CHAPTER 2 Discrete-Time Signals and Systems in the Time-Domain YANG Jian jianyang@ynu.edu.cn School of Information Science and Technology Yunnan University Outline • Discrete-Time Signals • Typical Sequences and Sequence Representation • The Sampling Process • Discrete-Time Systems • Time-Domain Characterization of LTI DiscreteTime Systems • Finite-Dimensional LTI Discrete-Time Systems • Correlation of Signals • Summary 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 2 Discrete-Time Signals • Basic signals – Unit sample or unit impulse sequence – Unit step sequence – Exponential sequence • Signal classification – Continuous-time / discrete-time signals – Deterministic / random signals – Energy signals • signals with finite energy – Power signals • signals with finite power – Energy signals have zero power, and power signals have infinite energy 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 3 Discrete-Time Signals • Time-Domain Representation – Sequence of numbers: • x(n) • n • — sequence — samples x ( n) — sample value or nth samples, a real or complex value – Figure of sequence: x(n) ,0.3,0.76,0,1, 2,0.92, • x ( n) is defined only for integer value of n 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 4 Discrete-Time Signals • Operation on sequences – Basic operation • Adder / Subtraction: x1 (n) x2 (n) y(n) • Scalar multiplication ( gain / attenuation ): • Delay / Advance: Ax( n) y( n) x(n n0 ) y(n) – Combination of Basic Operations • Multiplier: x1 (n) x2 (n) y(n) • Linear combination: 云南大学滇池学院课程:现代信号处理数字信号处理 a1 x1 (n) a2 x2 (n 3) y(n) Digital Signal Processing 5 Discrete-Time Signals • Operation on sequences – Sampling Rate Alteration ( special operations of for discrete-time signals ) • Up-sampling: x( n / L), y( n) 0, n 0, L, 2 L, , otherwise , • Down-sampling: y( n) x( nM ) 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 6 Discrete-Time Signals • Classification of Sequences – The number of sequences: finite / infinite • Finite-length sequences: x(n) 0, n N1 and n N 2 – Symmetry • conjugate-symmetric ( even ): • conjugate-antisymmetric ( odd ): x ( n) x ( n) x ( n) x ( n) – Periodity: periodic / aperiodic • Periodic sequence: x(n) x(n kN ), 云南大学滇池学院课程:现代信号处理数字信号处理 for all n, k is any integer. Digital Signal Processing 7 Discrete-Time Signals • Classification of Sequences – Energy and Power Signals energy : power : x x ( n) 2 n K 1 2 P lim x ( n ) k 2 K 1 n K energy signals : power signals : 云南大学滇池学院课程:现代信号处理数字信号处理 x <, P x , P Digital Signal Processing 8 Discrete-Time Signals • Classification of Sequences – Other types of Classification • Bounded: x ( n) Bx • Absolutely summable: n • Square-summable: x ( n) 2 x ( n) n 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 9 Typical Sequences and Sequence Representation • Some Basic Sequences – Unite sample sequence: 1, 0, ( n) n0 n0 • An arbitrary sequence can be represented by unite sample sequence in time-domain – Unite step sequence: ( n) n (k ), 1, ( n) 0, n0 n0 ( n) ( n) ( n 1) k 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 10 Typical Sequences and Sequence Representation • Sinusoidal and Exponential Sequences – The real sinusoidal sequence: x(n) A cos(0 n ), n – The exponential sequence: x( n) A n Ae ( 0 j0 ) n A e 0ne j (0n ) A e 0n cos(0 n ) j A e 0n sin(0 n ) • The sinusoidal sequence are periodic of period N as long as N is an integer multiple of 2 . The smallest 0 possible N is the fundamental period of the sequence. 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 11 Typical Sequences and Sequence Representation • Some Typical Sequences – Regular window sequence: 1, w R ( n) 0, 0 n N 1 otherwise – Real exponential sequence: x ( n) a n ( n ) • Representation of an Arbitrary Sequence – An arbitrary sequence can be represented as a weight sum of basic sequence and its delayed version. x ( n) x(k ) (n k ) k 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 12 The Sampling Process • Uniform sampling: – Often the discrete-time sequence is developed by uniformly sampling a continuous-time signal xa ( t ): x(n) xa (nT ) • F 1 T , the sampling T frequency • 2 F , the sampling T T angular frequency 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 13 The Sampling Process • Aliasing: – When T 2 MAX , a continuous-time sinusoidal signal of higher frequency would acquire the identity of a sinusoidal sequence of lower frequency after sampling. e.g. 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 14 Discrete-Time System • Discrete-time system y( n) H [ x( n)] Input x(n) - n H[] Output y(n) • Simple Discrete-Time Systems – The accumulator – The M-point moving-average filter – The factor-of-L interpolator 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 15 Discrete-Time System • Classification of Discrete-Time System – Linear system: if x1 (n) y1 (n), x2 (n) y2 (n), then x1 (n) x2 (n) y1 (n) y2 (n) – Shift-Invariant System: if x(n) y(n), then x( n n0 ) y( n n0 ) – LTI System: The linear time-invariable discrete-time system satisfies both the linear and the time-invariable properties. 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 16 Discrete-Time System • Classification of Discrete-Time System – Causal System: In a causal discrete-time system, the n0 th output sample y( n0 ) depends only on input samples x ( n) for n n0 and does notdepend on input samples for n n0 . if u1 ( n) y1 ( n) then and u2 ( n) y2 ( n) {u1 ( n) u2 ( n), for n N } implies also that { y1 ( n) y2 ( n), for n N } 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 17 Discrete-Time System • Classification of Discrete-Time System – Stable System: Definition of bounded-input, bounded-output ( BIBO ) stable. x( n) Bx , n if then y( n) B y , n • Passive and Lossless Systems – The passivity: n – The losslessness: 2 y ( n) 2 x ( n) n the same energy 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 18 Discrete-Time System • Impulse and Step Responses – Input sequence → output sequence – Impulse response h( n) : – Step response s( n) : 云南大学滇池学院课程:现代信号处理数字信号处理 ( n) h( n) ( n) s( n) Digital Signal Processing 19 Time-Domain Characterization of LTI Discrete-Time Systems • Input-Output Relationship – The response y(n) of the LTI discrete-time system to x(n) will be given by the convolution sum: y ( n) k k x(k )h(n k ) x(n k )h(k ) x( n) h( n) – The operation h( k ) h( k ) h( k ) h(n k ) Step 2, shift n sampling period: • Step 1, time-reverse: • • Step 3, product: x( k )h( n k ) v( k ) • Step 4, summing all samples: 云南大学滇池学院课程:现代信号处理数字信号处理 k k v(k ) x(k )h(n k ) Digital Signal Processing 20 Time-Domain Characterization of LTI Discrete-Time Systems • Some useful properties of the convolution operation – Commutative: x1 (n) x2 (n) x2 (n) x1 (n) – Associative for stable and single-sided sequences: x1 (n) [ x2 ( n) x3 ( n)] [ x1 ( n) x2 ( n)] x3 ( n)] – Distributive: x1 (n) [ x2 (n) x3 (n)] x1 (n) x2 (n) x1 (n) x3 ( n)] 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 21 Time-Domain Characterization of LTI Discrete-Time Systems • Simple Interconnection Schemes – Cascade Connection: h(n) h1 (n) h2 (n) h1 (n) h2 (n) h2 (n) h1 (n) h1 (n) h2 (n) – Parallel Connection: h1 ( n) h2 ( n) – Inverse System: 云南大学滇池学院课程:现代信号处理数字信号处理 h(n) h1 (n) h2 (n) h1 (n) h2 (n) h1 (n) h2 (n) (n) Digital Signal Processing 22 Time-Domain Characterization of LTI Discrete-Time Systems • Stability Condition in Terms of the Impulse Response – An LTI digital filter is BIBO stable if only if its impulse response sequence h( n) is absolutely summable, i.e.: S h( n) n • Causality Condition in Terms of the Impulse Response – An LTI discrete-time system is causal if and only if its impulse response is a causal sequence satisfying the condition: h( k ) 0, 云南大学滇池学院课程:现代信号处理数字信号处理 for k 0 Digital Signal Processing 23 Finite-Dimensional LTI Discrete-Time Systems • The difference equation: – An important subclass of LTI discrete-time systems is characterized by a linear constant coefficient difference equation of the form: N d k 0 M k y( n k ) pk x ( n k ) k 0 – The order of the system is given by max( N, M ) 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 24 Finite-Dimensional LTI Discrete-Time Systems • Total Solution Calculation – The complementary solution • The homogeneous difference equation: • The characteristic equations: 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 25 Finite-Dimensional LTI Discrete-Time Systems • Total Solution Calculation – The particular solution y p ( n) is of the same form as specified input x ( n) . – The total solution: 云南大学滇池学院课程:现代信号处理数字信号处理 y( n) yc ( n) y p ( n) Digital Signal Processing 26 Finite-Dimensional LTI Discrete-Time Systems • Zero-Input Response and Zero-State Response – zero-input response = complementary solution with initials; – zero-state response = the convolution sum of x(n) and h(n). if x( n) 0 the solution is yzi ( n), and if applying the specified input with all initial conditions set to zero the solution is yzs ( n), then the total solution is : 云南大学滇池学院课程:现代信号处理数字信号处理 y zi (n) y zs ( n) Digital Signal Processing 27 Finite-Dimensional LTI Discrete-Time Systems • Impulse Response Calculation 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 28 Finite-Dimensional LTI Discrete-Time Systems • Impulse Response Calculation – The solutions 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 29 Finite-Dimensional LTI Discrete-Time Systems • Location of Roots of Characteristic Equation for BIBO Stability – A casual LTI system characteristic of a linear constant coefficient difference equation is BIBO stable, if the magnitude of each of the roots its characteristic equation is less than 1. – The necessary and sufficient condition: k 1 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 30 Finite-Dimensional LTI Discrete-Time Systems • Classification of LTI System – Based on impulse response length • Finite impulse response ( FIR ): h(n) 0, for n N1 and n N 2 , with N 1 N 2 y( n) N2 h(k ) x(n k ) k N1 • Infinite impulse response ( IIR ): n y( n) x ( k )h( n k ) k 0 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 31 Finite-Dimensional LTI Discrete-Time Systems • Classification of LTI System – Based on the output calculation process • Non-recursive system: If the output sample can be calculated sequentially, knowing only the present and pass input samples. • Recursive system: If the computation of the output involves past output samples. – Remarks: • FIR — Non-recursive • IIR — Recursive 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 32 Correlation of Signals • Definitions – A measure of similarity between a pair of energy signals, x(n) and y(n), is given by the cross-correlation sequence defined by: rxy ( l ) ryx ( l ) x(n) y(n l ), n y( n) x ( n l ) n rxy ( l ) – l 0, 1, 2, m y(m l ) x (m ) rxy ( l ) y( n) x[( l n)] y( l ) x ( l ) n The autocorrelation sequence of x(n) is given by: rxx ( l ) 云南大学滇池学院课程:现代信号处理数字信号处理 x ( n) x ( n l ) n Digital Signal Processing 33 Correlation of Signals • Properties of Autocorrelation and Cross-correlation Sequences – Set rxx (0) x 0 and ryy (0) y 0 as energies of the sequences x(n) and y(n) , then we can get rxx (0)ryy (0) rxy2 (l ) 0 or equivalently rxy (l ) rxx (0)ryy (0) x y – If y(n) = x(n), then rxy (l ) rxx (0) x • The sample value of the autocorrelation sequence has its max value at zero lag ( l = 0 ). 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 34 Correlation of Signals • Properties of Autocorrelation and Cross-correlation Sequences – If y( n) bx( n N ) , where N is integer and b>0 is an arbitrary number. In this case y b2 x, so brxx (0) rxy ( l ) brxx (0) 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 35 Correlation of Signals • Normalized Forms of Correlation: r (l ) xx ( l ) xx , rxx (0) xy (l ) rxy ( l ) rxx (0)ryy (0) • Correlation Computation for Power and Periodic Signals – Power signals: K 1 rxy ( l ) lim x( n) y( n l ), K 2 K 1 n K K 1 rxx ( l ) lim x ( n) x ( n l ) K 2 K 1 n K – Periodic signals: 1 r (l ) xy N N x(n) y(n l ), n 0 云南大学滇池学院课程:现代信号处理数字信号处理 1 r (l ) xx N N x ( n) x ( n l ) n 0 Digital Signal Processing 36 Summary • The LTI system has numerous applications in practice. • The LTI system can be described by an inputoutput relation composed of a linear constant coefficient difference equation. • The LTI discrete-time system is usually classified in terms of the length of its impulse response. 云南大学滇池学院课程:现代信号处理数字信号处理 Digital Signal Processing 37