Continuous and Discrete Signals and Systems 전북대학교 전자공학부 송민호 교수 전자공학부 백흥기 교수님의 강의자료를 바탕으로 구성하였음을 알립니다. Ch. 1. Representing Signals 1.1 Introduction 1.2 Continuous-Time vs. Discrete-Time Signals 1.3 Periodic vs. Aperiodic Signals 1.4 Energy and Power Signals 1.5 Transformations of the Independent Variable 1.5.1 The Shifting Operation 1.5.2 The Reflection Operation 1.5.3 The Time-Scaling Operation 1.6 Elementary Signals 1.6.1 The Unit Step Function 1.6.2 The Ramp Function 1.6.3 The Sampling Function 1.6.4 The Unit Impulse Function 1.6.5 Derivatives of the Impulse Function 1.7 Other Types of Signals 1.8 Summary 2 Signals & Systems Prof. M. Song 1.1 Introduction § Signal : detectable physical quantities or variables by which messages or information can be transmitted. • Examples of signals : − − − − human voice, television pictures, teletype data, and atmosphere temperature. Electrical signals are the most easily measured and the most simply represented type of signals. Therefore, many engineers prefer to transform physical variables to electrical signals. 3 Signals & Systems Prof. M. Song § Mathematically signals are represented as functions of one or more independent variables. • Examples of signals : − AC current or voltage (time) ex.: oscilloscope − the vibration of a rectangular membrane, image (x, y coordinates) − electric field intensity (time and space) In this S&S course, we focus attention on signals involving one independent variable, time t. 4 Signals & Systems Prof. M. Song 1.2 Continuous-Time vs. Discrete-Time Signals § The signal is a continuous-time signal, if the independent variable is continuous. x(t ) 연속(시간) 신호 t § The signal is a discrete-time signal, if the independent variable is discrete. 이산(시간) 신호 x[n] n 5 Signals & Systems Prof. M. Song A continuous-time signal which is continuous or discontinuous § x(t) is continuous at t = t1 , if x(t1- ) = x(t1+ ) = x(t1 ) § x(t) is discontinuous at t = t1 , otherwise § x(t) is continuous, if x(t) is continuous at all points t. t1 t1 6 Signals & Systems Prof. M. Song A piecewise continuous signal : continuous-time signal that is not continuous § x(t) is piecewise continuous, if the jump in amplitude at each discontinuity is finite. • Rectangular pulse function æt rect ç èt ö ì1, | t |< t / 2 ÷=í ø î0, | t |> t / 2 • Pulse train x(t) 1 -4 -3 -2 -1 0 1 2 3 4 5 t 7 Signals & Systems Prof. M. Song A discrete-time signal § If the independent variable takes on only discrete values t=kTs, the corresponding signal x(kTs) is called a discrete-time signal. § More details, in Ch. 6 x[n] n 8 Signals & Systems Prof. M. Song 1.3 Periodic vs. Aperiodic Signals § x(t) is a periodic signal, if 주기신호 x(t ) = x(t + nT ), n = 1, 2,3, L where T is the fundamental period. § Periodic Aperiodic (not periodic) 기본주기 § Real-valued sinusoidal signal x(t ) = A sin (w0t + f ) • x(t) is also periodic with period 2T, 3T, 4T,… • The fundamental radian frequency is w0 = 2p T K차 고조파 • The sinusoidal signal with wk = kw0 is called as k-th harmonic. 9 Signals & Systems Prof. M. Song Ex. 1.3.1 fk (t ) = exp [ jkw0t ] § fk (t ) is periodic with period T = 2p / w0 fk (t + T ) = exp [ jkw0 (t + T )] = exp [ jkw0t ] exp [ jkw0T ] = exp [ jkw0t ] = fk (t ) é 2p ù exp [ jkw0T ] = exp ê jk T = exp [ jk 2p ] = 1 ë T úû • A periodic signal is also periodic with period lT for any positive integer l. 10 Signals & Systems Prof. M. Song The sum of two periodic signals may or may not be periodic. • Let x(t) and y(t) be periodic signals with periods T1 and T2 x(t ) = x ( t + kT1 ) , y (t ) = y ( t + lT2 ) z (t ) = ax(t ) + by (t ) = ax ( t + kT1 ) + by ( t + lT2 ) z (t + T ) = ax(t + T ) + by (t + T ) To be periodic, T = kT1 = lT2 z (t ) = z (t + T ) Þ T1 l = T2 k 유리수 • z(t) is periodic if T1/T2 is a rational number. • z(t) is aperiodic if T1/T2 is a irrational number. 11 Signals & Systems Prof. M. Song Ex. 1.3.2 Which of the following signals are periodic? 2p 2p 2p t Þ T= = =3 3 w0 2p / 3 (a) x1 (t ) = sin (b) 2p ö æ 2p ö æ 4p ö 1 æ 14p x2 (t ) = sin ç t ÷ cos ç t ÷ = ç sin - sin ÷ 5 3 2 15 15 è ø è ø è ø 2p 15 2p T1 = = , T2 = = 15 14p /15 7 2p /15 T1 1 l 1 = Þ = Þ l = 1, k = 7 Þ T = kT1 = lT2 = 15 T2 7 k 7 12 Signals & Systems Prof. M. Song Ex. 1.3.2 (c) (d) x3 (t ) = sin 3t Þ T = 2p w0 = æ 2p x4 (t ) = x1 (t ) - 2 x3 (t ) = sin ç è 3 T1 2p 9 T1 = 3, T2 = Þ = 3 T2 2p 2p 3 ö t ÷ - 2sin 3t ø (irrational number) Þ aperiodic signal 13 Signals & Systems Prof. M. Song Ex. 1.3.3 In general, nonlinear operations on periodic signals can produce higher order harmonics? x(t ) = cos w1t , y (t ) = cos w2t z (t ) = x(t ) y (t ) = cos w1t cos w2t = 1 éëcos (w1 - w2 ) t + cos (w1 + w2 ) t ùû 2 § If w1 = w2 = w x(t ) y (t ) = cos 2 wt = 1 [1 + cos 2wt ] 2 § Nonlinear operations on periodic signals can produce higher order harmonics. 14 Signals & Systems Prof. M. Song § A periodic signal is a signal of infinite duration. Therefore, all practical signals are aperiodic. Nevertheless, the study of the system response to periodic inputs is essential in the process of developing the system response to all practical inputs. 15 Signals & Systems Prof. M. Song 1.4 Energy and Power Signals § If x(t) is a real-valued signal, • the energy over a time interval of length 2L : L E2 L = ò | x(t ) |2 dt -L • The total energy over t (-¥, ¥) : L E = lim ò | x(t ) |2 dt L ®¥ - L • The average power : é 1 L ù P = lim ê ò | x(t ) |2 dt ú L ®¥ 2 L - L ë û 16 Signals & Systems Prof. M. Song L 2 E = lim ò | x(t ) | dt L ®¥ - L § § § § § é 1 L ù P = lim ê ò | x(t ) |2 dt ú L ®¥ 2 L - L ë û Signal is said to be an energy signal, if E exists and 0 < E < ¥ Signal is said to be an power signal, if P exists and 0< P<¥ Energy signals have zero power. Power signals have infinite energy. Periodic signals have infinite energy and finite average power, and they are power signals. § In contrast, bounded finite duration signals are energy signals. 17 Signals & Systems Prof. M. Song Ex. 1.4.1 § The average power of the periodic signal with period T 1 T 2 P = ò x(t ) dt T 0 If x(t) is periodic with period T, then the integral is the same over any interval of length T. 1 T 2 P = ò sin wt dt T 0 1 T 1 - 2cos wt = ò dt 0 T 2 1 T1 1 T cos wt = ò dt + ò dt 0 0 T 2 T 2 1 = 2 18 Signals & Systems Prof. M. Song Ex. 1.4.2 Determine whether the signals are energy or power signals. energy signal not power signal x1(t) x2(t) A A exp[-t] 0 t not energy signal power signal exp[-t] 0 t A2 E1 = ò A exp [ -2t ] dt = 0 2 A2 é 1 L 2 ù P1 = lim ê ò A exp [ -2t ] dt ú = lim =0 L ®¥ 2 L 0 L ®¥ 4 L ë û ¥ 2 0 L 1 é ù 2 é E2 = lim ò A dt + ò A2 exp [ -2t ] dt ù = lim A2 ê L + (1 - exp [ -2 L ]) ú = ¥ úû L ®¥ ë 0 L ®¥ ê ë -L 2 û 2 L 1 é 0 2 A 2 P2 = lim A dt + ò A exp [ -2t ] dt ù = ò úû 2 0 L ®¥ 2 L ê ë -L 19 Signals & Systems Prof. M. Song Ex. 1.4.3 x(t ) = A sin (w0t + f ) § The average power of x(t) 1 T 2 2 P = ò A sin (w0t + f ) dt T 0 A2w0 2p / w0 é 1 1 ù = cos 2 w t + 2 f dt ( ) 0 ò ê ú 0 2p ë2 2 û A2 = 2 20 Signals & Systems Prof. M. Song Ex. 1.4.4 L 2 1 L ®¥ - L E1 = lim ò x (t )dt = ò t /2 -t / 2 A2 dt = A2t L E2 = lim ò A2 exp [ -2a | t |] dt L ®¥ - L A2 A2 = lim (1 - exp [ -2aL]) = a L ®¥ a § x1(t) and x2(t) are the energy signals § Almost all time-limited signals are energy signals. 21 Signals & Systems Prof. M. Song 1.5 Transformation of the Independent Variable 1.5.1 The Shifting Operation 전이연산 § Signal x(t - t0 ) represents a time-shifted version of x(t). x(t) x(t-t0) A A t0 t1 t0 t0 t2 0 x(t-t0) t t1 0 A t2 t 0 t0 t1 t0 t0 t2 delay advance t0 < 0 t0 > 0 t 22 Signals & Systems Prof. M. Song Ex. 1.5.1 x(t ) Þ x(t - 2), x(t + 3) 23 Signals & Systems Prof. M. Song Ex. 1.5.2 § Vibration sensors are mounted on the front and rear axle of a moving vehicle. Distance between two sensors : 6 ft § Sensor signals § Speed of vehicle is d = vt Þ v= d t = 6 ft = 50 ft/sec 0.12 sec 24 Signals & Systems Prof. M. Song Ex. 1.5.3 § Transmitted and received radar signals § Round-trip delay is t= 2R 2 ´ 45 = = 0.556 ms C 161875 25 Signals & Systems Prof. M. Song 1.5.2 The Reflection Operation 반전연산 § Signal x(-t) is obtained from the signal x(t) by a reflection about t = 0. x(t) x(-t) 3 -1 0 1 2 3 t -2 -1 0 1 t • If x(t) represents a video signal, then x(-t) is the signal when the rewind switch is pushed on. 26 Signals & Systems Prof. M. Song Ex. 1.5.4 § The operations of shifting and reflecting are not commutative. reflect x(t ) Þ shift x(-t ) Þ shift x(t ) Þ x(-(t - 3)) = x(-t + 3) reflect x(t - 3) Þ x(-t - 3) 27 Signals & Systems Prof. M. Song 우대칭 § Signal x(t) is even symmetric, if 기대칭 x(-t ) = x(t ) § Signal x(t) is odd symmetric, if x(-t ) = - x(t ) § An arbitrary signal x(t) can be expressed as sum of even and odd 우대칭, 기대칭 신호 signals. x(t ) = xe (t ) + xo (t ) 1 [ x(t ) + x(-t )] 2 1 xo (t ) = [ x(t ) - x(-t ) ] 2 xe (t ) = 28 Signals & Systems Prof. M. Song Ex. 1.5.5 ì1, t > 0 x(t ) = í î0, t < 0 1 xe (t ) = , t ¹ 0 2 x(0) = 1 Þ 2 ì-1/ 2, t < 0 xo (t ) = í t >0 î1/ 2, 1 xe (0) = , xo (0) = 0 2 29 Signals & Systems Prof. M. Song Ex. 1.5.6 ì A exp [ -a t ] , t > 0 x(t ) = í t<0 î0, ì1 ïï 2 A exp [ -a t ] , t > 0 xe (t ) = í ï 1 A exp [a t ] , t < 0 ïî 2 1 = A exp [ -a | t |] 2 ì1 ïï 2 A exp [ -a t ] , t > 0 xo (t ) = í ï- 1 A exp [a t ] , t < 0 ïî 2 30 Signals & Systems Prof. M. Song 1.5.3 The Time-Scaling Operation 시간척도조절 연산 § x(at) is a compressed version of x(t) if | a |> 1 § x(at) is an expanded version of x(t) if | a |< 1 -1 x(t) x(3t) x(t/2) 2 2 2 1 1 1 0 t 1 -1/3 1/3 t -2 0 2 t Recorded video Played back 3 times faster Played back @ ½ speed 31 Signals & Systems Prof. M. Song Ex. 1.5.7 x(3t - 6) = x [3(t - 2) ] scaling shifting Scaling first, shifting next!! 32 Signals & Systems Prof. M. Song Ex. 1.5.8 x(t ) = 1 - A exp [ -a t ] cos (w0t + f ) § Settling time: the time after which the signal stays within ±5% x(t ) = 1 - A exp [ -a t ] é 0.05 ù ln ê a ë A úû ( ts = 4.6052 for A = 0.5, a = 0.5) ts = - 1 x(t ) = 1 - 2.3exp [ -10.356t ] cos5t A = 2.3, a = 10.356, ts = 0.3697 s x(t / 2) = 1 - 2.3exp [ -5.178t ] cos 2.5t Þ ts = 0.7394 x(2t ) = 1 - 2.3exp [ -20.712t ] cos10t Þ ts = 0.1849 33 Signals & Systems Prof. M. Song Order of Performing Operations x ( a t + b ) = x (a ( t + b / a ) ) 1. 2. 3. 4. Scale by a. If a is negative, reflect about the vertical axis. Shift to the right by b/a if a, b have different signs. Shift to the left by b/a if a, b have different signs. § Operation of reflecting and time scaling is commutative. § Operation of shifting and reflecting or shifting and time scaling is not commutative. 34 Signals & Systems Prof. M. Song 1.6 Elementary Signals 1.6.1 Unit Step Function § Continuous-time unit step function 단위계단 함수 ì1, t > 0 u (t ) = í î0, t < 0 35 Signals & Systems Prof. M. Song Ex. 1.6.1 Rectangular pulse signal æt rect ç èt ö ì1, | t |< t / 2 ÷=í ø î0, | t |> t / 2 æ t ö A rect ç ÷ = A [u (t + a ) - u (t - a ) ] è 2a ø ætö A rect ç ÷ = A [u (t + 1) - u (t - 1) ] è2ø 36 Signals & Systems Prof. M. Song Ex. 1.6.2 Signum function is one of the most often used signals in communication and in control theory. sgn(t) t >0 ì1, ï sgn t = í0, t = 0 ï-1, t < 0 î 1 sgn t = -1 + 2u (t ) 0 t -1 37 Signals & Systems Prof. M. Song 1.6.2 Ramp Function 램프 함수 ì t, t ³ 0 r (t ) = í î0, t < 0 The ramp function is obtained by integrating the unit step function. ò t -¥ u (t )dt = r (t ) dr (t ) = u (t ) dt 38 Signals & Systems Prof. M. Song Ex. 1.6.3 Expression of arbitrary signals x(t ) = u (t + 2) - 2u (t + 1) + 2u (t ) - u (t - 2) - 2u (t - 3) + 2u (t - 4) t y (t ) = ò x(t )dt -¥ \ y (t ) = r (t + 2) - 2r (t + 1) + 2r (t ) - r (t - 2) - 2r (t - 3) + 2r (t - 4) 39 Signals & Systems Prof. M. Song 1.6.3 Sampling Function § Sampling function § Sinc function 샘플링 함수 Sa( x) = sinc x = sin x x sin p x = Sa(p x) px § Sampling function is a damped sine wave. § Sinc function is a compression version of Sa(x). 40 Signals & Systems Prof. M. Song 1.6.4 Unit Impulse Function 단위 임펄스 함수 § Unit impulse signal, Dirac delta function, delta function ò t2 t1 x(t )d (t )dt =x(0), t1 < 0 < t2 § Many physical phenomena such as point sources, point charges, concentrated loads on structures, and voltage or current sources acting for very short time can be modeled as delta functions. § Properties of unit impulse function 1. d (0) ® ¥ 2. d (t ) = 0, t ¹ 0 ¥ d (t ) dt = 1 3. ò 4. d (t ) is an even function; i.e., d (t ) = d (-t ) -¥ 41 Signals & Systems Prof. M. Song Examples of Delta Function p1(t) p2(t) p3(t) 1/ 1/ 1/ 0 t 0 d (t ) = lim pi (t ) e ®0 t 2 0 pt ö æ 1 p3 (t ) = e ç sin ÷ e ø è pt 2 t 2 § All function have the following properties for "small" ε 1. The value at t=0 is very large and becomes infinity as ε approaches zero. 2. The duration is very short and becomes zero as ε becomes zero. 3. The total area under the function is constant and equal to 1. 4. The functions are all even. 42 Signals & Systems Prof. M. Song Ex. 1.6.4 2 2 pt ö 1 æ sin p t / e ö 1 æ 1 2æ t ö p (t ) = lim+ e ç sin ÷ = lim+ ç sinc ç ÷ ÷ = elim + e ®0 e ® 0 ® 0 e ø e è pt / e ø e è pt èe ø • Property 1, 2, 4 • Property 2 ò ¥ -¥ p3 (t )dt = 1 >> syms x t; >> e = 0.01; >> p3 = e*(1/(pi*t)*sin(pi*t/e))^2; >> Sp3 = int(p3,t,-inf,inf) Sp3 = 1 43 Signals & Systems Prof. M. Song Important Properties – 1. Sifting Property ò t2 t1 ò t2 t1 x(t )d ( t - t0 ) dt = ò t 2 - t0 t1 -t0 x (t + t0 ) d (t )dt = x ( t0 ) , t1 < t0 < t2 ì x ( t0 ) , t1 < t0 < t2 x(t )d ( t - t0 ) dt = í otherwise î0, • Signal x(t) can be expressed as a continuous sum of weighted delta function. ¥ ¥ -¥ -¥ x(t ) = ò x(t )d (t - t )dt = ò x(t )d (t - t )dt ¥ = ò x(t )d (t - t )dt -¥ 44 Signals & Systems Prof. M. Song Graphical Interpretation of Sifting Property xˆ (t ) x(t) 0 x(k ) rect((t-k )/ ) k t ¥ æ t - kD ö x ( k D ) rect å ç ÷ D è ø k =-¥ 1 é (t - k D) ù rect ê ® d (t - t ) ú D ë D û ¥ x(k D) ® x(t ) é1 æ t - k D öù = å x(k D) ê rect ç ÷ ú [ k D - (k - 1)D ] k D - (k - 1)D ® dt è D øû k =-¥ ëD xˆ (t ) = ¥ ß ¥ x(t ) = lim xˆ (t ) = ò x(t )d (t - t )dt D® 0 ¥ å ·®ò · k =-¥ -¥ -¥ 45 Signals & Systems Prof. M. Song 2. Sampling Property • If x(t) is continuous at t0, then, x(t0) rect((t-t0)/ ) x(t )d ( t - t0 ) = x ( t0 ) d ( t - t0 ) x(t) t0 t lim x(t ) D® 0 1 æ t - t0 ö rect ç ÷ = x(t0 )d (t - t0 ) D è D ø • In the sifting property, the output is the value of the function evaluated at some point. • In the sampling property, the output is still a delta function with strength equal to x(t0). 46 Signals & Systems Prof. M. Song 3. Scaling Property • Property 1 d (at ) = 1 d (t ) |a| • Property 2 d (at + b) = 1 æ bö d çt + ÷ |a| è aø The pulse p(at) is a compressed (expanded) version of p(t) if a> 1 (a<1), and its area is 1/|a|. 47 Signals & Systems Prof. M. Song Ex. 1.6.5 § Gaussian pulse é -t 2 ù p (t ) = exp ê 2 ú 2 2pe ë 2e û 1 ò ¥ -¥ p (t )dt = ò ¥ -¥ é -t 2 ù exp ê 2 ú dt = 1 2 2pe ë 2e û 1 lim p (t ) = d (t ) e ®0 48 Signals & Systems Prof. M. Song Ex. 1.6.6 (a) 1 ò ( t + t )d (t - 3)dt = 0 2 -2 (b) 4 ò ( t + t )d (t - 3)dt = ( t + t ) 2 2 -2 t =3 = 3 + 32 = 12 (c) 1 1 1 exp t 2 d (2 t 4) dt = exp t 2 d ( t 2) dt = exp 0 = [ ] [ ] [ ] ò0 ò0 2 2 2 (d) ì1, t > 0 ò-¥ d (t )dt = íî0, t < 0 3 3 t ò t -¥ d (t )dt = u (t ), d u (t ) = d (t ) dt 49 Signals & Systems Prof. M. Song 1.6.5 Derivatives of Impulse Function § The derivative of the impulse function, or unit doublet, d ¢(t ) ò d ¢(t ) t2 t1 x(t )d ¢ ( t - t0 ) dt = - x¢ ( t0 ) , t1 < t0 < t2 § The unit doublet d '(t ) samples the derivative of the signal at time t=0. § High-order derivatives of d (t ) ò t2 t1 x(t )d ( n ) ( t - t0 ) dt = (-1) n x ( n ) ( t0 ) , t1 < t0 < t2 50 Signals & Systems Prof. M. Song Ex. 1.6.7 i(t) [A] § The current through an inductor of 1-mH 10 i (t ) = 10 exp [ -2t ] u (t ) - d (t ) A t 0 -1 § Voltage drop across the inductor d v(t ) = 10 10 exp [ -2t ] u (t ) - d (t )} { dt = -2 ´ 10-2 exp [ -2t ] u (t ) + 10-2 exp [ -2t ] d (t ) - 10-3 d ¢(t ) -3 v(t) [mV] 10-3 10-2 = -2 ´ 10-2 exp [ -2t ] u (t ) + 10-2 d (t ) - 10-3 d ¢(t ) V 0 t -10-2 2x10-2 51 Signals & Systems Prof. M. Song Ex. 1.6.8 [Error in Textbook] (a) æ 1 1ö 2 ¢ ( t 2) d ç - t + ÷ dt ò-4 è 3 2ø 4 4 1 1ö æ æ 3ö 2 ¢ ¢ ( t 2) d t + dt = 3( t 2) d ç ÷ ç t - ÷ dt = -6(t - 2) t =3/ 2 = 3 ò-4 ò 4 è 3 2ø è 2ø 4 2 (b) ò 1 -4 ò 1 -4 t exp [ -2t ] d ¢¢(t - 1)dt t exp [ -2t ] d ¢¢(t - 1)dt = ( t exp [ -2t ])¢¢ t =1 = (4t - 4) exp [ -2t ] t =1 = 0 52 Signals & Systems Prof. M. Song