2 Review of z-Transform 2.1 INTRODUCTION Laplace transform is an extension of the continuous-time Fourier transform. This extension was motivated by the fact that Laplace transform can be applied to a broader class of signals than the Fourier transform. Since there are many signals for which the Fourier transform does not converge but it does for the Laplace transform. z-transform for discrete-time signals is the discrete-time counterpart of the Laplace transform for contiunous-time signals, and both of them have a similar relationship to the corresponding Fourier transform. The motivation for introducing generalization of Discrete-time Fourier ‘Transform (DTFT) are 1, Discrete-time Fourier Transform (DTFT) does not convere® for all sequences but its generalization (z-transform) can be applied to a broader class of signals. 2. Asecnod advantage is thatin notation is often conveniently us ‘Transform (DTFT). Itis worth noting that transform techniques are used for analysis of various signals and Linear-Time-Invariant (LT]) systems. Comparison between Transforms and Logarithms Jace transform, Fourier transform, In various transforms such as Lap cr immeform etc, we use one transformation formula which converts signals or sequences from time-domain to another corresponding domain depending upon a particular transform that domain is easy and suitable for computations and al: an extract more information in that domain which is necessary an inverse transform formula to 9g: analytical problems the z- transform ‘ed than the Discrete-time Fourier 50 we C for signal processing. We use ain convert back that signal into time-domain. 23Digital Siena! Procesting yg, 2 oe yses of transti Purpose of Transformation. There are vo purpo: formay (@) For ‘oiresting ‘more information from the transformed domes (8) For the purpote of simplification of computations in that dom, Logaritimisured for simplification and making easy computay cent ie not used forthe purpose of extracting any informay. Bea Rquence or signal that was done in transfo techniques. _ ‘As welenow, traneform techniques are used for analy$is of signaty Li} onctome ‘Ihe z-transform plays the same role in the analygy discrete-time signals and LTI systems as the Laplace transform in the analysis of continuous-time ignals and LTI systems, 2.2 DEFINITION OF THE =-TRANSFORM s-transforme are of two types 1. Two-sided z-transform 2. One-sided z-transform : ‘The two-sided z-transform of a discrete-time signal s(n) is defined a, the power series, Play ‘Say = z-transform of sequence s(n) = z[sto]= Eee where z is a complex variable. ze rel where r= [21 = magnitude of 2 and @ is the angle of z. It is also called bilateral z-transform. One-sided z-transform of a discrete-time signal s(n) is defined as Se)= 2[ol]= Salma It is also called unilateral z-transform, ‘The difference between two-sided z-transform and one-sided s-transform is that lower limit of summation is zero for one-sided 2+transform and —e for two-sided z-transform. Generally one-sided = Atansform i sed for solving difference equations of discrete-time LT For causal sequences the two-sided and one-sided led and one-sided z-transforms are equivalent. Two-sided z-transform can be defined only for a particular reaion of convergence (ROC). The ROC is not important for one-sided z-transform. One-sided z-transform is used for solving linear difference ‘equations with non-zero initial conditions. (2.2) Reviews of 2 Transform *Maien Po 2.4 REGION OF CONVERGENCE (ROC) For some values of, the power series in Eqn. (2.1) does not eraverge toa finite valve fen in Ban. (2.1) converges ‘The portion ofthe e-plane for which the sci in En. (2. calcd the region of convergence (ROC). The ROC depends upon the ivcaltude of The ROC cannot contain any pole, since the wees seers tninite atthe poles. The ROC can be circle, interior ofa circle the exterior of circle, an annulus, or the entice plane. the ROC for the z-transform 2.3.1 Possible Configurations of ‘These configurations are : 1. Interior of a circle (Fig. 2.1(a)] 2, Exterior of a circle (Fig. 2.1(6)] 3. An annulus (Fig. 2.16) 4. The entire z-plane (Fig. 2.1(¢)]. A ime) DD Wj D Pete) Qj © @ Fig. 2.1 Possible configurations of the ROC for the z-transform. (a) Interior ofa cirele (b) Exterior of a cirele (c) An annulus (d) The entire xplane,Digital Signal Processin, 26 nea 2.3.2 Properties Properties ofthe region of ‘Table 2.1. form of ROC for #-trans! convergence for the z-transform jg big, ‘Table 2.1. Entire z-plane S Entire z-plane except z «9 Discrete-time sequence, 1. Unit-sample sequence, &(n) 2. Finite-duration causal sequence 3. Finite-duration sequence with) s(n) # 0 for some n <0 ‘sin) = 0 for all n > 0 4, Finite-duration sequence with ‘(n) # 0 for some n <0 Entire z-plane except for 2. Entire z-plane except forz=Oand2 = = | s(n) = 0 for some n > 0 | 5. Right-sided sequence Outward from outermost poy n) = 0 for some n <0 ; 6. Right-sided sequence Outward from outermost pole, z = © is excluded, Inward from innermost pole s(n) #0 for some n <0 7. Left-sided sequence s(n) = 0 for all n > 0 8. Leftsided sequence s(n) # 0 for some n > 0 9. Two-sided sequence Inward from innermost pole 2 = 0is excluded. Annulus Example 2.1 Determine the z-transform: i id ramps 21 De anaforms of the folowing Tale (@) sm) = (2,4, 5,6,1) © stm) =(1,2,9,4,5) © sin) = (0, 0,1, .0.1,2,8) (a) sin) = (a) (© sin) = 82-2), k>O0 OH sin) = 8(r+h), k>0 Solution. (a) transform is defined as Se) = & [a(n)]= So(njen Review of Transform Given s(n) = (2,4, 5,611 2 or Say= Seta) amen = 5(0)2-0 + (1)z71 +9 (2)2°? +8(8)29 + 5(4)e* $527 462% +1 Q1+de 24427 +522 462% +2 ROC : Entire z-plane except z= (b) Given s(n) = (1,2, 3, 4.51 transform is dofined as si)= & [aq]= Een Boe (-2)2? +5(-t1)z" +5 (0)2°+5()27 +9(2)2* = 12242284 3.14424 452% = 22422484424 +527 ROC : Entire z-plane except z = and z= 0 (©) Given sin) =(9,0,1,2,8) z-transform is defined as St) = % [s(n)]= Za = 8(0)2° +s(1)2- +5(2)2-? +5(8)28 +5(4)2 + = 0.294021 +127 +229 +324 zte2e943r4 ROC : Entire z-plane except z = 0 A, n=0 0, n #0 z-transform is defined as sc) = & [a(n)]= Eee" Sateen (2) Given sn) = (2) 2 ot s(-D)22 +9(-1)at +8(0)29 +5 (az +s(Q)2 2 + e022 402441294029 4028 ad or S@)=1 ROG : Entire z-planeDigital Signal Process; 28 ist receasing 0 s(n) = 8(n-K),k>0 , z-transform is defined as (©) Given Ste) = &[s(n)]= Ealn)enm = 28m Substituting m =n—-k or n=m+k,we get siz) = _Ea(mjerh) = Lo(mjerment =z Salmo = where S8(m)e-" =1 Se) = %[5(n-h) ROC : Entire z-plane except z = 0. | () Given a(n) = 5(n+h),k>0 z-transform is defined a: Se) = Z[5r+A)]= Lanta Substituting m =n +k or n= m-~—k, we get Se) = LS(m)em-") Ta(m)e-mek = 2 S8(m)e-m azknzh where E (mem a1 S@) = %[8(n+k)]=24 ROC : Entire z-plane except z = Review of 2-Transform Example 2.2 Determine 29 the z-transform of the following discrete-time signals (@) &(n) = win) @) ln) = Arun) (©) on) =~ Arun 1) (d) on) = coa(won)u(n) (© s(n) = sin(@yn)u(n) sn) = Atu(n)+B%u(-n—1) @) s(n) = nu(n) a(n) = un = 2) Solution. (a) Given s(n) = u(n) and u(n) is defined as neo un) = ona (6) Given s(n) = A"u(n) S-transform is defined as Se) = B[s]= Esme eae tee te gem Spon = = Sarum) = Zaren= (assy, neo no (This is a geometric series for infinite number of points) (42-8) (Ae) (Ae ee (det 1 * TatDigital Signal Processing (Dg, 0 roc: eet or |e] > [Al where A scalar quantity: ROC ofthez-transform ofthe sequence ain) = Atu(n) is [2] > 1AL ‘This is a causal sequence. ‘The ROC is the exterior of & circle having radius || ROC of thez-transformof the sequence s(n) = A7u(n) ean be Shown graphically on plane and ie illustrated in Fig. ‘Given sin)=- Attn D Fig. 22 ROC of the z-transform of the ‘sequence s(n) = A”u(n. {3 20 = |-ann<0 transform is defined a8 se) = &[s(0)] Eee" = E-Atu(n-2)e" . i arenes 3 ety Substituting ! = -2 (ie Fane [fares (ary (arte) “| (Sum of geometric series fa inifinite number of points.) Roc :|A“Hf[e!%|>1 and feet ]z|>|e]>1 It we st |e ROC of thez-transform ofthe sequence s(n) = cos (a4) u(n) is exterior to the unit circle, It ean be shown graphically on z-plane. Its ROC is shown in Fig. 24