ E ss(0) = Z s(t) 2 dt Mean signal power: P s = lim T 1 2T T T s( t) 2 dt Energy signals: Es < and Ps = 0 Power signals: Es = and Ps < wt 10/2005 9

Basic signals (1) Rectangular impulse: rect(t) = ½ 1 0 for otherwise t 1 2 Triangular impulse: Λ(t) = ½ 1 t for t 1 0 otherwise Sine x over x: si(πt) = sin(πt) πt Gaussian impulse: 1 t e 2π 2 2 wt 10/2005 10

Basic signals (2) Step function: ε(t) = ½ 1 0 for t 0 otherwise Dirac delta function δ(t): Z δ(τ)dτ =1 Z s(t τ)δ(τ)dτ = s(t) Sha function: III(t) = X i δ(t i) wt 10/2005 11

Overview (Signals & Systems) Signals: definition & classification properties basic signals Signal transformations Fourier transformation signal space representation Stochastic signals properties white Gaussian noise Systems definition & classification linear time-invariant systems wt 10/2005 12

Signal Transformations Time reflection: even signal: odd signal: Scaling: Time shift (delay element): Fourier transform: s( t) s ( t) = s( t) s( t) = s( t) s(at) s( t t0) S(f)= Z s(t)= Z s(t)e j2πft dt S(f)e j2πft df wt 10/2005 13

Properties of the Fourier Transform Linearity Convolution Multiplication Modulation Time Shift Similarity Similarity & Shift Differentiation t-multiplication Complex Conjugate Parseval s Theorem Time Domain Frequency Domain c 1s1( t) + c2s2( t) c 1S1( f ) + c2s2( f ) s1( t) s2( t) S1( f ) S2( f ) S f ) S ( s1( t) s2( t) 1( 2 f ) j2πf t S( f f0) 0 s( t) e s ( t t ) j2πft0 0 S( f ) e 1 f s (at) S( a ) a s a ( t t 1 f j2πft0 )) S( a ) e ( 0 d n n s(t) ( j2πf ) n S( f ) dt n n j2πt ) s( ) d S( f ) ( t s * ( ± t) S *( f ) x ( t), s( t) = X ( f ), S( f ) a df n wt 10/2005 14

Symmetry Properties of the Fourier Transform Time Domain s(t) Frequency Domain S(f) real even real odd imaginary even imaginary odd real even imaginary odd imaginary even real odd wt 10/2005 15

Basic Signals and their Spectra (1) Basic waveform Spectrum Informationstechnik Universität Ulm wt 10/2005 16

Basic Signals and their Spectra (2) Basic waveform Spectrum wt 10/2005 17

Signal Space Representation Coordinates in signal space: a i =(s(t),b i (t)) i =0, 1,,M 1 = Z s(t) b i(t) dt Linear combination: s(t) = M 1 X i=0 a i b i (t) Signal vectors: a s =(a 0,a 1,,a M 1 ) wt 10/2005 18

Overview (Signals & Systems) Signals: definition & classification properties basic signals Signal transformations Fourier transformation signal space representation Stochastic signals properties white Gaussian noise Systems definition & classification linear time-invariant systems wt 10/2005 19

Stochastic Signals: Properties (1) Linear ensemble mean: m s (t 0 )= 1 s(t 0 )= lim n n nx k= 1 k s(t 0 ) Ensemble mean square: P = s 2 (t 0 )= n 1 lim n nx k= 1 k s 2 (t 0 ) Variance: σs(t 2 0 )= s(t0 ) m s (t 0 ) 2 ensemble average s 2 (t 0 )=σ 2 s(t 0 )+m 2 s (t 0 ) wt 10/2005 20

Stochastic Signals: Properties (2) time average Linear time mean: k m = m t (k 0 ) = k s(t) = lim T Z 1 T 2 T T k s (t) dt Time mean square: Z T k 2 1 s (t) = lim k 2 s (t) T 2 T T dt Stationarity: ensemble averages are time invariant Ergodicity: time average and ensemble average coincide ergodicity stationarity wt 10/2005 21

Stochastic Signals: Properties (3) Autocorrelation function: ϕ ss (t 0,τ)= s(t 0 ) s(t 0 + τ) ϕ ss (t 0, 0) = s(t 0 ) s(t 0 )= s 2 (t 0 ) Autocovariance function: μ ss (t 0,τ)= s(t0 ) m s (t 0 ) s(t 0 + τ) m s (t 0 +τ ) Power density spectrum: Φ ss (t 0,f)= Z ϕ ss (t 0,τ) e j2πfτ dτ Φ ss ( 0 t0, f ) ϕss( t, τ ) wt 10/2005 22

Stochastic Signals: Properties (4) Crosscorrelation function: ϕ sg (t 0,τ)= s(t 0 ) g (t 0 + τ) Crosscovariance function: μ sg (t 0,τ)= s(t0 ) m s (t 0 ) g(t 0 + τ) m g (t 0 + τ) wt 10/2005 23

Stochastic Signals: Properties (5) Cumulative distribution function (cdf) (Verteilungsfunktion): P s (x, t 0 )= k s(t 0 ) x Prob 1 P s (x 1,t 0 ) P s (x 2,t 0 ) 1 P s (,t 0 ) = 1 P s (,t 0 ) = 0 for x 1 x 2 Probability density function (pdf) (Verteilungsdichtefunktion): p s (x, t 0 )= d dx P s(x, t 0 ) wt 10/2005 24

p e 2 2 6 Gaussian Distribution (1) Probability density function (pdf): p s (x) = 1 (x m) 2 e 2 πσ 2 2σ 2 p s (x) 6 1 p 2 2 1 x - m ; m m + Informationstechnik Universität Ulm wt 10/2005 25 P s (x)

1 e 2 2 Gaussian Distribution (2) p Cumulative distribution function (cdf): m ; m m + P s (x) = Z x p s (ϑ) dϑ - x P s (x) 6 1 1 2 mean value - x Mittelwert wt 10/2005 26

Gaussian Distribution (3) Cdf of Gaussian distribution: P s (x) = Z x = 1 2 p s (ϑ) dϑ m x erfc 2 σ 2 Error function: erf (x) = Z x 2 e ϑ 2 π dϑ 0 Error function complement: erfc(x) =1 erf (x) wt 10/2005 27

White Gaussian Noise Process Probability distribution function: Gaussian Ergodic Process Mean value: m s = 0 Variance: 2 σ s = ϕ ss (0) Power density spectrum: Φ ss (t 0,f)=Φ ss (f) =N 0 white Autocorrelation function: ϕ ss (τ) =N 0 δ(τ) wt 10/2005 28

Overview (Signals & Systems) Signals: definition & classification properties basic signals Signal transformations Fourier transformation signal space representation Stochastic signals properties white Gaussian noise Systems definition & classification linear time-invariant systems wt 10/2005 29

Systems Definition (cf Oppenheim Signals and Systems ) A system can be viewed as any process that results in the transformation of signals Input Signal s(t) System g(t) $ $ Output Signal Examples: filter wire Transformation/ Mapping wt 10/2005 30

A System Can Be Continuous-time or discrete-time Deterministic or stochastic Memory-less or with memory Causal or non-causal Stable or unstable Linear or nonlinear Time-invariant or time-variant wt 10/2005 31

Block-diagram Representation of Systems s(t) = s 1 (t) g 1 (t) = s 2 (t) g 2 (t) system 1 system 2 s(t) g(t) = g 2 (t) + g 3 (t) system 3 s(t) = s 3 (t) g 3 (t) wt 10/2005 32

Basic Systems (1): n(t) + g(t) = s(t) + n(t) (addition of two signals) s(t) g(t) a g(t) = a s(t) (multiplication by a s(t) g(t) constant coefficient) T s(t) g(t) g(t) = s(t-t) (delay element) n(t) x g(t) = s(t) n(t) (multiplication of two s(t) g(t) signals) wt 10/2005 33

Basic Systems (2): SIN g(t) = sin [s(t)] (nonlinear operation) s(t) g(t) h(t) s(t) g(t) g(t) = s(t) * h(t) (LTI-system) s(t) h 1 (t) g(t) (feedback system) h 2 (t) wt 10/2005 34

Linear Time-Invariant (LTI) Systems (1) Linear: s i(t) $ g i(t) $ X s i (t) $ X g i (t) $ i i Superposition principle Time-Invariant: Impulse response: Z s(t) $ g(t) $ s(t τ) $ g(t τ) $ δ(t) s(t τ) δ(τ) dτ h(t) Z s(t τ) h(τ) dτ Convolution integral: Z s(t τ) h(τ) dτ = s(t) h(t) wt 10/2005 35

Linear Time-Invariant (LTI) Systems (2) Input Signal s(t) $ h(t) Output Signal g(t) =s(t) h(t) $ impulse response Properties of the convolution integral: commutative law: s t) h( t ( ) = h( t) s( t) associative law: distributive law: s s ( 2 t) [ h1 ( t) h2 ( t)] = [ s( t) h1 ( t)] h ( t) ( 2 t) [ h1 ( t) + h2 ( t)] = [ s( t) h1 ( t)] + [ s( t) h ( t)] wt 10/2005 36

Some Properties of LTI Systems Transfer function: h(t) H ( f ) g( t) = s( t) h( t) G( f ) = S( f ) H ( f ) without memory: h( t) = K δ ( t) causality: h( t) = 0 t < 0 for wt 10/2005 37

Special LTI Systems (1) s(t) δ(t) g(t) =s(t) Ideal system: h ideal (t)=δ(t) Output signal of an ideal system: g(t) =δ(t) s(t) = s(t) wt 10/2005 38

Special LTI Systems (2) Ideal lowpass filter: h TP (t)=2f g si(π2f g t) H TP (f)= rect( f ) 2f g f g : cutoff frequency Ideal bandpass filter: h BP (t)=f si(πf t) 2cos(2πf 0 t) H BP (f)= ( ( f f 0 )+ ( f + f 0 rect rect )) f f f : f 0 : bandwidth center frequency wt 10/2005 39

Special LTI Systems (3) Short time integration: Integration: h KI ( t) g( t) g( t) = t h Int (t)=ε(t) t T = rect T t T = rect T T T s(τ ) dτ 2 2 * s( t) integration time g(t) =ε(t) s(t) = Z t s(τ) dτ wt 10/2005 40

LTI Systems with Stochastic Input Signals (1) Signals, sample functions: k s(t) h(t) k g(t)=h(t) k s(t) sample function of input signal Autocorrelation functions (Wiener Lee relation) Power density spectrum (Wiener-Khintchine theorem) sample function of output signal ϕ gg (τ)=ϕ ss (τ) ϕ E hh(τ) Φ gg (f)=φ ss (f) H(f) 2 wt 10/2005 41

LTI Systems with Stochastic Input Signals (2) Example: WGN at the input Power density spectrum: Autocorrelation function: h(t) n (t) n e (t) Φ ne n e (f) =N 0 H(f) 2 ϕ ne n e (τ) =N 0 ϕ E hh(τ) Variance at the output: σ 2 n e = N 0 ϕ E hh(0) = N 0 E h LTI system: ideal lowpass filter Φ ne n (f) =N e 0 rect ( f ) 2f g ϕ ne n e (τ) =N 0 2f g si(π2f g t) 2 σ ne = N 0 2f g wt 10/2005 42

LTI Systems with Stochastic Input Signals (3) Generalized Wiener Lee relation k s(t) h ( t 1 ) k g ( t 1 ) h ( t 2 ) k g ( t 2 ) Crosscorrelation function: WGN at the input: ϕ g1 g 2 (τ)=ϕ ss (τ) ϕ E h 1 h 2 (τ) ϕ ne1 n e2 (τ)=n 0 ϕ E h 1 h 2 (τ) h ( ) ( ) and orthogonal: 1 t h ϕ (τ)= 2 t n e1 n e2 0 wt 10/2005 43