Signal Processing and Linear Systems-B.P.Lathi copy

# c d etermine t he n umber o f binary d igitss bitss

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: l-order signals. Signals which d o n ot s atisfy t his c ondition m ay n ot b e L aplace t ransformable. 361 362 6 C ontinuous-Time S ystem Analysis Using t he Laplace Transform 6.1 T he Laplace Transform 363 4>(t) = f (t)e- ut as depicted in Fig. 6.1a. T he signal 4>(t) is now Fourier transformable a nd i ts Fourier components are o f t he form ejwt w ith frequencies w varying from w = - 00 t o 0 0. T he e xponential components ejwt a nd e - jwt in t he s pectrum a dd t o give a sinusoid of frequency w. T he s pectrum c ontains a n infinite number of such sinusoids, each having an infinitesimal amplitude. I t would be very confusing t o d raw all these components; hence, in Fig. 6.1b, we show j ust two typical components. Addition of all such components (infinite in number) results in 4>(t), i llustrated in Fig. 6.1a. T he e xponential spectral components of 4>(t) are of t he form ejwt w ith complex frequencies j w l ying on t he imaginary axis from w = - 00 t o 0 0, as shown in Fig. 6.1c. Figure 6.1a shows a signal 4>(t) = f (t)e- ut , Fig. 6.1b shows two of its infinite spectral components, a nd Fig. 6.1c shows t he location in t he complex plane of the frequencies of all t he s pectral components of 4>(t). Now, t he desired signal f (t) can be obtained by multiplying 4>(t) w ith eut . T his fact means t hat we c an synthesize f (t) by multiplying each spectral component of 4>(t) w ith eO't a nd a dding them. B ut m ultiplying t he s pectral c omponents of 4>(t) (sinusoids in Fig. 6.1b) with eut results in t he e xponentially growing sinusoids as shown in Fig. 6.1e. Addition of all such exponentially growing sinusoids (infinite in number) results in f (t) in Fig. 6.1d. T he s pectral components of 4>(t) a re of t he form ejwt . Multiplication of these components with eut results in t he s pectral components of the form eutejwt = e(u+jw)t. Therefore, a component of frequency j w in t he s pectrum of 4>(t) is transformed into a component of frequency a + j w in t he s pectrum of f (t). T he l ocation in the complex plane o f t he frequencies a + j w is along a vertical line, depicted in Fig. 6.1f. I t is clear t hat a signal f (t) c an be synthesized with growing everlasting exponentials lying along t he p ath a + j w, w ith w varying from - 00 t o 0 0. T he value of a is flexible. For example, if f (t) = e2t u(t), t hen 4>(t) = f (t)e- ut c an be made Fourier transformable if we choose a > 2. T hus, there are infinite choices for a . T his means t he s pectrum of f (t) is n ot unique, a nd t here are infinite possible ways of synthesizing f (t). Nevertheless, a has a certain minimum value ao for a given f (t) [ao = 2 for f (t) = e2t u(t)]. T he region in t he complex plane where a > ao (Fig. 6.1g) is c alled t he r egion o f c onvergence (or e xistence) o f t he resulting transform of f (t ). All these conclusions reached here by pure heuristic reasoning will now be derived analytically. T he frequency j w in t he Fourier transform will be generalized cP(t) f (t) = f ( t)e- at t-(a) (d...
View Full Document

## This note was uploaded on 04/14/2013 for the course ENG 350 taught by Professor Bayliss during the Spring '13 term at Northwestern.

Ask a homework question - tutors are online