Properties of Exponential Form of Fourier Series Coefficients Note : c, and d, are exponential form of Fourier series coeflicients of x(t) and y() n Property Continuous time Fourier series periodic signal coefficients Linearity A x(t) + B y(t) Time shifting x(t -,) e Conjugation x*(t) Time reversal x(-t) x(at) ; a>0 jm20t x(t) (x(0) is period with period T/a) Multiplication x(t) y(t) d Differentiation -x(t) d x(t) dt Integration Ac, + B d, jns 20to Frequency shifting Time scaling respectively. n- m C C, No change in Fourier coefficient) too - m =-0 jn2,Cn 1 jns2, (Finite valued and periodic only if a, 0) Periodic convolution x(t) y(t -) dt Te, d Cn = C-n Symmetry of real signals x(t) is real Re{cn) = Re{c-n Im{cn} -Im{c-n) Real and even x(t) is real and even C, are real and even Real and odd x(t) is real and odd c, are imaginary and odd Parseval's relation Average power, P of x(t) is defined as, P = The average power, P in terms of Fourier series coefficients is, P= n=-0 Note : l. The term lc. respresent the power in n" harmonic component ofx(t). The total average power in aperiodic signal is equal to the sum of power in all of its harmonics. 2. The term lc. for n= 0, l, 2, ...is the distribution of poweras a function of frequency and so itis called power density spectrum or power spectral density of the periodic signal Gibbs Phenomenon The exponential form of Fourier series of a continuous time periodic signal x(t) is given by +00 x(t)=c, emag! n=-0 The above equation is frequency domain representation of the signal x(t) as a sum of infinite serio with each term in the series representing a harmonic frequency component. When the signal x() i reconstructed or synthesised with only N number of terms of the infinite seires, the reconstructed sional exhibits oscillations (or overshoot or ripples), especially in signals with discontinuities. x(t) A Fig 4.17a. 0 2 n=1 Fig 4.17b. t x(t)4 n=3 A Fig 4.17c. x(t)A n=5 A Fig 4.17à ,(t)4 n=59 A Fig 4.17e. T 0 2 Fig 4.17 : Successively closer to a square pulse signal. Consider a periodic square pulse signalapproximations shown in fig 4.17a. The reconstructed signal using of Fourier series are shown in fig N-tems 4.17b, c, d and e. (Refer MATLAB program 4.5 in be observed that the reconstructed signal exhibits oscillations and the oscillations are section 4.18). It can points of discontinuities with increasing value of N. compressed towards Also it can be observed that, at the points ot discontinuity, the Fourier series converges to average value of the signal on either side of discontinuy This phenomenon was named after a famous mathematician, Josiah Gibbs, as Gibbs phenomenon and the oscillations are called Gibbs oscillations. (Josiah Gibbs is a famous mathematician who first mathematical explanation for this phe x(t)+ Example Determine the trigonometric form of Fourier series of the waveform shown in fig4.1.1. t Solution The waveform shown in fig 4.1.1 has even symmetry, half wave symmetry and quarter wave symmetry. b, = 0 and a, = The mathema x(t) cosnslt dt Fig 4.1. 1. iation of the square wave is, x(t) = r T t = 0 to T t = A T to 4 2 Evaluation ofa, T/4 A COsnsz,t dt = T/2 Acosns2,t dt + ;|-A) cosn2,t dt T/4 4A qT/4 2n. ]T/4 4A sinn2gt qTi2 sinng,! T Jo no sinn 4A T JTI4 n 4A 2r = 2T T sin sin 0 T 2T 4A T 2n T 2A 4A T n n sin n 4A n sin 2 sin n 4A Sinr 2 2 For even values of n, sin= 0 For odd values of n, sin= 1 2 a = 0 a, = .. a, = ; for even values of n 4A sin 4A 1x T 4A 3 x TU ag = a7 = 4A 5 X T 4A 7 x T ; for odd values of n 2 4A T sin 2 sin sin sin TU 3n 4A 2 3T 5t 4A 2 5 7n 4A 2 7n JTI4 2T T T2 2 T 2nT 2A 2 T Jo 4A 2 7T/2 sinnt T and so on. 2nT sin sin 0 = 0 sin n =0 for integer n Fourier Series The trigonometric form of Fourier series of x(t) is, x(t) = b, sinn2,t Xa, cosnl,t + + 2 n=1 n = 1 Here, a, =0,b =0 and a exists only for oddvalues of n. n a, cosns2,t .. x(t) = n= odd = a, cos,t + a, COs 32,t + a, cos 52,t + a, cos7lot + 4A = 4A T cOs2t cos2,t 4A 3T cos312%t + 4A 5t cos502,t cos 32,t cos 502,! 3 5 4A 7 cos72t + cos792,t .... 7 + Fourier Transform Development of Fourier Transform From Fourier Series The exponential form of Fourier series representation of a periodic signal is given by, x(t) = )c,eintlg +00 D=-0 T/2 where, c, = -jns20t T dt -T/2 In the Fourier representation using equation (4.29), the c for various values of n are the spectral components of the signal x(), located at intervals of fundamental frequency B Therefore the frequency spectrumn is discrete in nature. The Fourier representation of a signal using cquation (4.29) is applicable for periodic signals. For rourier representation of non-periodic signals, let us consider that the fundamental period tends to ninity. When the fundamental period tends to infinity, the tundamental frequency 2, tends to zero or becomes very small. Since fundamental frequency 9is very small, the spectral components will lie very ach other andso the frequency spectrum becomes continuous. In order to obtain the Fourier representation of a non-periodic signal let us consider that the fundamental frequency S2, is very small. Let. 2, On replacing S2, by A2 in cquation (4.29) we get. N() = ) c,cn n On substituting for c, in the abOve cquationfrom cquation (4.30)(by taking t as dummy variable for integration) we get. T dte'inAt jnAs2r x(t) = ..431) -T/2 2r We know that, S2, = 2nE, = : T Since S2, ’ AQ. T 2T ..4.32) 2T On substituting for 1 from equation (4.32) in equation (4.31) we get, T/2 +00 X(t) enAlr drejnA2t x(t)= ) 2T 2 -T/2 T/2 x(t) ejn dten AN n=--T/2 For non-periodic signals, the fundamental period T tends to infinity. On letting limit T tends to infinity in the above equation we get, x(t)= T’ Lt 1 21 + T/2 -jnA2r drejnat A n=- -T/2 When T ’ ; -jns2r do .. x(t) = eJn do e' 1 2T where, Xj2) = Xx(t) e h dr = The equation (4.34) is Fourier .(4.33) Since t is a dummy variable, Let r =t. dt ..(4.34) transtorm of x() and equation (4.33) is nverse Fourier transtom Since the equalion (4.34) extracts the frequency components of the signal, usng equation (4.34) is also called analysis of the signal x(). Since the cquation (4,33) transtomation combines the frequency components of the signal, ihe nverse transtormalion using equation is also called synthesis of the (4.33) signal x(). of x(t). Definition of Fourier Transform Let, x(t) = Continuous tme signal X(i) =Fourier transform of x() The Fourier transform of continuous time signal, x(t) is defined as, X(G2) = x()e t Also, X(iS) isdenoted as F{x(t)} where "F" is the symbol used todenote the Fourier transform operation. F{x()} = X(j2) = x(t) e Mdt ..(4.35) Note : Sometimes the Fouiertransformis erpressed as afunction of cyclic frequency F, rather than radian frequency S2. The Fourier transform as afunction of cyclic frequency F, is defined as, +00 XGF) = x() e dt Condition for Existence of Fourier Transform The Fourier transform of x() exists if it satisfies the following Dirichlet condition. 1. The x(t) be absolutely integrable. +0 i.e., x(t) dt < o Thex(t) should have a finitenumber of maxima and minima within any finite interval. 3The x() can have a finite number of discontinuities within any interval. Definition of Inverse Fourier Transform The inverse Fourier transform of X(2) is defined as, +00 x(t) = F{X(j2)} = 1 2 X(j2) e dQ ....(4.36) The signals x(t) and X(/2) are called Fourier transform pair and can be expressed as shown below, x(t) F X(j2) Note : When Fourier transform is expressed as afunction ofcyclic frequency F he inverse Fourier transform is defined as, Properties of Fourier Transform 1. Linearity Let, F{x, (t)} - X,(G2): F{x,()} - X,((2) The linearity property of Fourier transform says that., Fla, x,() +a, x,(t)} =a, X,(2) ta, X,(j2) Proof: By definition of Fourier transform, X,j2) = x,(t) e- dt and X,() = x, t) e t ..4.40) Consider the linear combination a x(t)+a, x,(t). On taking Fourier transform of this signalwe get, T{o, x,(t) + 4, x,(1} = | a, x,t) + a, X,t)] e dt = , x, (t) e dt + a, x,t) e dt + a,x, t) e 2. Time shifting If F{x()}= Xj2) then F{x(t-t,)} = e ig! Xi2) dt Using equation (4.40)| = 0, Xj2) + a, X,i2) The time shifing property of Fourier transform says dt that, Proof: By definition of Fourier transform, F{x()} = Xlij2) = | x(t) e t ....14.4) Let, - : F{t - t,)} =(t - t,)e" dt =x) e ..t=t+f tol dr On differentiating dt = de 4 |xt) e xe odr =e Mo xe)e de Since t is a dummy variable for ingetration we can change to t. - +00 = eMox(t) e t, =t dt = eMo X(j2) Usingequation (4.41) 3. Time scaling The time scaling property of Fourier transform says that, If F{x(t)} =X(j2) then 1 F{x(a)} x2) Proof: By definition of Fourier transform, +00 xt) e dt F{x) = Xlj2) = F{xlat)) = xlat) e xt)e dt = |xt:)e)dr |Put, at=t dt = : .. t= The term -;dt = x\c) e dt d is The above transform is applicable for positive values of "a". similar to the form of Fourier If "a" happens to be negative then it can be proved that, transform except that replaced by (). dr = X) Fklat} = Hence in general, Fxlat)) = G ) for both positive and negative values of"a' 4. Time reversal The time reversal property of Fourier transform says that, If F{x(t)} =X(j2) then F {x(-t)} - X(-j2) Proof: From time scaling property we know that, FIat)) = 2) Let, a =-1. .:. FIx(-)) = X(H) is S. Conjugation The conjugation property ot Fourier transforn says that. If F{x(t)} = X(ÇQ) then Fx(0) = X'(-j2) Proof: By definition of Fourier transform, F{xtM} = Xi2) =xlt) e dt The term.xlt)e dt is similar to the form of Fourier transform except that 2is replaced by -S2 | x(t)e-'dt = X(- i2) - [X-i2)] = X*(-2) 6. Frequency shifting says that, The frequency shifting property of Fourier transform If F{x(t)} = X(j2) then F{ePal x(1)} - XG(2-2,) Proof: By definition of Fourier transform, F{xt)} = Xij2) = x(t) e" dt The term | xt) e il0-20" dt is similar to FfePo' xtn} [|eo' xt] e" t= xnefo' e dhe formof Fourier transformexcept that 2is replaced by 2-S [x(0) e- -00" dt = x(i(2 - lo)) 7. Time differentiation The differentiation property of Fourier transform says that, If F{x()} X(j9) then = j2 X(j2) Proof: Consider the definition of Fourier transform of xt). F{x(n} = Xij2) =xt) e t dt ...4.42) dt d = exloo) - e*" x-o) + j2 = j9 x)=0 x(t) e " t e0 x(th e dt = 2Xi) |Using equation (442) 8. Time integration The integration property of Fourier transform says that, IfF{x()} -X(j2) and X(0) =0then j2 X(j2) Proof: Considera continuoUs time signal x(t). Let Xlj) be Fourier transform of xt). Since integration and differentiafion are inverse operations, xt) can be expressed as shown below. = x(t) dt On taking Fourier transform of the above equation we get, Using time differentiation =F{at1} 1 property of Fourier transform. F{xt} =X(2) X(j2) 9. Frequency differentiation The frequency differentiation property of Fourier transform says that, If F{x(t)} =X(2), then d F{t x(t)} = jdS2 X(j2) Progf: By definition of Fourier transform, X(js2) = Fit) = xlt) e dt On differentiating the above equation with respect to 2 we get, d + -X(j2) = dQ + x(t) e dt d do e dt |Interchanging the order of integration and differentiation da Using definition of Fourier transform. d 10. Convolution theorem The convolution theorem of Fourier transform says that, Fourier transform of convolution of two signals is given by the product of the Fourier transform of the individual signals. i.e., if F{x,()} = X, (i2) and F{x,()} X,(j2) then, ...4.43) F{x,() * x,()} = X,(2) X,(S) The equation (4.43) is also known a convolution property of Fourier transforrn. With reference to chapter-2, section -2.9 we get, x,() x,(t - ) dr x(t) * x, (t) = .(4.44) where t is a dummy variable used for integration. Proof: Let x{t) and x,() be two time domain signals. Now, by definition of Fourier transform, X(2) = Fo,t)} = x,t) e dt X,(i2) = Fh,tt} = dt x,(t) e ..44.45) ...4.46) Using definition of Fourier transform we can write, e- dt = Let, e e x e xe : e xe-d = e-t where, M=t- and so, dM = dt ...4.47) ....4.48) ....4.49) Using equations (4.48) and (4.49), the equation (4.47) can be written as, +0 - x) e dt xX, (M) ekM dM ...4.50) In equation (4.50), tand Mare dummy variables used for integration,and so they can be changed to t. Therefore equation (4.50) can be writen as, F{*,#) " x, }=x,() e dt xX t) e dt = X(i2) X, (i2) Using equations |(4.45) and (4.46) |1. Frequecy convolution Let. F{x,(tO} = X,((2): F{x,(t)} - X,(G2). The frequency convolution property of Fourier transform says that. F{ (0) x,)) X,() x,(0 - 2) d 2 Proof: By definition of Fourier transtorm, F(xth} = X-) = xt e t F{*,)x,)} = x4)x,t) e d 4.5) By the definition of inverse Fourier transform we get, 1 |X,2) e dQ = 2r x,t) =FX,(2)} = 2 (4.521 On substituting for X,(t) from equation (4.52) in equation (4.5l] we get, |Here 2is the vairable USed for ingetrafion. Let us change 2 to À ) 2 x,(t) e e dt da 1 X,t) e - dt dA X,à) 2 - |Interchanging the order of integration. x,it) et- dt |The term, is similar to the fon of Fourier transforn except thgt 2is reploced by S2 2 12. Parseval's relation The Parseval's relation says that, IfF{x(0)} =X(j2) then x(t)}' dt = Proof: 2 Let x(t) be a continuous time signal and x") be conjugate of x(). Now, lxtI' = x) x*) On integrating the above equation with respect to twe get, (4.53) By definiticn of inverse Fourier transform, we can write, xt) =F{Xje} = 21 transform Fourier definitionof |Using .:X(i2)=2rx(12) xfi). x-2)= function x,(-i2) 2rx,l-j2) even For Note: :.X,lj2) 2n = = d dn x,t)e :x2)= 2 (4.55)equation |Using interchanging andt| is2 Replacingt -tby x1-)-X2)e +3 2 x,=(t) transform, Fourier inverse definition of T By 4.55) form. X,i2) F{x,))= similar in (-j2) functions similar functions below. are X,j2) and x,I) and X,(2) F{x,l1}= Let, let, Proof: jS2) X,( then 2rx, (j2) X, ’(t) x, If propertyis duality Alternatively shown expressed as are j2) rx,(2are (j2) X, 2rx,(-j2), i.c., similar are (j2) X,and x,(t) i.e., (G2), X,(j2) (t)}- (j) = X, then X,x,(t)= ifand and X,(j2) F{x, F{x,(0}= If Duality called isit soand x(). signal function s2 of energy term The density X(j2) energy spectrum or epresents the the density of spectral energy as distribution of 13. Not:e 2n S2-0 = 2r d2 XijQ) |X*i2) Xlj X*{j)= X(jS) transform. definition of 1 2n Fourier Using tegration. erchanging the 1 2n order of writlen as, beca(4.53) n equation the(4.54)equation dn )e .4.54) gei. equation we above x| Using 2n thconjugate e of x*(t) = faking On 14. Area under a time domain signal Area under x(t) =x() dt If x(t) and X(i2) are Fourier transfom pair, then,x(t) dt= X(0) where, X(0)= j2Lt’0 X((2) Proof: By definition of Fourier transform, Xi2) =xl|) e d :.X(0) = Xj2) = Lt x(0) e- dt = (x() e d=) dt i.x() dt = X{0) 15. Area under a frequency domain signal +o0 Area under X(j2) =X((2) d2 If x(t) and X(() are Fourier transform pair, then,XG2) dQ =2r x(0) where, x(0)= t ’ Lt0 x(t) Proof: By definition of inverse Fourier transform, x[t) = 1 2r .. x0)= Lt x(t) = LI 1-0 2 2 | X(jS2) d = 2r x(0) Summary of Properties of Fourier Transform Let, F{x()} = X((): Fx.)} = X.G); F{(t)} = ,0s Frequency domain signal Time domain signal Property a, X,() +a, X,G2) Linearity a, x,() +a,x,(0) Time shi x(t - ) Time scaling x(at) Time reversal x() Conjugation x() Frequency shifting X(9) X(j) X(-jN) XG(2-)) eo! x(t) x(t) Time differentiation jS2 XGL) Time integration x(t) dr Frequency differentiation t x(t) Xj) = X(0) S(9) 400 Time convolution (t)*x,(t)= x() X, (t- t)dt X,G2) X,i2) 0 1 Frequency convolution (or Multiplication) 2 x,() x,() X(j2)) - X(2) LXj2) = -LX-j2) Symmetry of real signals x(t) is real Re(X(j2)) Re(X(-j2)) Im(X((2)) - Im{X(-j)) Real and even x(t) is real and even X(i2) are real and even Real and odd x(t) is real and odd X(i2) are imaginary and odd Duality [ie., x,() and X,(2) are similar functions] then X,(j)= 2rx(jS2) [i.e.,X,() and 2rx,(j) are similar functions If x,() = X,G) +00 = 2r x(0) Area under a domain signalfrequency Area under a time domain signal x() t =X(o) Energy in frequency domain is, Energy in time domain is, +00 2T Parseval's relation 1 2T