Principles of Communication

Prof. V. Venkata Rao

1

CHAPTER 1

Representation of Signals

1.1 Introduction
The process of (electronic) communication involves the generation, transmission and reception of various types of signals. The communication process becomes fairly difficult, because: a) the transmitted signals may have to travel long distances (there by undergoing severe attenuation) before they can reach the destination i.e., the receiver. b) c) d) of imperfections of the channel over which the signals have to travel of interference due to other signals sharing the same channel and of noise at the receiver input1.

In quite a few situations, the desired signal strength at the receiver input may not be significantly stronger than the disturbance component present at that point in the communication chain. (But for the above causes, the process of communication would have been quite easy, if not trivial). In order to come up with appropriate signal processing techniques, which enable us to extract the desired signal from a distorted and noisy version of the transmitted signal, we must clearly understand the nature and properties of the desired and undesired signals present at various stages of a communication system. In this lesson, we begin our study of this aspect of communication theory.

1

Complete statistical characterization of the noise will be given in chapter 3, namely, Random

Signals and Noise.

1.1
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Signals physically exist in the time domain and are usually expressed as a function of the time parameter1. Because of this feature, it is not too difficult, at least in the majority of the situations of interest to us, to visualize the signal behavior in the Time Domain. In fact, it may even be possible to view the signals on an oscilloscope. But equally important is the characterization of the signals in the Frequency Domain or Spectral Domain. That is, we characterize the signal in terms of its various frequency components (or its spectrum). Fourier analysis (Fourier Series and Fourier Transform) helps us in arriving at the spectral description of the pertinent signals.

1.2 Periodic Signals and Fourier Series
Signals can be classified in various ways such as: a) b) c) d) Power or Energy Deterministic or Random Real or Complex Periodic or Aperiodic etc.

Our immediate concern is with periodic signals. In this section we shall develop the spectral description of these signals.

1.2.1 Periodic signals
Def. 1.1: A signal x p ( t ) is said to be periodic if

xp (t ) = xp (t + T ) ,
for all t and some T . ( denotes the end of definition, example, etc.)

(1.1)

1

We will not discuss the multi-dimensional signals such as picture signals, video signals, etc.

1.2
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Let T0 be the smallest value of T for which this is possible. We call T0 as the period of x p ( t ) .

Fig. 1.1 shows a few examples of periodic signals.

Fig. 1.1: Some examples of periodic signals

1.3
Indian Institute of Technology Madras

we have. where Re [ x ] denotes the real part of x . As the radian frequency. j 2 πf t namely.4 Indian Institute of Technology Madras . A e ( + ϕ) or A exp ⎡ ⎣ j ( 2 π f t + ϕ)⎤ ⎦ . ω (in units of radians/ sec). the complex exponential can also written as A e j ( ωt + ϕ) . e j ω t = cos ω t + j sin ω t . From Euler’s relation.2. is equal to 2 π f . we ⎣ ⎦ have cos ω t = e j ωt + ( e j ωt ) 2 j ωt ∗ ( ∗ denotes the complex conjugate) e j ωt + e − = 2 The term e − j ωt or e − j 2 πf t is referred to as the complex exponential at the negative frequency − ω (or − f ). ⎡ π ⎞⎤ ⎛ Fourier analysis uses cos ω t ⎢or sin ω t = cos⎜ ω t − ⎟ ⎥ in the represen2 ⎠⎦ ⎝ ⎣ tation of real signals. As cos ω t is the Re ⎡e j ω t ⎤ . 1. f (or ω ) and ϕ to denote the specific values of these parameters.2 Fourier series 1. where A : Amplitude (in Volts or Amperes) f : Cyclical frequency (in Hz) ϕ : Phase angle at t = 0 (either in radians or degrees) Both A and f are real and non-negative. V. Venkata Rao The basic building block of Fourier analysis is the complex exponential. We use subscripts on A .Principles of Communication Prof.

Principles of Communication Prof. Fourier series decomposes x p ( t ) in to DC.5 Indian Institute of Technology Madras . 1. Using Eq. the argument (or the angle). T0 is not defined.2(a) states that x p ( t ) . 1. fundamental and its higher harmonics. The coefficients { xn } are in general complex. we have. 1. in general. ∞ xp (t ) = n = −∞ ∑ xn e j 2 π n f0 t (1. hence xn = xn e j ϕn (1. V. Venkata Rao Let x p ( t ) be a periodic signal with period T0 . Eq. n = 1 results in the fundamental).2(a). where n is an integer (for n = 0 . Then f0 = 1 is called the T0 fundamental frequency and n f0 is called the n th harmonic. we have the DC component and for the DC singal.2(a) is referred to as the Exponential form of the Fourier series. 1. fundamental and its various higher harmonics. xp (t ) = n = −∞ ∞ j ( 2 π n f0 t + ϕn ) ∑ xn e Eq.2a) The coefficients x p ( t ) as { xn } constitute the Fourier series and are related to xn = 1 T0 ∫T xp (t ) e − j 2 π n f0 t dt (1. xn is the magnitude 1. namely. Most often.3) where xn denotes the magnitude of the complex number and ϕn . we ⎛ T T ⎞ use the interval ⎜ − 0 .2b) 0 where ∫ T0 denotes the integral over any one period of x p ( t ) .3 in Eq. is composed of the frequency components at DC. T0 ) . 0 ⎟ or 2⎠ ⎝ 2 ( 0 .

6 Indian Institute of Technology Madras . for a real periodic signal. It is important to note that the spectrum of a periodic signal exists only at discrete frequencies. x− n = x n ϕ. then Eq.4b) Properties of Eq. The plot of xn vs. n (or n f0 ) is called the phase spectrum.4a) (1.4 holds. ⋅ ⋅ ⋅ . 1. Venkata Rao of the component in x p ( t ) at frequency n f0 and ϕn . Let x p ( t ) be real. ± 1. and ϕn vs.4 are part of an if and only if (iff) relationship.4 holds and if Eq. then x− n = 1 T0 ∫T xp (t ) e j 2 π n f0 t dt 0 ∗ = xn That is. x 0 .Principles of Communication Prof. at n f0 . 1. let the only nonzero coefficients of a periodic signal be x ± 2 . π⎞ π⎞ ⎛ ⎛ x p ( t ) = 4 cos ⎜ 4 π f0 t − ⎟ + 6 cos ⎜ 2 π f0 t − ⎟ + 1 4⎠ 3⎠ ⎝ ⎝ which is a real signal. The above form of representing x p ( t ) . namely. etc. x0 is real and let j π 3 ∗ = x1 and x0 = 1. in terms of cosines is called the Trigonometric form of the Fourier series. we have the two symmetry properties. namely. n = 0. n (or n f0 ) is called the magnitude spectrum. That is. x± 1 . if x p ( t ) is real. 1. 1. Then. x0 x− 2 = 2 e j π 4 * =X 0 implies. This is because the complex exponentials at ( n f0 ) and − ( n f0 ) can be combined into a cosine term.n = − ϕn (1. As an example. then x p ( t ) has to be real. V. ± 2. its phase. π 3 π 4 ∗ = x2 and x − 1 = 3 e xp (t ) = 2 e j π 4 e − j 4 π f0 t + 3e j π 3 e − j 2 π f0 t + 1 + 3e − j e j 2 π f0 t + 2e − j e j 4 π f0 t Combining the appropriate terms results in.

1 For the unit amplitude rectangular pulse train shown in Fig. Venkata Rao We shall illustrate the calculation of the Fourier coefficients using the periodic rectangular pulse train (This example is to be found in almost all the textbooks on communication theory). let us compute the Fourier series coefficients.2: Periodic rectangular pulse train x p ( t ) has a period T0 = 4 milliseconds and is ON for half the period and OFF during the remaining half.2(b). Example 1. we have xn = 1 T0 T0 2 T − 0 2 T0 4 ∫ e− j 2π n f0 t dt = 1 T0 − 0 ∫ T e− j 2 π n f0 t dt 4 ⎟ 1 sin ⎜ ⎝ 2 ⎠ = T0 n π f0 ⎛ nπ⎞ 1. V.7 Indian Institute of Technology Madras . 1.2. Fig. 1.Principles of Communication Prof. 1. From Eq. The fundamental frequency f0 = 250 Hz.

Principles of Communication Prof. Fig. Venkata Rao ⎛nπ⎞ sin ⎜ ⎟ ⎝ 2 ⎠ = nπ As can be seen from the equation for xn . Hence ϕn is either zero or ± π for all n . 1.3.3: Magnitude and phase spectra for the x p (t) of example 1.8 Indian Institute of Technology Madras .1 From Fig. all the Fourier coefficients are real but could be bipolar (+ve or –ve). 1. we observe: 1.3 shows the plots of magnitude and phase spectrum. V. Fig. 1.

Venkata Rao i) x0 . is undefined as xn = 0 for these n . V.Principles of Communication Prof. with f0 = 250 Hz . This is indicated with a cross on the phase spectrum plot. λ along with a table of values are given in appendix A1.5) A plot of the sinc λ vs. ii) spectrum exists only at discrete frequencies.9 Indian Institute of Technology Madras .3(a) is the envelope of the magnitude spectrum. frequency is symmetric and the plot of ϕn vs. Such a spectrum is called the discrete spectrum (or line spectrum). namely.1. frequency is anti-symmetric. This is because x p ( t ) is real. at the end of the chapter. the average value is 1 T0 ∫ x p ( t ) dt . iv) the plot of xn vs. For any periodic 2 signal. f = n f0 . iii) the curve drawn with broken line in Fig. 1. ± 4 etc. v) ϕn at n = ± 2 . The envelope consists of several lobes and the maximum value of each lobe keeps decreasing with increase in frequency. One of the functions that is useful in the study of Fourier analysis is the sinc ( ) function defined by sin ( π λ ) πλ sinc λ = sinc ( λ ) = (1. 1. the average or the DC value of the pulse train is T0 2 T − 0 2 1 .

frequency range over which the spectrum can be measured etc.4.1. the Fourier coefficient of example 1. depend on that particular instrument).Principles of Communication Prof. V. which can be used to obtain the magnitude spectrum of periodic signals (frequency resolution.4: x p ( t ) of Exercise 1. 1. xn = 1 ⎛n⎞ sinc ⎜ ⎟ . ⎡x ⎤ The spectral plots 1 to 4 give the values of 20 log10 ⎢ n ⎥ for the ⎢ x1 ⎥ ⎣ ⎦ waveforms 1 to 4 respectively. as indicated by a spectrum analyzer. The units for the above quantities are in decibels (dB).1 can be written as. Venkata Rao In terms of sinc ( λ ) . 1. 2 ⎝2⎠ Exercise 1. show that xn = ⎜ ⎟ sinc ( n f0 τ ) ⎝ T0 ⎠ Fig.1 ⎛ τ ⎞ For the x p ( t ) of Fig .10 Indian Institute of Technology Madras . We have given below a set of four waveforms (output of a function generator) and their line spectra.1 Spectrum analyzer is an important laboratory instrument.

Venkata Rao 1.11 Indian Institute of Technology Madras .Principles of Communication Prof. Waveform 1 Spectral Plot 1 1. V.

V. Venkata Rao Waveform 1: A cosine signal (frequency 10 kHz). namely at ± 10 kHz.12 Indian Institute of Technology Madras . x − 1 and x 1 . Waveform 2 1. Hence the spectral plot has only two lines.Principles of Communication Prof. Also. x1 2. we have x−1 = x1 . namely. Comments on spectral plot 1: Waveform 1 has only two Fourier coefficients. and their values are 20 log10 x1 = 0 dB.

13 Indian Institute of Technology Madras .6 ) ⎤ 20 log10 ⎢ ⎥ ⎢ sin c ( 0.8 ) ⎤ 20 log10 ⎢ ⎥ ⎢ sin c ( 0.362 ⎡ sin c ( 0. V.2 ) ⎥ ⎦ = 20 log10 ( 0 ) = − ∞ iv) fourth Harmonic: v) fifth harmonic: because of the limitations of the instrument.1 msec. T0 = 0. Venkata Rao Waveform 2: Periodic square wave with τ 1 = .935 ⎦ = 20 log10 (0.504 ⎤ = 20 log10 ⎢ ⎥ ⎣ 0.924 dB iii) third harmonic: ⎡ sin c ( 0. we see a small spike at − 60 dB.2 ) ⎦ ⎥ ⎣ ⎡ 0. the values of other components can be calculated.809 = − 0.539) dB = − 5.2 ) ⎦ ⎥ ⎣ = − 12. T0 5 Comments on Spectral Plot 2: Values of various spectral components are: i) ii) fundamental: 0 dB second harmonic: ⎡ sin c ( 0.4 ) ⎤ 20 log10 ⎢ ⎥ ⎢ sin c ( 0.2 ) ⎦ ⎥ ⎣ = 20 log10 0. 1. Similarly.04 dB ⎡ sin c (1) ⎤ 20 log10 ⎢ ⎥ ⎢ ⎣ sin c ( 0.Principles of Communication Prof.

14 Indian Institute of Technology Madras . Venkata Rao 3.Principles of Communication Prof. V. Waveform 3 Values of Spectral Components: Exercise 1.

Principles of Communication Prof. Venkata Rao 4. Waveform 4 Values of Spectral Components: Exercise 1.15 Indian Institute of Technology Madras . V.

denoted by Pxp . However. the only spectral components that have nonzero magnitudes are at 990 Hz and 1010 Hz.1 and the periodic waveforms 1 to 4 all have fundamental as part of their spectra. as Px p = 1 T0 ∫ T0 xp (t ) d t 2 Parseval’s (Power) Theorem relates Pxp to xn as follows: Pxp = n = −∞ ∑ ∞ xn 2 (The proof of this relation is left as an exercise. Based on this. let x p ( t ) = cos ( 20 π t ) cos ( 2000 π t ) . V.16 Indian Institute of Technology Madras . ie. We now define its (normalized) average power. we should not surmise that every periodic signal must necessarily have a nonzero value for its fundamental. As a counter to the conjuncture. the first 99 spectral components (inclusive of DC) are zero! Let x p ( t ) be a periodic voltage waveform across a 1 Ω resistor or a current waveform flowing in a 1 Ω resistor.) If x p ( t ) consists of a single complex exponential. This is periodic with period 100 msec. Venkata Rao The Example 1. That is. 1.Principles of Communication Prof.

if x p ( − t ) = − x p ( t ) A periodic signal x p ( t ) has half-wave symmetry. 1.Principles of Communication Prof. we have the following special forms for the coefficients xn : x p ( t ) even: xn ’s are purely real and even with respect to n x p ( t ) odd: xn ’s are purely imaginary and odd with respect to n x p ( t ) half-wave symmetric: xn ’s are zero for n even.2(a): Def. if T ⎞ ⎛ xp ⎜ t ± 0 ⎟ = − xp (t ) 2⎠ ⎝ (1.6b) With respect to the symmetries defined by Eq.2(b): Def. Parseval’s power theorem implies that the total average power in x p ( t ) is superposition of the average powers of the complex exponentials present in it.6. A periodic signal x p ( t ) is even.17 Indian Institute of Technology Madras . Fourier coefficients take special forms. A proof of these properties is as follows: i) x p ( t ) even: 1. if x p ( − t ) = x p ( t ) A periodic signal x p ( t ) is odd. When the periodic signal exhibits certain symmetries. Venkata Rao j 2 π n f0 t + ϕ ) x p ( t ) = xn e ( 2 then. 1. V.6c) Def. Let us first define some of these symmetries (We assume x p ( t ) to be real).6a) (1. 1. 1.2(c): (1. Px p = xn In other words.

t . and noting x p ( − t ) = x p ( t ) .18 Indian Institute of Technology Madras .Principles of Communication Prof. ii) x p ( t ) odd: xn 0 1⎡ ⎢ ∫ xp ( t ) e− = T0 ⎢ − T 2 ⎣ 0 j 2 π n f0 t dt + T0 2 0 ∫ xp ( t ) e− j 2 π n f0 t ⎤ d t⎥ ⎥ ⎦ Changing t to .t in the first integral. we have T 2 T0 2 1⎡0 j 2 π n f0 t − ⎢ ∫ − xp (t ) e d t + ∫ xp (t ) e = T0 ⎢ 0 0 ⎣ j 2 π n f0 t ⎤ d t⎥ ⎥ ⎦ = 1 T0 T0 2 0 ∫ x p ( t ) ⎡e ⎣ − j 2 π n f0 t −e j 2 π n f0 t ⎤ ⎦ dt 2j = − T0 T0 2 0 ∫ x p ( t ) sin ( 2 π n f0 t ) d t Hence the result. V. Venkata Rao xn = 1 T0 T0 2 − T0 2 ∫ xp (t ) e − j 2 π n f0 t dt T0 2 0 = 0 1 ⎡ ⎢ ∫ xp ( t ) e− T0 ⎢ − T 2 ⎣ 0 j 2 π n f0 t dt + ∫ x p (t)e − j 2 π n f0 t ⎤ d t⎥ ⎥ ⎦ Changing t to . in the first integral and noting that x p ( − t ) = − x p ( t ) . iii) x p ( t ) has half-wave symmetry: 1. T 2 1⎡0 ⎢ ∫ x p ( t ) e j 2 π n f0 t d t + = T0 ⎢ 0 ⎣ T0 2 0 ∫ xp (t ) e − j 2 π n f0 t ⎤ d t⎥ ⎥ ⎦ T /2 ⎤ 2 ⎡0 ⎢ ∫ x p ( t ) ( cos 2 π n f0 t ) d t ⎥ = T0 ⎢ 0 ⎥ ⎣ ⎦ The above integral is real and as cos ⎡ ⎣2π ( − n ) f0 t ⎤ ⎦ = cos(2 π n f0 t ) . x −n = x n .

n even 1. an infinite summation. Let xM ( t ) = n = −M ∑ M xn e j 2 π n f0 t (1. ii) iii) x p ( t ) has only a finite number of maxima and minima over any period T0 . Venkata Rao xn 0 1⎡ ⎢ ∫ xp (t ) e− = T0 ⎢ −T / 2 ⎣ 0 j 2 π n f0 t dt + T0 / 2 ∫ 0 xp (t ) e− j 2 πn f0 t ⎤ d t⎥ ⎥ ⎦ In the first integral. As such.7) 1 For examples of periodic signals that do not satisfy one or more of the conditions i) to iii). 2] listed at the end of this chapter. known as Dirichlet conditions that guarantee convergence. xn T /2 T0 / 2 1 ⎡0 − j ( n ω0 t +π n ) ⎢ ∫ x p (t + T0 /2) e d t + ∫ x p ( t )e − = T0 ⎢ 0 0 ⎣ j n ω0 t ⎤ d t⎥ ⎥ ⎦ The result follows from the relation ⎧− 1. replace t by t + T0 /2 . There are only finite number of finite discontinuities over any period1.19 Indian Institute of Technology Madras . It is easy to verify that the above condition results in xn < ∞ for any n . 1. There is a set of conditions. 1. in general. the function is absolutely integrable over any period. V. the representation of a periodic function in terms of Fourier series involves.3 Convergence of Fourier series and Gibbs phenomenon As seen from Eq.Principles of Communication Prof. the reader is referred to [1.2. These are stated below. the issue of convergence of the series is to be given some consideration.2. i) T0 ∫ xp (t ) < ∞ That is. n odd e− j π n = ⎨ ⎩ 1.

but they occur at isolated points. The limiting behavior of xM ( t ) at the points of discontinuity in x p ( t ) is somewhat interesting. From the figure. 1. we shall have examples of Fourier series for functions that voilate some of the Dirichet conditions.20 Indian Institute of Technology Madras . M → ∞ lim T0 ∫ eM ( t ) 2 dt = 0 (1. T0 ∫ xp (t ) 2 dt < ∞ . 1.8) Note that Eq. M → ∞ Dirichlet conditions are sufficient but not necessary. resulting in the M →∞ integral of Eq. V.5(a). Later on. that is lim eM ( t ) = 0 for all t. we see that 1. If x p ( t ) is not absolutely integrable but square integrable.8 does not imply that lim eM ( t ) is zero. This is illustrated in Fig. That is. that is.8. 1. regardless of x p ( t ) being absolutely integrable or square integrable. then the series converges in the mean. Venkata Rao and eM ( t ) = x p ( t ) − xM ( t ) then M → ∞ lim xM ( t ) converges uniformly to x p ( t ) wherever x p ( t ) is continuous. There could be M →∞ nonzero values in lim eM ( t ) . being equal to zero.Principles of Communication Prof.

In the limiting case.5(b): Gibbs ears at t = t1 1. Venkata Rao Fig.5(b). all the oscillations converge in location to the point t = t1 (the point of discontinuity) resulting in what is called as Gibbs ears as shown in Fig. 09 A) persist even as M → ∞ . These oscillations (with the peak overshoot as 2M well as the undershoot of amplitude 0. 1.Principles of Communication Prof.5(a): Convergence behavior of xM ( t ) at a discontinuity in x p ( t ) xM ( t ) passes through the mid-point of the discontinuity and has a peak overshoot (as well as undershoot) of amplitude 0. 1. 1. t > 0 ) is T0 .09A (We assume M to be sufficiently large). Fig. V.21 Indian Institute of Technology Madras . The period of oscillations (whose amplitudes keep decreasing with increasing t .

A few of the aperiodic signals occur quote often in theoretical studies. When Michelson tried this instrument on signals with discontinuities (with continuous signals. Venkata Rao (In 1898. t > T ⎪ 2 ⎩ (1. Albert Michelson. which was capable of computing the first 80 coefficients of the Fourier series. it behooves us to introduce some notation to describe their behavior. close agreement was found between the original signal and the synthesized signal).22 Indian Institute of Technology Madras . J. Gibbs.Principles of Communication Prof. HA could also be used a signal synthesizer. In other words. ga ⎜ ⎟ ⎝T ⎠ T ⎧ 1. W. developed an instrument called Harmonic Analyzer (HA). he observed a strange behavior: synthesized signal.9a) 1. 1. V. professor at Yale. t < ⎪ ⎛t ⎞ ⎪ 2 ga ⎜ ⎟ = ⎨ ⎝ T ⎠ ⎪0. based on the 80 coefficients.) The convergence of the Fourier series and the corresponding Gibbs oscillations can be seen from the animation that follows. Hence. These are: M = 10. hence the name Gibbs Phenomenon. it has the ability to self-check its calculations by synthesizing the signal using the computed coefficients. 25 and 75 . This behavior persisted even after increasing the number of terms beyond 80.3 Aperiodic Signals and Fourier Transform Aperiodic (also called nonperiodic) signals can be of finite or infinite duration. in the vicinity of the discontinuity. You have been provided with three options with respect to the number of harmonics M to be summed. a well-known name in the field of optics. exhibited ringing with an overshoot of about 9% of the discontinuity. i) ⎛t ⎞ Rectangular pulse. investigated and clarified the above behavior by taking the saw-tooth wave as an example.

1. centered at t = 0 .9d) 1. t = 0 ex1 ⎜ ⎟ = ⎨ ⎝T⎠ ⎪ 2 ⎪ 0 . t < 0 ⎩ Fig. Venkata Rao [Rectangular pulse is sometimes referred to as a gate pulse. hence the symbol ga ( ) ]. t = 0 ⎝T ⎠ ⎪ t ⎪ eT . tri ⎜ ⎟ ⎝T ⎠ ⎛t tri ⎜ ⎝T ⎧ t ⎞ ⎪1− . ex1 ⎜ ⎟ : ⎝T ⎠ ⎧ −t ⎪e T . t < 0 ⎪ ⎩ iv) ⎛t ⎞ Two-sided (symmetrical) exponential pulse. t ≤ T T ⎟=⎨ ⎠ ⎪ ⎩ 0 .Principles of Communication Prof. (1.23 Indian Institute of Technology Madras . In view of the above ⎛t ⎞ A ga ⎜ ⎟ refers to a rectangular pulse ⎝T ⎠ of amplitude A and duration T . t > 0 ⎪ ⎛ t ⎞ ⎪ 1 . t > 0 t ⎛ ⎞ ⎪ ex 2 ⎜ ⎟ = ⎨ 1 .9b) iii) ⎛t ⎞ One-sided (decaying) exponential pulse. outside (1. ex 2 ⎜ ⎟ : ⎝T ⎠ ⎧ −t ⎪e T . ii) ⎛t ⎞ Triangular pulse.9c) (1. V.6 illustrates specific examples of these pulses.

2(a) expresses x p ( t ) as a sum over a discrete set of frequencies. Its counterpart for the aperiodic case is an integral relationship given by 1. the frequency components constituting a given signal x ( t ) lie in a continuous range (or ranges). 1. we use the symbol E without any subscript to denote the energy quantity. that is. and quite often this range could be ( −∞.1 Fourier transform Like periodic signals. (When no specific signal is being referred to.Principles of Communication Prof. V. we have a continuous spectrum for the aperiodic case.3. as ∞ Ex = −∞ 2 ∫ x ( t ) dt (1. However. 1. 1.6: Examples of pulses defined by Eq. ∞ ) . unlike the discrete spectrum of the periodic case.9 Let x ( t ) be any aperiodic signal. We define its normalized energy E x . Venkata Rao Fig.) 1. Eq.10) An aperiodic signal with 0 < E x < ∞ is said to be an energy signal. aperiodic signals also can be represented in the frequency domain.24 Indian Institute of Technology Madras .

Venkata Rao x (t ) = ∞ −∞ ∫ X (f ) e j 2 πf t df (1. 1.11(a) is X (f ) = −∞ ∞ − j 2 πf t dt ∫ x (t ) e (1.11(b) is referred to as the Fourier Transform (FT) relation or. 1. X ( f ) is. a complex quantity. we know x ( t ) and would want X ( f ) .25 Indian Institute of Technology Madras . X (f ) = XR (f ) + j XI (f ) (1.12a) (1. If we sum a large number of such complex exponentials.12b) ⎡ ⎣ X ( f )⎤ ⎦ Eq.Principles of Communication Prof. V. That is. or forward transform relation.11(a) is called the synthesis relation or Inverse Fourier Transform (IFT) relation. the analysis equation.12(a) and Eq. Let X ( fi ) ∆f be the incremental complex amplitude of e j 2πfi t at the frequency f = fi .11a) where X ( f ) is the Fourier transform of x ( t ) . leads to signal representation with a sum of complex exponentials replaced by an integral. Eq.11(a) is given the following interpretation. carried to its natural conclusion. where a continuous range of frequencies. We use the notation X (f ) = F ⎡ ⎣ x ( t )⎤ ⎦ x (t ) = F −1 (1. The companion relation to Eq. 1. in general. 1.11b) Eq.12c) 1. x ( t ) ←⎯ → X (f ) . Quite often. 1. Eq.12(b) are combined into the abbreviated notation. the resulting signal should be a very good approximation to x ( t ) . Let the integral be treated as a sum over incremental frequency ranges of width ∆f . with the appropriate complex amplitude distribution will synthesize the given signal x ( t ) . namely. This argument. 1.

Venkata Rao = X (f ) e j θ( f ) X R ( f ) = Real part of X ( f ) . 1. 1. ± 2 etc. V.1 contains the tabulated values of sin c ( λ ) . λ = 0 Fig.26 Indian Institute of Technology Madras . 2 where sin c ( λ ) = sin ( π λ ) πλ Appendix A1. Example 1. Let us compute and sketch X ( f ) . f . known as the phase spectrum.7(a) sketches the magnitude spectrum of the rectangular pulse for AT = 1 . Fig. A1. f .Principles of Communication Prof. λ = ± 1. X I ( f ) = Imaginary part of X ( f ) X ( f ) = magnitude of X ( f ) = 2 XR ( f ) + X I2 ( f ) ⎡ XI (f ) ⎤ θ ( f ) = arg ⎡ ⎣ X ( f )⎤ ⎦ = arc tan ⎢ X ( f ) ⎥ ⎢ ⎥ ⎣ R ⎦ Information in X ( f ) is usually displayed by means of two plots: (a) X ( f ) vs. Its behavior is shown in ⎧1 . Note that sin c ( λ ) = ⎨ ⎩0 . ⎠ T X (f ) = − ∫ T 2 Ae− j 2 πft dt = AT sin c (f T ) . known as magnitude spectrum and (b) θ ( f ) vs.2 ⎛t Let x ( t ) = A ga ⎜ ⎝T ⎞ ⎟ .1.

As the T T magnitude spectrum is always positive. 1.7. each with its own amplitude and phase.7: Spectrum of the rectangular pulse Regarding the phase plot. sin c ( f T ) is negative.7(b). But when we synthesize a signal using the complex exponentials with the magnitudes and phases as given in Fig. Remarkable balancing act: A serious look at the magnitude and phase plots reveals a very charming result. However it could be bipolar. Each of these complex exponentials exist for all t . m m +1 < f < . Venkata Rao Fig. they add up to 1.Principles of Communication Prof. we find that a rectangular pulse is composed of frequency components in the range −∞ < f < ∞ .27 Indian Institute of Technology Madras . 1. 1. as shown in Fig. V. From the magnitude spectrum. with m odd. sin c ( f T ) is real. negative values of sin c ( f T ) are taken care of by making θ ( f ) = ±1800 for the appropriate ranges. During the interval.

3 Let x ( t ) = ex1( t ) . This is called the phenomenon of reciprocal spreading. T Hence. V. Venkata Rao a constant for t < T indeed! 2 and then go to zero forever. this is true of other pulse types. Example 1. the range of strong spectral components) of the signal lies in the interval f < 1 . where T is the duration of the rectangular pulse. X ( f ) = ∫ e − t e − j 2πf t dt = 0 ∞ 1 1 + j 2πf Hence. the more wide-spread it would be in the frequency domain and vice versa. 1. That is. if T is reduced.Principles of Communication Prof. more compact is the signal in the time-domain.28 Indian Institute of Technology Madras . 1. A very fascinating result Fig. 1.8. (that is. we see that most of the energy. other than the rectangular). From the figure.7(a) illustrates another interesting result. Let us find X ( f ) and sketch it. X ( f ) = 1 1 + ( 2πf ) 2 θ ( f ) = − arc tan ( 2πf ) A plot of the magnitude and the phase spectrum are given in Fig. then its spectral width increases and vice versa (As we shall see later.

Principles of Communication

Prof. V. Venkata Rao

Fig. 1.8: (a) Magnitude spectrum of the decaying exponential (b) Phase spectrum

1.3.2 Dirichlet conditions
Given X ( f ) , Eq. 1.11(a) enables us to synthesize the signal x ( t ) . Now
s the question is: Is the synthesized signal, say x ( ) ( t ) , identical to x ( t ) ? This

leads to the topic of convergence of the Fourier Integral. Analogous to the Dirichlet conditions for the Fourier series, we have a set of sufficient conditions, (also called Dirichlet conditions) for the existence of Fourier transform, which are stated below:

1.29
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

(i)

x ( t ) be absolutely integrable; that is,

−∞

∫ x ( t ) dt < ∞

This ensures that X ( f ) is finite for all f , because
X (f ) =
∞ − j 2 πf t dt ∫ x (t ) e ∞

−∞

X (f ) =

−∞

− j 2 πf t dt ∫ x (t ) e

−∞

∫ x (t ) e

− j 2 πf t

dt =

−∞

∫ x ( t ) dt

<∞

(ii)

x ( t ) is single valued and has only finite number of maxima and minima

with in any finite interval. (iii)
x ( t ) has a finite number of finite discontinuities with in any finite interval.

If

x

(s )

( t ) = Wlim →∞ ∫

W

X ( f ) e j 2 πf t df , then

s x( ) (t )

converges to

x (t )

−W

uniformly wherever x ( t ) is continuous. If x ( t ) is not absolutely integrable but square integrable, that is,

−∞

∫ x (t )

2

dt < ∞ (Finite energy signal), then we have the convergence in the

mean, namely
∞ −∞

2 s x ( t ) − x ( ) ( t ) dt = 0

1.30
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Regardless of whether x ( t ) is absolutely integrable or square integrable,
s x ( ) ( t ) exhibits Gibbs phenomenon at the points of discontinuity in x ( t ) , always

passing through the midpoints of the discontinuities.

1.4 Properties of the Fourier Transform
Fourier Transform has a large number of properties, which are developed in the sequel. A thorough understanding of these properties, and the ability to make use of them appropriately, helps a great deal in the analysis of various signals and systems.

P1)

Linearity

Let x1 ( t ) ←⎯ → X1 ( f ) and x2 ( t ) ←⎯ → X2 (f ) . Then, for all constants a1 and a2 , we have a1 x1 ( t ) + a2 x2 ( t ) ←⎯ → a1 X1 ( f ) + a2 X 2 ( f ) It is very easy to see the validity of the above transform relationship. This property will be used quite often in the development of this course material.
P2a) Area under x ( t )

If x ( t ) ←⎯ → X ( f ) , then

−∞

∫ x ( t ) dt = X ( 0 )

The above property follows quite simply by setting f = 0 in Eq. 1.11(b). As an example of this property, we have the transform pair ⎛t ga ⎜ ⎝T ⎞ → T sin c ( fT ) ⎟ ←⎯ ⎠

By inspection, area of the time function is T , which is equal to T sin c ( f T ) | f = 0 .
P2b) Area under X ( f )

1.31
Indian Institute of Technology Madras

Proof: Exercise 1 ⎛f ⎞ X ⎜ ⎟ .9(b). we have ∞ −∞ 2πf ∫ X ( f ) df = ∫ 1 + ( 2πf )2 df 1 = 1 .9(a). which is the value of ex1( t ) | t = 0 2 P3) Time Scaling If x ( t ) ←⎯ → X ( f ) . x ( αt ) is a time reversed version of x ( t ) . we have x ( t ) = ex1( t ) ←⎯ → X (f ) = ∞ ∞ 1 1 + j 2πf ∞ 1 Hence ∫ X ( f ) df = ∫ df = 1 + j 2πf −∞ −∞ ∞ −∞ ∫ 1 + ( 2πf )2 df 1 − j 2πf Noting that ∞ −∞ −∞ ∫ 1 + ( 2πf )2 df = 0 . If α > 1 .Principles of Communication Prof. If α = − 1 . 1.32 Indian Institute of Technology Madras . then x ( αt ) ←⎯ → constant. As an illustration of this property.11(a). we have a time expanded version of x ( t ) . V. 1. where as if 0 < α < 1 . 1. x ( αt ) is a time compressed version of x ( t ) . Then x ( −2t ) would be as shown in Fig. Venkata Rao If x ( t ) ←⎯ → X ( f ) . Let x ( t ) be as shown in Fig. 1. where α is a real α ⎝α⎠ The value of α decides the behavior of x ( αt ) . then x ( 0 ) = ∞ −∞ ∫ X ( f ) df The proof follows by making t = 0 in Eq.

Principles of Communication

Prof. V. Venkata Rao

Fig. 1.9: A triangular pulse and its time compressed and reversed version α = − 1,

For

the

special

case

we

have

the

transform

pair

x ( − t ) ←⎯ → X ( − f ) . That is, both the time function and its Fourier transform undergo reversal. As an example, we know that ex1( t ) ←⎯ → 1 1 + j 2πf 1 1 − j 2πf

Hence ex1( −t ) ←⎯ →

ex1( t ) + ex1( − t ) = ex 2 ( t )

Using the linearity property of the Fourier transform, we obtain the transform pair ex 2 ( t ) = exp ( − t ) ←⎯ →
=

1 1 + 1 + j 2πf 1 − j 2πf

2 1 + ( 2πf )
2

Consider x ( αt ) with α = 2 . Then, x ( 2t ) ←⎯ → 1 ⎛f ⎞ X ⎟ 2 ⎜ ⎝ 2⎠

1.33
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Let us compare X ( f ) with
y ( t ) = ex1( 2t ) . That is,

1 ⎛f ⎞ X ⎜ ⎟ by taking x ( t ) = ex1( t ) , and 2 ⎝ 2⎠

⎧ e − 2t , t > 0 ⎪ ⎪ 1 , t =0 y (t ) = ⎨ ⎪ 2 ⎪ 0 , t <0 ⎩ Y (f ) = = ⎤ 1 ⎡ 1 ⎢ ⎥ 2 ⎣ ⎢ 1 + j 2π ( f 2 ) ⎥ ⎦ 1 2 + j 2πf

θ f Let X ( f ) = X ( f ) e x ( ) and

Y (f ) = Y (f ) e

θy ( f )

, where Y ( f ) =

1 2 1 + π2 f 2

θy ( f ) = − arc tan ( π f )

Fig. 1.10 gives the plots of X ( f ) and Y ( f ) . In Fig. 1.10(a), we have the plots X ( f ) vs. f and Y ( f ) vs. f . In Fig. 1.10(b), we have the plot of Y ( f ) normalized to have the maximum value of unity. This plot is denoted by YN ( f ) . Fig. 1.10(c) gives the plots of θ x ( f ) and θy ( f ) . From Fig. 1.10(b), we see that i)
y ( t ) = x ( 2t ) has a much wider spectral width as compared to the spectrum

⎛f ⎞ of the original signal. (In fact, if X ( f ) is band limited to W Hz, then X ⎜ ⎟ ⎝ 2⎠ will be band limited to 2W Hz.)

1.34
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Fig 1.10: Spectral plots of ex1( t ) and ex1( 2t ) 1.35
Indian Institute of Technology Madras

P4a) Time shift If x ( t ) ←⎯ → X ( f ) then. if the audio is compressed. This property of the FT will now be demonstrated with the help of a recorded audio signal.Principles of Communication Prof. then x ( t − t0 ) is a delayed version of x ( t ) and if t0 is negative. time expansion would result either in the loss or attenuation of higher frequency components. Then the value of E ( f ) is dependent on f . In any case. In other words. the remaining spectral components are distorted (both in amplitude and phase). higher frequency components (if the original signal is band limited) or making the latter part of the original spectrum much more significant. These higher frequency components will impart shrillness to the audio. Then x ( 2t ) will have a spectral components upto 20 kHz. V. the x ( t − t0 ) is an advanced version of x ( t ) .10(c). that is. Let x ( t ) represent an audio signal band limited to 10 kHz. besides distorting the original signal. time shifting 1. x ( t − t0 ) ←⎯ → e − j 2 π f t0 X ( f ) If t0 is positive. we have loss of “sharpness” in the resulting signal and severe distortion. Y ( f ) is a distorted version of X ( f ) . we observe that θ y ( f ) ≠ θ x ( f ) and their difference is a function of frequency. the original spectral magnitudes are modified by different amounts at different frequencies (Note that Y ( f ) ≠ k X ( f ) where k is a constant). In summary. Venkata Rao (ii) Let E ( f ) = X ( f ) − Y ( f ) . that is θy ( f ) is a distorted version of θ x ( f ) . and distortion of the remaining spectrum. time compression would result either in the introduction of new. 1. On the other hand. (iii) From Fig. Similarly.36 Indian Institute of Technology Madras .

Then. Fig. Proof: Exercise As an application of the above result. From the frequency shift theorem. then Y ( f ) will be as shown in Fig. F⎡ ⎣ x ( t − t0 ) ⎤ ⎦= ∞ − j 2 πf ( λ + t ) dλ ∫ x (λ) e 0 −∞ = e − j 2 π f t0 ∞ −∞ − j 2 πf λ dλ ∫ x (λ)e = e − j 2 π f t0 X ( f ) P4b) Frequency Shift If x ( t ) ←⎯ → X ( f ) . we have ⎣ ⎦ y ( t ) = 2 cos ( 2πfc t ) x ( t ) ←⎯ →Y ( f ) = X ( f − fc ) + X ( f + fc ) .11(a).37 Indian Institute of Technology Madras . 1.11: Illustration of modulation theorem 1. let us consider the spectrum of y ( t ) = 2 cos ( 2πfc t ) x ( t ) .Principles of Communication Prof. then e ± j 2 πfc t x ( t ) ←⎯ → X ( f ∓ fc ) where fc is a real constant. we want the Fourier transform of ⎡e j 2 πfc t + e − j 2 πfc t ⎤ x ( t ) .11(b) for fc = W . Proof: Let λ = ( t − t0 ) . This implies that x ( t ) and x ( t − t0 ) have the same magnitude spectrum.1. 1. If X ( f ) is as shown in Fig. (This property is also known as modulation theorem). V. that is. Venkata Rao will simply result in the multiplication of X ( f ) by a linear phase factor.

Duality theorem helps us in creating additional transform pairs. which is stated below.11a. namely.11a where as it is negative in Eq.1. the duality property. 1. ' f ' is the variable of integration where as in Eq. The proof of the second part of the property is similar. where the variable t is associated with a function denoted using the upper case letter and the variable f is associated with a function denoted using a lower case letter.11b. V.4 If z ( t ) = A sin c 2W t .38 Indian Institute of Technology Madras . 1. We shall illustrate the duality property with the help of a few examples. from the given set. Venkata Rao P5) Duality If we look fairly closely at the two equations constituting the Fourier transform pair. Both t and f are variables of the continuous type. 1. then X ( t ) ←⎯ → x ( − f ) and X ( − t ) ←⎯ → x (f ) Note: This is one instance. If x ( t ) ←⎯ → X ( f ) . we find that there is a great deal of similarity between them. 1. let us use duality to find Z ( f ) . it is the variable ' t ' .Principles of Communication Prof. This results in an interesting property. The sign of the exponent is positive in Eq. Example 1. Proof: x ( t ) = x (−t ) = ∞ −∞ j 2 πf t ∫ X ( f ) e df ∞ −∞ − j 2 πf t df ∫ X (f ) e The result follows by interchanging the variables t and f . In Eq.11b.

if we replace f by t . we have 1+ f 2 Z ( f ) = 2π x ( − f ) = 2π e − 2π f 1. Venkata Rao We look for a transform pair. ⎠ Example 1. V.39 Indian Institute of Technology Madras .Principles of Communication Prof. We know that if x ( t ) = A ⎛ t ga ⎜ 2W ⎝ 2W ⎞ ⎟ . ⎠ X ( f ) = A sin c ( 2W f ) and X ( t ) = A sin c ( 2W t ) = z ( t ) ⇒ Z (f ) = x ( − f ) = ⎛ f ⎞ A ga ⎜ − ⎟ 2W ⎝ 2W ⎠ As ga( − f ) = ga( f ) . then Y ( f ) = 2 1+( 2πf ) 2 ⎛ f ⎞ 1 Y⎜ ⎟ 2π ⎝ 2π ⎠ 1 2 2π 1 + f 2 2 1+ f 2 or 2π X ( f ) = As z ( t ) = 2 with f being replaced by t . we have ⎛ A ⎞ ⎛ f Z (f ) = ⎜ ⎟ ga ⎜ ⎝ 2W ⎠ ⎝ 2W ⎞ ⎟. x ( t ) ←⎯ → X ( f ) where in X ( f ) . Then X ( f ) = = − t . we have z ( t ) = A sin c 2W t . with α = 2π .5 Find the Fourier transform of z ( t ) = 2 . then. 1+ t 2 We know that if y ( t ) = e Let x ( t ) = y ( αt ) .

if x ( − t ) = x ∗ ( t ) . namely x ∗ ( − t ) ←⎯ → X ∗ (f ) Exercise 1. then x ( t ) is even if x ( − t ) = x ( t ) .2: Let x ( t ) = xR ( t ) + j xI ( t ) where xR ( t ) is the real part and xI ( t ) is the imaginary part of x ( t ) . From the time reversal property. Venkata Rao Hence 2 − ←⎯ → 2π e 2 1+ t 2π f P6) Conjugate functions If x ( t ) ←⎯ → X ∗ (−f ) → X ( f ) . If x ( t ) is real.40 Indian Institute of Technology Madras . then x ∗ ( t ) ←⎯ Proof: X ( f ) = X ∗ (f ) = ∞ − j 2 πf t dt ∫ x (t ) e ∞ −∞ −∞ ∫ x ∗ ( t ) e j 2 πf t dt ∞ X ∗ (−f ) = −∞ ∫ x ∗ ( t ) e − j 2 πf t dt Hence the result. 1. 1.Principles of Communication Prof. we get the additional relation. V.3(a): A signal x ( t ) is called conjugate symmetric. Show that xR ( t ) ←⎯ → xI ( t ) ←⎯ → 1 ⎡ X ( f ) + X ∗ ( − f )⎤ ⎦ 2 ⎣ 1 ⎡ X ( f ) − X ∗ ( − f )⎤ ⎦ 2j ⎣ Def.

V. that for real signals.S represent the convolution of X1 ( f ) and X 2 ( f ) . the spectrum for the negative frequencies is the complex conjugate of the positive part of the spectrum. 1.41 Indian Institute of Technology Madras . x1 ( t ) x2 ( t ) ←⎯ → ∞ −∞ ∫ X1 ( λ ) X 2 ( f − λ ) d λ = ∞ −∞ ∫ X 2 ( λ ) X1 ( f − λ ) d λ The integrals on the R.3(b): A signal x ( t ) is said to be conjugate anti-symmetric if x ( − t ) = − x∗ (t ) .H. then X ( f ) is also real and even. (Example: e − t ←⎯ → 2 1 + (2 π f ) 2 ). If x ( t ) is real. If x ( t ) is real. its transform is purely imaginary and odd (See Example 1. As a result. Hence. Venkata Rao Def. We denote the convolution of X1 ( f ) and X 2 ( f ) by X1 ( f ) ∗ X 2 ( f ) . X ( − f ) = X (f ) θ ( − f ) = − θ (f ) Going one step ahead. X ( f ) = X ∗ ( − f ) or X ∗ ( f ) = X ( − f ) . P7a) Multiplication in the time domain If x1 ( t ) ←⎯ → X1 ( f ) x2 ( t ) ←⎯ → X2 (f ) then. (Note that ∗ in between two functions represents the convolution of the two quantities where as a superscript. if x ( t ) is real and odd. it denotes the complex conjugate) 1. we can show that if x ( t ) is real and even. Similarly. This implies.Principles of Communication Prof.7). then x ( t ) is odd if x ( − t ) = − x ( t ) . then x ( t ) = x ∗ ( t ) .

x3 ( t ) = ∞ −∞ ∫ x1 ( λ ) x2 ( t − λ ) d λ ∞ −∞ ⎡ x3 ( t ) ⎦ ⎤= X3 (f ) = F ⎣ ∫ x3 ( t ) e − j 2 πf t ⎡∞ ⎤ dt = ∫ ⎢ ∫ x1 ( λ ) x2 ( t − λ ) d λ ⎥ e − ⎥ −∞ ⎢ ⎣− ∞ ⎦ ∞ j 2π f t dt Rearranging the integrals. V. F −1 ⎡ ⎣ X1 ( f ) X 2 ( f ) ⎤ ⎦= = ∞ −∞ ∞ ∫ x1 ( λ ) x2 ( t − λ ) d λ ∫ x2 ( λ ) x1 ( t − λ ) d λ −∞ As any one of the above integrals represent the convolution of x1 ( t ) and x2 ( t ) . 1. we have x1 ( t ) ∗ x2 ( t ) ←⎯ → X1 ( f ) X 2 ( f ) Proof: Let x3 ( t ) = x1 ( t ) ∗ x2 ( t ) That is. From the property P4(a).42 Indian Institute of Technology Madras . X3 (f ) ⎡∞ ⎤ = ∫ x1 ( λ ) ⎢ ∫ x2 ( t − λ ) e − j 2 π f t dt ⎥ d λ ⎢ ⎥ −∞ ⎣− ∞ ⎦ ∞ But the bracketed quantity is the Fourier transform of x2 ( t − λ ) .Principles of Communication Prof. we have x2 ( t − λ ) ←⎯ → e − j 2 πf λ X 2 ( f ) Hence. Venkata Rao Proof: Exercise P7b) Multiplication of Fourier transforms Let x1 ( t ) ←⎯ → X1 ( f ) x2 ( t ) ←⎯ → X2 (f ) then.

is one of the very useful properties of the Fourier transform. the input and output of a linear.invariant system are related by the convolution integral. convolution integral etc.Principles of Communication Prof. the student is advised to refer to [1 .6 ⎛t ⎞ In this example. Example 1. As will be shown later. we will find the Fourier transform of T tri ⎜ ⎟ . The concept of convolution is very basic in the theory of signals and systems. That is. ⎝T ⎠ ⎝T ⎠ ⎛t ⎞ ⎛t ga ⎜ ⎟ ∗ ga ⎜ ⎝T ⎠ ⎝T ⎞ ⎛t ⎞ ⎟ = T tri ⎜ T ⎟ ⎠ ⎝ ⎠ ⎛t ⎞ As ga ⎜ ⎟ ←⎯ → T sin c ( f T ) .3]. time. we have ⎝T ⎠ 2 ⎛t ⎞ T tri ⎜ ⎟ ←⎯ →⎡ T sin c ( f T ) ⎤ ⎣ ⎦ ⎝T ⎠ 1. known as the Convolution theorem.43 Indian Institute of Technology Madras . ⎝T ⎠ ⎛t ⎞ ⎛t ⎞ T tri ⎜ ⎟ can be obtained as the convolution of ga ⎜ ⎟ with itself. For a fairly detailed treatment of the properties of systems. Venkata Rao X3 (f ) = ∞ −∞ − j 2 πf λ dλ ∫ x1 ( λ ) X 2 ( f ) e ∞ = X2 (f ) −∞ − j 2 πf λ dλ ∫ x1 ( λ ) e = X1 ( f ) X 2 ( f ) Property P7(b). V.

Principles of Communication Prof. then. dt Example 1. d ⎡ x ( t )⎤ → j 2πf ⎡ ⎦ ←⎯ ⎣ X ( f )⎤ ⎦ dt ⎣ d n x (t ) dt n Generalizing. we see that F − 1 ⎡ ⎣ j 2πf X ( f ) ⎤ ⎦ is d x ( t ) . Venkata Rao P8) Differentiation P8a) Differentiation in the time domain Let x ( t ) ←⎯ → X (f ) .44 Indian Institute of Technology Madras . 1. x (t ) = ∞ j 2 πf t ∫ X ( f ) e df −∞ ∞ ⎤ d d ⎡ j 2 πf t ⎡ ⎤ ⎢ ⎥ = x t X f e df ( )⎦ ( ) dt ⎣ dt ⎢ −∫∞ ⎥ ⎣ ⎦ Interchanging the order of differentiation and integration on the RHS.12 below. 1.7 Let us find the FT of the doublet pulse x ( t ) shown in Fig. d ⎡ x ( t )⎤ ⎦= dt ⎣ = ∞ −∞ ∞ ⎡ d j 2 πf t ⎤ e ∫ X (f ) ⎢ ⎥ df ⎣ dt ⎦ j 2 πf t −∞ ⎣ j 2πf X ( f ) ⎤ ⎦e ∫⎡ df From the above. Hence the property. We have. ←⎯ → ( j 2πf ) X ( f ) n Proof: We shall prove the first part. V. generalization follows as a consequence this.

(using the result of example 1.Principles of Communication Prof.45 Indian Institute of Technology Madras . Hence.6) ( j 2πf )T 2 ( π fT )( π fT ) sin ( π fT ) sin2 ( π fT ) = j 2T sin ( π fT ) ( π fT ) = j 2T sin c ( π fT ) sin ( π fT ) As a consequence of property P8(a). ⎠⎦ ⎡ ⎛ t ⎞⎤ X ( f ) = j 2πf F ⎢T tri ⎜ ⎟ ⎥ ⎝ T ⎠⎦ ⎣ = = ( j 2πf )T 2 sin c 2 ( f T ) .12: x ( t ) of Example 1. V. Let x3 ( t ) = x1 ( t ) ∗ x2 ( t ) Then X 3 ( f ) = X1 ( f ) X 2 ( f ) ( j 2πf ) X 3 ( f ) = ⎡ ⎣( j 2πf ) X1 ( f ) ⎤ ⎦ X2 (f ) = X1 ( f ) ⎡ ⎣ j 2πf X 2 ( f ) ⎤ ⎦ That is. we have the following interesting result. 1. Venkata Rao Fig 1.7 x (t ) = d dt ⎡ ⎛t ⎢T tri ⎜ T ⎝ ⎣ ⎞⎤ ⎟ ⎥ .

Venkata Rao d d ⎡ ⎡ x1 ( t ) ⎤ x1 ( t ) ∗ x2 ( t ) ⎤ = ⎣ ⎦ ⎦ ∗ x2 ( t ) dt dt ⎣ = x1 ( t ) ∗ d ⎡ x2 ( t ) ⎤ ⎦ dt ⎣ P8b) Differentiation in the frequency domain Let x ( t ) ←⎯ → X (f ) . ex1( t ) ←⎯ → Hence. ⎠ We have.46 Indian Institute of Technology Madras . Then.Principles of Communication Prof.8 ⎛t Find the Fourier transform of x ( t ) = t ex1⎜ ⎝T ⎞ ⎟. ⎡ → ⎣( − j 2πt ) x ( t ) ⎤ ⎦ ←⎯ d X (f ) df d n X (f ) n ⎡ ⎤ → Generalizing. V. ( − j 2πt ) x ( t ) ←⎯ ⎣ ⎦ df n Proof: Exercise The generalized property can also be written as n ⎛ j ⎞ d X (f ) t x ( t ) ←⎯ →⎜ ⎟ df n ⎝ 2π ⎠ n n Example 1. 1 1 + j 2πf 1 ⎛t ⎞ ex1⎜ ⎟ ←⎯ →T 1 + j 2πfT ⎝T ⎠ ⎤ j d ⎡ T ⎛t ⎞ t ex1⎜ ⎟ ←⎯ → ⎢ 2π df ⎣ 1 + j 2πfT ⎥ ⎝T ⎠ ⎦ ←⎯ → j T 2π − ( j 2πT ) (1 + j 2πfT )2 1.

∞ −∞ ∫ ∗ x1 ( t ) x2 ( t ) dt = ∞ −∞ ∫ ∗ X1 ( f ) X 2 ( f ) df Proof: We have ∞ −∞ ∫ x1 ( t ) ∗ x2 ⎡∞ ⎤ ( t ) dt = ∫ x1 ( t ) ⎢ ∫ X 2 ( f ) e j 2πf t df ⎥ dt ⎢ ⎥ −∞ ⎣− ∞ ⎦ ∞ ∞ ∗ = −∞ ∞ ∫ ∫ ∗ X2 ⎡∞ ⎤ ( f ) ⎢ ∫ x1 ( t ) e − j 2πf t dt ⎥ df ⎢ ⎥ ⎣− ∞ ⎦ = ∗ X2 ( f ) X1 ( f ) df −∞ If x1 ( t ) = x2 ( t ) = x ( t ) . 1. Venkata Rao ⎡ ⎤ T2 ⎥ ←⎯ → ⎢ 2 ⎢ ⎥ 1 j 2 f T + π ( ) ⎣ ⎦ P9) Integration in time domain This property will be developed subsequently. energy of the signal x ( t ) . namely. P10) Rayleigh’s energy theorem This theorem states that. then ∞ −∞ ∫ x (t ) 2 ∞ dt = E x = −∞ ∫ X ( f ) df 2 Note: If x1 ( t ) and x2 ( t ) are real.Principles of Communication Prof. then. V. is ∞ Ex = −∞ ∫ X ( f ) df 2 This result follows from the more general relationship. E x .47 Indian Institute of Technology Madras .

∞ Ex = −∞ ∫⎡ ⎣2 AW sin c ( 2W t ) ⎤ ⎦ ⎞ ⎟ . X ( f1 ) df is the energy in the incremental frequency interval d f . From Example ⎛ f 1. Consider the rectangular pulse ga ⎜ ⎝T The first nulls of the magnitude spectrum occur at f = ± ⎞ ⎟. it would be easier to compute E x based on X ( f ) . this may be easier than computing the energy in the time domain. 2 W centered at f = f1 . (Some authors refer to this result as Parseval’s theorem) Example 1.9 E x . 2 That is. The evaluation of T 1.Principles of Communication Prof.4. Rayleigh’s energy theorem enables to treat X ( f ) as the energy spectral density of x ( t ) . Venkata Rao ∞ −∞ ∫ x1 ( t ) x2 ( t ) d t = ∞ −∞ ∫ X1 ( f ) X 2 ( − f ) d f = ∞ −∞ ∫ X 2 ( f ) X1 ( − f ) d f Property P10 enables us to compute the energy of a signal from its magnitude spectrum. In a few situations. ⎠ 1 .48 Indian Institute of Technology Madras . ⎠ 2 dt In this case. V.9 Let us find the energy of the signal x ( t ) = 2 AW sin c ( 2W t ) . Then. X ( f ) = A ga ⎜ ⎝ 2W W Ex = −W ∫ A2 df = 2W A2 More important than the calculation of the energy of the signal. Let −W ∫ X ( f ) df = 0. Hence. 90 percent of the energy of 2 ⎛t signal is confined to the interval f ≤ W .

5 Unified Approach to Fourier Transform So far. This would also enable us to have Fourier transforms for signals that do not satisfy one or more of the Dirichlet’s conditions (for the existence of the Fourier transform). f ( y ) ]. Venkata Rao 1 T − ∫ 1 X ( f ) df = 2 1 T − T ∫ 1 T sin c 2 ( f T ) df will yield the value 0. 1. The question arises: is it possible to unify these two approaches and talk only in terms of say. V. 1. A distribution is defined in terms of its effect on another function. It is a distribution or generalized function.Principles of Communication Prof. Hence. Fourier transform? The answer is yes provided we are willing to introduce Impulse Functions both in time and frequency domains.13a) (1. T ⎟ can ⎠ ⎝ ⎠ T ⎛t percent of the total energy of ga ⎜ ⎝T ⎛ 2 2⎞ be taken to be the spectral width of the rectangular pulse. The symbol δ ( t ) is fairly common in the technical literature to denote the unit impulse. We define the unit impulse as any (generalized) function that satisfies the following conditions: (i) (ii) (iii) δ ( t ) = 0.49 Indian Institute of Technology Madras . [The interval ⎜ − . which is 92 ⎞ ⎛ 1 1⎞ ⎟ . takes a number y and a produces another number. the frequency range ⎜ − T . t = 0 (1.92 T . ⎟ ⎝ T T⎠ may result in about 95 percent of the total energy].13b) Let p ( t ) be any ordinary function.5.1 Unit impulse (Dirac delta function) Impulse function is not a function in its strict sense [Note that a function f( ). t ≠ 0 δ ( t ) = ∞. then 1. we have represented the periodic functions by Fourier series and the aperiodic functions by Fourier transform.

namely.13. 1. A number of conventional functions have a limiting behavior that approaches δ ( t ) . 1.Principles of Communication Prof. As such δ ( t ) falls between a function and a transform (A transform operates on a function and produces a function). i = 1. 1. We cite a few such functions below: Let (a) p1 ( t ) = ε→0 1 ⎛t ⎞ ga ⎜ ⎟ ε ⎝ε⎠ (b) p2 ( t ) = 1 ⎛t⎞ tri ⎜ ⎟ ε ⎝ε⎠ (c) p3 ( t ) = 1 ⎛t⎞ sin c ⎜ ⎟ ε ⎝ε⎠ Then. 3 . p ( 0 ) . we see that δ ( t ) operates on a function such as p ( t ) and produces the number. lim pi ( t ) = δ ( t ) .13(c).13d) From Eq. ε > 0 (1.50 Indian Institute of Technology Madras . 2.13c) ( ε could be infinitesimally small) If p ( t ) = 1. Venkata Rao ∞ −∞ ∫ p ( t ) δ ( t ) dt = −ε ∫ p ( t ) δ ( t ) dt ε = p (0). then we have −ε ∫ ε δ (t ) d t = ∞ −∞ ∫ δ (t ) d t = 1 (1. for t ≤ ε . p3 ( t ) is shown below in Fig. V.

⎟ ⎜ Note that ε sin c ⎜ ε ⎟ ←⎯ ⎝ ⎠ ⎝ ⎠ From the above examples.13: sin c ( ) with the limiting behavior of δ ( t ) ⎛ ⎞ 1 ⎛t⎞ → ga ( ε f ) . we see that the shape of the function is not very critical.14) with the weight or area of the impulse shown in parentheses very close to it. 1. Venkata Rao Fig.Principles of Communication Prof. This is normally shown as a spear (Fig. V. we have A δ ( t − t0 ) . 1. 1. By delaying δ ( t ) by t0 and scaling it by A . Hence the area under the time function = 1.51 Indian Institute of Technology Madras . its area should remain at 1 in order to approach δ ( t ) in the limit.

every value of p ( t ) will be sampled and shifted by t0 resulting in p ( t − t0 ) . the values of p ( t ) . (Note: Some authors use this property as the operational definition of impulse function. Note that for the sampling property. V. that in the process of convolution. Then.52 Indian Institute of Technology Madras . Then for a < t0 < b .14: Symbol for A δ ( t − t0 ) Some properties of unit impulse P1) Sampling (or sifting) property Let p ( t ) be any ordinary function.) 1. Proof follows from making the change of variable t − t0 = τ and noting δ ( τ ) is zero for τ ≠ 0 . ∫ p ( t ) δ ( t − t0 ) dt a b = p ( t0 ) (This is generalization of condition (iii)). p ( t ) ∗ δ ( t − t0 ) = p ( t − t0 ) The proof of this property follows from the fact. P2) Replication property Let p ( t ) be any ordinary function. Venkata Rao Fig.Principles of Communication Prof. t ≠ t0 are of no consequence. 1.

that is. V.Principles of Communication Prof. Special Case: If α = − 1 . It is also possible to come up with delta functions as a limiting case of functions that are not even. Venkata Rao P3) Scaling Property δ ( αt ) = 1 δ (t ). as a limiting case of one-sided functions. which are all even functions of t . α > 0 ∞ −∞ ∫ δ ( αt ) dt = 1 1 δ ( τ) d τ = α α −∞ ∞ ∫ 1 = δ ( t ) dt α −∫ ∞ 1 = α ∞ ∞ −∞ ∫ δ ( t ) dt (ii) Let α < 0 . that is α = − α . ∞ −∞ ∫ δ ( αt ) dt = ∞ −∞ ∫ 1 1 δ ( τ) d τ = α α = 1 α ∞ −∞ ∫ δ ( t ) dt 1 It is easy to show that δ ⎡ ⎣ α ( t − t0 ) ⎤ ⎦ = α δ ( t − t0 ) . The above result is not surprising. In such a situation we have a left-sided delta function or right-sided delta function etc. Left-sided delta function will prove to be useful in the context of probability density functions of certain random variables. α ≠ 0 α Proof: (i) Let α t = τ. subject matter of chapter 2. 1. especially if we look at the examples p1 ( t ) to p3 ( t ) . and let − α t = τ .53 Indian Institute of Technology Madras . we have the result δ ( − t ) = δ ( t ) . Hence some authors call this as the even sided delta function.

(b) As the range of integration includes the impulse. F⎡ ⎣δ ( t ) ⎤ ⎦ = ∞ −∞ ∫ δ (t ) e − j 2 πf t dt = 1 (1.9 Example 1. V. 5.1 = 5 3 ∫ δ ( t − 5 ) dt = 125 4.11 ⎛t⎞ Let p ( t ) = tri ⎜ ⎟ . Hence the integral is zero. we have. ⎝4⎠ ⎡ ⎛ 1 ⎞⎤ 1⎡ ⎛ 1 ⎞⎤ δ ( 2t − 1) = δ ⎢2 ⎜ t − ⎟ ⎥ = δ ⎜ t − ⎟⎥ ⎢ 2 ⎠⎦ 2⎣ ⎝ 2 ⎠⎦ ⎣ ⎝ p (t ) ∗ 1 2 ⎡ ⎛ 1 ⎞⎤ 1 ⎛ 1⎞ 1 ⎛t − 1 2⎞ ⎢δ ⎜ t − 2 ⎟ ⎥ = 2 p ⎜ t − 2 ⎟ = 2 tri ⎜ 4 ⎟ ⎠⎦ ⎝ ⎠ ⎝ ⎠ ⎣ ⎝ Let us now compute the Fourier transform of δ ( t ) . we have a nonzero value for the product t 3 δ ( t − 5 ) .9 (a) δ ( t − 5 ) is nonzero only at t = 5 . The range of integration does not include the impulse. Hence.1 ∫ t δ ( t − 5 ) dt 3 5. 1.Principles of Communication Prof.54 Indian Institute of Technology Madras .1 (b) 3 ∫ t δ ( t − 5 ) dt 4. As δ ( t − 5 ) occurs at t = 5 .11(b).9 4.10 Find the value of (a) 3 ∫ t δ ( t − 5 ) dt 4 −4 5. Find ⎡ ⎣ p ( t ) ∗ δ ( 2t − 1) ⎤ ⎦. Venkata Rao Example 1. we can write t 3 δ ( t − 5 ) = 53 δ ( t − 5 ) . From Eq.14a) 1.

all with the same magnitude (unity in this case) and zero phase shift. we obtain another transform pair (from modulation theorem) e j 2 π f0 t ←⎯ → δ ( f − f0 ) e− j 2 π f0 t (1. its impulse response and the output).55 Indian Institute of Technology Madras . That is. a constant signal has no time variations and hence its spectral content ought to be confined to f = 0 . we make the following observation: a constant in one domain will transform into an impulse in the other domain. δ ( f ) is the proper quantity for the transform because it is zero for f ≠ 0 and its inverse transform yields the required constant in time (note that only an impulse can yield a nonzero value when integrated over zero width). a fascinating result indeed! Hence exciting any electric network or system with a unit impulse is equivalent to exciting the network simultaneously with complex exponentials of all possible frequencies. (We shall see a little later that if the network is linear and time invariant. The dual of the Fourier transform pair of Eq. all with unity magnitude and zero phase shift. 1. 1 ←⎯ → δ (f ) . 1. Because of the transform pair.15b) ←⎯ → δ ( f + f0 ) 1.14(a) gives us 1 ←⎯ → δ ( − f ) = δ (f ) (1. Venkata Rao How do we interpret this result? The spectrum of the unit impulse consists of frequency components in the range ( − ∞. Eq. 1. 1. V.Principles of Communication Prof. ∞ ) .14(a) and Eq.14(b).14b) Based on Eq.15a) (1. the unit impulse response of a linear network is the synthesis of responses to the individual complex exponentials and we intuitively feel that the impulse response of a network should be able to characterize the system in the time domain.14(b) is intuitively satisfying. a simple relation exists between the input to the network.

1. where the magnitude plot is identical to that shown in Fig. one for magnitude and the other for phase. 1. or else we can make two separate plots.56 Indian Institute of Technology Madras .15.17) F⎡ ⎣cos ω0 t ⎤ ⎦ and F [ sin ω0 t ] are shown in Fig.16) cos ω0 t ←⎯ → 1 ⎡δ ( f − f0 ) + δ ( f + f0 ) ⎤ ⎦ 2 ⎣ Similarly.15(b). Fig. Venkata Rao As cos ω0 t = e j 2 π f0 t + e − 2 j 2 π f0 t . 2 2 In summary.15(b) have weights that are complex.15: Fourier transforms of (a) cos ω0 t and (b) sin ω0 t Note that the impulses in Fig. we have developed the Fourier transforms of the periodic signals such 1. 1.Principles of Communication Prof. sin ω0 t ←⎯ → 1 ⎡δ ( f − f0 ) − δ ( f + f0 ) ⎦ ⎤ 2j ⎣ (1. 1. V. we have. 1. we have found the Fourier transform of δ ( t ) (a time function with a discontinuity that is not finite). (1. and using impulses in the frequency domain. It is fairly conventional to show the spectrum of sin ω0 t as depicted in Fig.15(a) and the phase plot has values of − π π at f = f0 and + at f = − f0 .

Let. we require impulses in the spectrum. V. xp (t ) = n = −∞ ∑ ∞ xn e j 2 π n f0 t Taking the Fourier transform on both the sides. ⎡ ∞ ⎤ j 2 π n f0 t ⎤ F⎡ x t X f F x e = = ( ) ( ) ⎢ ⎥ ∑ p p n ⎣ ⎦ ⎢ ⎥ ⎣n = − ∞ ⎦ = n = −∞ ∑ ∑ ∞ ∞ j 2 π n f0 t ⎤ xn F ⎡ ⎣e ⎦ = n = −∞ xn δ ( f − nf0 ) (1. 1. xn is a function of the discrete variable n . as X p ( f ) is purely impulsive.18 are only minor in nature. 1. spectral components exist only at f = n f0 . cos ω0 t and sin ω0 t . with complex weights xn . whereas. in the FS representation of x p ( t ) . 1. for a periodic signal x p ( t ) .Principles of Communication Prof.1 and spectral representation given by Eq. As such.18) FT of x p ( t ) is a function of the continuous variable f . the differences between the line spectrum of sec. We are now in a position to present both Fourier series and Fourier transform in a unified framework and talk only of Fourier transform whether the signal is aperiodic or not. Venkata Rao as e ± j 2 π f0 t . They both provide the same information. However. There is an interesting relation between xn and the Fourier transform of one period of a periodic signal. we have the Fourier series relation. which are neither absolutely integrable nor square integrable.57 Indian Institute of Technology Madras . As inversion of X p ( f ) requires integration. differing essentially only in notation. This is because.

⎩ T0 − T0 T < t < 0 2 2 outside j 2 π n f0 t xn = 1 T0 − ∫ T 0 2 xp (t ) e− dt 2 ⎡ T0 1 ⎢ 2 x (t ) e− = ∫ ⎢ T0 T0 ⎢ ⎣− 2 ∞ 1 ⎡ ⎢ ∫ x (t ) e− = T0 ⎢ − ∞ ⎣ j 2 π n f0 t ⎤ dt ⎥ ⎥ ⎥ ⎦ ∞ j 2 π n f0 t ⎤ 1 dt ⎥ = T0 ⎥ ⎦ −∞ ∫ x (t ) ⎛n⎞ − j 2 π⎜ ⎟t ⎝ T0 ⎠ e dt The bracketed quantity is X ( f ) ⎛n⎞ 1 X⎜ ⎟ T0 ⎝ T0 ⎠ f = n T0 Hence.16: Uniform impulse train 1.16 below. x (t ) = ⎨ p ⎪ 0 .58 Indian Institute of Technology Madras . Venkata Rao ⎧ ⎪x (t ) . 1.Principles of Communication Prof.19) Example 1.12 Find the Fourier transform of the uniform impulse train xp (t ) = n = −∞ ∑ δ ( t − nT0 ) ∞ shown in Fig 1. Fig. xn = (1. V.

we have another interesting result: A uniform periodic impulse train in either domain will transform into another uniform impulse train in the other domain. ∞ −∞ ∫e − j 2 πf t df = δ ( t ) That is. But as δ ( t ) ←⎯ → 1. 1.Principles of Communication Prof. V. T0 X p (f ) 1 = T0 n = −∞ ∑ δ (f ∞ − n f0 ) (1. from Eq. we have From the transform pair. 2 2 x ( t ) = δ ( t ) for this example. ˆ (t ) = Let x ∞ −∞ ∞ j 2 πf t ∫ X ( f ) e df = ⎡∞ x ( λ ) e− ∫⎢ ∫ −∞⎢ ⎣− ∞ ∞ ∞ j 2πf λ ⎤ d λ ⎥ e j 2 π f t df ⎥ ⎦ dλ = −∞ −∞ ∞ j 2 π f (t − λ ) df ∫ ∫ x (λ) e = ⎡ ∞ j 2 πf t − λ ⎤ ( ) ⎢ ⎥ dλ λ x e df ( ) ∫ ∫ ⎢ ⎥ −∞ ⎣− ∞ ⎦ 1.19) we have. we have. → 1.21) Using Eq. Venkata Rao Let x ( t ) be one period of x p ( t ) in the interval.21. xn = 1 for all n . δ ( t ) ←⎯ F −1 [1] ∞ = −∞ ∫e j 2 πf t df = δ ( t ) ∞ As δ ( − t ) = δ ( t ) .59 Indian Institute of Technology Madras . we show that x ( t ) and X ( f ) constitute a transform pair. Hence. 1.20) From Eq.20. (1. Then. − T0 T < t < 0 . −∞ ∫e ± j 2πf t df = δ ( t ) (1.

x (t ) = x (t ) ∗ δ (t ) = Let ∞ −∞ ∫ x ( τ) δ (t − τ) d τ R ⎡ ⎣δ ( t )⎤ ⎦ denote the output (response) of the LTI system.22) That is. Venkata Rao ∞ = −∞ ∫ x (λ ) δ (t − λ ) dλ = x (t ) ˆ ( t ) = x ( t ) . That is. V. From the replication property of the impulse. ⎡ x ( τ ) δ ( t − τ )⎦ ⎤ = x ( τ) h (t − τ) ⎣ As the system is linear.2 Impulse response and convolution Let x ( t ) be the input to a Linear. when δ ( t ) is input to an LTI system. We shall now establish a relation between x ( t ) and y ( t ) .Principles of Communication Prof. As the system is time R ⎡ ⎣δ ( t − τ )⎤ ⎦ = h (t − τ) .5. R ⎡ ⎣δ ( t )⎤ ⎦ = h ( t ) . y ( t ) . its output y ( t ) = invariant. Time-Invariant (LTI) system resulting in the output.60 Indian Institute of Technology Madras . y ( t ) = x ( t ) ∗ h ( t ) The following properties of convolution can be established: Convolution operation P1) P2) is commutative is associative 1. when the input is exited by the unit impulse δ ( t ) . and R R ⎡ x ( t )⎦ ⎤ = y (t ) = ⎣ R ⎢ ∫ x ( τ) δ (t ⎢ ⎣− ∞ ⎡ ∞ ⎤ − τ ) d τ⎥ = ⎥ ⎦ ∞ −∞ ∫ x ( τ) h (t − τ) dτ (1. As x 1. we have X ( f ) uniquely representing x ( t ) . This is generally denoted by the symbol h ( t ) and is called the impulse response of the system. we have.

P2) implies that ⎡ ⎣ x1 ( t ) ∗ x2 ( t ) ⎤ ⎦ ∗ x3 ( t ) = x1 ( t ) ∗ ⎡ ⎣ x 2 ( t ) ∗ x3 ( t ) ⎤ ⎦ . we have. complex. f . −∞ ∫ x1 ( τ ) x2 ( t − τ ) d τ = ∞ −∞ ∫ x2 ( τ ) x1 ( t − τ ) d τ . (As H ( f ) = Y (f ) X (f ) . As H ( f ) is. Venkata Rao P3) distributes over addition P1) implies that x1 ( t ) ∗ x2 ( t ) = x2 ( t ) ∗ x1 ( t ) ∞ That is. Note that the properties P1) to P3) are valid even if the independent variable is other than t .61 Indian Institute of Technology Madras . in general.23) → H ( f ) . we assume that every convolution pair gives rise to bounded output.22. the impulse response of any LTI system can be used to uniquely characterize the system in the time domain. −1 If H1 ( f ) ≠ H2 ( f ) . P3) implies that x1 ( t ) ∗ ⎡ ⎣ x 2 ( t ) + x3 ( t ) ⎤ ⎦ = x1 ( t ) ∗ x2 ( t ) + x1 ( t ) ∗ x3 ( t ) . we then have F − 1 ⎡ ⎣H1 ( f ) ⎤ ⎦ ≠ F ⎡ ⎣H2 ( f ) ⎤ ⎦ .Principles of Communication Prof. where the bracketed convolution is performed first. h1 ( t ) ≠ h2 ( t ) . Of course. Y (f ) = X (f ) H (f ) (1. namely. 1. 1. f and the phase response: arg ⎡ ⎣H ( f ) ⎤ ⎦ vs. the magnitude response: H ( f ) vs. In other words. V. The quantity H ( f ) is referred to (quite obviously) as where h ( t ) ←⎯ the frequency response of the system and describes the frequency domain behavior of the system. it is normally shown as two different plots. it is also referred to as the transfer function of the LTI system). Taking the Fourier transform on both sides of Eq. That is.

when x ( t ) = e j 2 π f0 t 1 + j 2 π f0 R C Generalizing this result. y ( t ) = e j 2 π f0 t 1 R+ j 2 π f0 C or y (t ) = 1 e j 2 π f0 t . H (f ) = That is. H (f ) = 1 1 + (2 π f R C ) 2 1 1 + j 2πf RC (1. 1.24) (1. Fig. This network is shown in Fig. Venkata Rao Example 1.62 Indian Institute of Technology Madras . Let us find its frequency response as well as the impulse response.Principles of Communication Prof.25b) θ ( f ) = arg ⎡ ⎣H ( f ) ⎤ ⎦ = − arc tan ( 2 π f R C ) 1. then the output y ( t ) is also a complex exponential given by y ( t ) = H ( f0 ) e j 2 π f0 t . we obtain.17: The RC-lowpass filter One of the important properties of any LTI system is: if the input x ( t ) = e j 2 π f0 t .17. V.25a) (1. 1 j 2 π f0 C But. 1.13 RC-lowpass filter (RC-LPF) is one among the quite often used LTI systems in the study of communication theory.

2πRC H (f ) = 1 2 ⎛f ⎞ 1+ ⎜ ⎟ ⎝F ⎠ (1. Venkata Rao Let 1 = F . f and arg ⎡ ⎣H ( f ) ⎤ ⎦ vs. V.63 Indian Institute of Technology Madras .Principles of Communication Prof. 1. Fig. f is shown in Fig. 1.18.26) A plot of H ( f ) vs.18 RC-LPF: (a) Magnitude response (b) Phase response 1. Then.

⎛ t ⎞ 1 ⎡ ⎣h ( t )⎤ ⎦ RC − LPF = R C ex1 ⎜ R C ⎟ ⎝ ⎠ This is shown in Fig. (1. We have the transform pair ex1( t ) ←⎯ → Hence. Venkata Rao Let us now compute h ( t ) = F − 1 ⎡ ⎣H ( f ) ⎤ ⎦ . 1.20. V. 1.19.19: Impulse response of an RC-LPF Example 1. That is. Let us find the output y ( t ) of the system. 1. 1.27) 1 1 + j 2πf ⎛ t ⎞ 1 1 → ex1⎜ ⎟ ←⎯ 1 + j 2πf RC RC ⎝ RC ⎠ Fig.64 Indian Institute of Technology Madras .Principles of Communication Prof.14 The input x ( t ) and the impulse response h ( t ) of an LTI system are as shown in Fig.

that is. x ⎡ ⎣ (t − τ )⎤ ⎦ is actually x ⎡ ⎣ − ( τ − t )⎤ ⎦ . where convolution is to be performed. These have been 1. the time instant for which y ( t ) is desired. This completes the operations in step (i).20: The input x ( t ) and the impulse response h ( t ) of an LTI system Let y ( t ) = ∞ −∞ ∫ h ( τ) x (t − τ) d τ To compute h ( t ) we have to perform the following three steps: i) ii) iii) Obtain h ( τ ) and x ( t − τ ) for a given t = t1 . we first reverse x ( τ ) to get x ( − τ ) and then shift by t . Take the product of the quantities in (i). h ( τ ) is the same as h ( t ) with the change of variable from t to τ . 1.65 Indian Institute of Technology Madras . V. Venkata Rao Fig. The operations involved in steps (ii) and (iii) are quite easy to understand. In quite a few situations.Principles of Communication Prof. Note that τ is the variable of integration. Integrate the result of (ii) to obtain y ( t1 ) . it would be of great help to have the plots of the quantities in step (i).

we see that if t < − 1 . namely t = 0 . t = − 1 and t = 2. For − 1 < t ≤ 0 . overlap of h ( τ ) and x ( t − τ ) increases as t increases and the integral of the product (which is positive) increases linearly reaching a value of 1 for t = 0 . 1. For 0 < t ≤ 2 .21 for three different values of t . net area of the product h ( τ ) x ( t − τ ) .1. then h ( τ ) and x ( t − τ ) do not overlap.Principles of Communication Prof. Venkata Rao shown in Fig. for t < − 1 . V. we have. keeps decreasing and at t = 2 . 1. that y ( t ) = 0 .66 Indian Institute of Technology Madras . From Fig.21(c). ∞ −∞ ∫ h ( τ) x (t − τ) d τ = 0.

(c). 1.21: A few plots to implement step (i) of convolution: (a): h ( τ ) (b).67 Indian Institute of Technology Madras . V.Principles of Communication Prof. t = − 1 and t = 2 respectively. 1. (d): x ( t − τ ) for t = 0 . Venkata Rao Fig.

otherwise ⎪ ⎩ ⎧ ⎪ h (t ) = ⎨ ⎪ ⎩ e − βt . could be very tricky and might even be sticky1. t ≥ 0 Let x ( t ) = ⎨ 0 .Principles of Communication Prof. t ≥ 0 0 . The result of the convolution is indicated in Fig. see [3].) Exercise 1. Venkata Rao Using the similar arguments. page 60). 1. For an interesting cartoon expressing the student reaction to the convolution operation. March 1991. Fig.14 (Sometimes. y ( t ) can be computed for t > 2 .68 Indian Institute of Technology Madras . otherwise where α. 1. V. computing y ( t ) = x ( t ) ∗ h ( t ) .22: Complete output y ( t ) of Example 1.22.1. Find y ( t ) = x ( t ) ∗ h ( t ) for (i) α = β and (ii) α ≠ β 1 Some people claim that convolution has driven many electrical engineering students to contemplate theology either for salvation or as an alternative career (IEEE Spectrum.3 ⎧ ⎪ e − αt . β > 0 .

Exercise 1. ⎧ 1. f2 = 2050 Hz and f3 = 20 Hz Verify that Y ( f1 ) = Y ( f2 ) = 2 and Y ( f3 ) = 1. 1. 1.3 Signum function and unit step function Def. Venkata Rao Exercise 1.5 Let X ( f ) ⎧1 ⎪5 ⎪ ⎪1 = ⎨ ⎪10 ⎪0 ⎪ ⎩ .Principles of Communication Prof.69 Indian Institute of Technology Madras . t > 0 ⎪ sgn ( t ) = ⎨ 0 . 950 < f < 1050 Hz . t < 0 ⎩ (1.4 Find y ( t ) = x ( t ) ∗ h ( t ) where ⎧ 2 . x (t ) = ⎨ ⎩ 0. 0 . 1. − 1050 < f < − 950 Hz . elsewhere (a) (b) Compute and sketch Y ( f ) = X ( f ) ∗ X ( f ) Let f1 = 50 Hz . V. 1.5.4: Signum Function We denote the signum function by sgn ( t ) and define it as.28) Def. ⎧ ⎪ h (t ) = ⎨ ⎪ ⎩ t < 2 outside t ≥ 0 outside 2 e− t . t = 0 ⎪ − 1. and define it as.5: Unit Step Function We denote the unit step function by u ( t ) .

23: sgn ( t ) as a limiting case of x ( t ) of Eq. t > 0 ⎪1 ⎪ u (t ) = ⎨ . we have 2 ⎣ 1. t < 0 We shall now develop the Fourier transforms of sgn ( t ) and u ( t ) . Fig. 1.29) Let x ( t ) = e − α t u ( t ) − e α t u ( − t ) α→0 (1.30 X (f ) = 1 1 − j 4 πf − = 2 α + j 2πf α − j 2πf α2 + ( 2 π f ) α→0 F⎡ ⎣sgn ( t ) ⎤ ⎦ = lim X ( f ) = 1 j πf (1. V.Principles of Communication Prof. 1.23.30) where α is a positive constant. 1. Then sgn ( t ) = lim x ( t ) . as can be seen from Fig.70 Indian Institute of Technology Madras .31) F⎡ ⎣u ( t ) ⎤ ⎦ As u ( t ) = 1 ⎡1 + sgn ( t ) ⎤ ⎦ . Venkata Rao ⎧1 . t = 0 ⎪2 ⎪ ⎩0 . F⎡ ⎣ sgn ( t ) ⎤ ⎦: (1.

then −∞ t ∫ x ( τ) d τ ∫ x ( τ) d τ t ←⎯ → (ii) ←⎯ → −∞ 2 Let p 1 ( t ) = 1 ⎛t ⎞ ga ⎜ ⎟ .32) We shall now state and prove the FT of the integral of a function. Then. Hence.33) Consider x ( t ) ∗ u ( t ) = x (t ) ∗ u (t ) = t t ∞ −∞ ∫ x ( τ) u (t − τ) d τ . Properties of FT continued. ⎡ 1 δ (f ) ⎤ τ τ ←⎯ → + x d X f ( ) ( ) ⎢ ⎥ ∫ 2 ⎦ ⎣ j 2πf −∞ X (f ) j 2πf X (f ) j 2 πf + X (0) δ (f ) Here there are two possibilities: (i) X ( 0 ) = 0 . ε→0 ε ⎝ε⎠ p1 ( t ) dt = 2 But lim ε→0 − ∫ε 0 −∞ ∫ p1 ( t ) dt 0 = 1 2 1.Principles of Communication Prof.. As u ( t − τ ) = 0 for τ > t . we know that lim p1 ( t ) = δ ( t ) . Venkata Rao U (f ) = 1 1 δ (f ) + 2 j 2πf (1. −∞ ∫ x ( τ ) d τ.71 Indian Institute of Technology Madras .. then X ( 0 ) ≠ 0 . But F⎡ ⎣ x ( t ) ∗ u ( t )⎤ ⎦ = X (f ) U (f ) . Proof: −∞ ∫ x ( τ) d τ t ←⎯ → 1 j 2πf X (f ) + X (0) 2 δ (f ) (1. V. P9) Integration in the time domain → X (f ) Let x ( t ) ←⎯ Then.

−∞ ∫ δ ( τ) d τ dt t = u (t ) (1. Hence. dt j 2 π f U (f ) = 1 or U (f ) = 1 j 2πf But this is valid only for f ≠ 0 because of the following argument.34b) We shall now give an alternative proof for the FT relation.72 Indian Institute of Technology Madras . f = 0 ⎪ ⎪2 = ⎨ ⎪ 1 . ε→0 − ∫ε 2 p1 ( t ) dt = ∞ 2 −∞ ∫ p1 ( t ) dt = 1. j 2πf 2 As d u (t ) = δ (t ) . V. 2 U (f ) ⎧1 δ (f ) .Principles of Communication Prof. u ( t ) ←⎯ → 1 1 δ (f ) + .34a) or d u (t ) = δ (t ) (1. U (f ) + U ( − f ) = δ (f ) As δ ( f ) is nonzero only for f = 0 . Venkata Rao ε and lim That is. u ( t ) + u ( − t ) = 1. we have U ( 0 ) + U ( − 0 ) = 2 U ( 0 ) = δ ( f ) or U (0) = 1 δ ( f ) . Therefore. f ≠ 0 ⎪ ⎩ j 2 πf 1.

24: Block schematic of a ZOH h ( t ) of ZOH: a) When x ( t ) = δ ( t ) .Principles of Communication Prof. Fig. ZOH). (b) If two such systems are cascaded.24. Venkata Rao Example 1. what is the overall impulse response (cascade of two ZOHs is called a First-Order-Hold. Hence. we have v (t ) = δ (t ) − δ (t − T ) Hence y ( t ) = ⎡ ⎣h ( t )⎤ ⎦ ZOH .15 (a) For the scheme shown in Fig. is −∞ ∫ δ ( τ) d τ − ∫ δ ( τ − T ) d τ −∞ t t = u (t ) − u (t − T ) T⎞ ⎛ ⎜t − 2 ⎟ = ga ⎜ ⎟ ⎜ T ⎟ ⎝ ⎠ b) Impulse response of two LTI systems in cascade is the convolution of the impulse responses of the constituents. ⎡ ⎣h ( t )⎤ ⎦ FOH T⎤ T⎤ ⎡ ⎡ ⎢t − 2 ⎥ ⎢t − 2 ⎥ = ga ⎢ ⎥ ∗ ga ⎢ ⎥ T ⎢ ⎥ ⎢ T ⎥ ⎣ ⎦ ⎣ ⎦ ⎛t −T ⎞ = T tri ⎜ ⎟ ⎝ T ⎠ 1. the impulse response of the ZOH. V. 1. 1.73 Indian Institute of Technology Madras . FOH). find the impulse response (This system is referred to as Zero-Order-Hold.

34(b) helps in finding the derivatives of signals with discontinuities. That is. Venkata Rao Eq.74 Indian Institute of Technology Madras . 1. V. 1. −∞ t ∫ δ ( τ) d τ ∫ δ ( τ) d τ t ←⎯ → 1 1 + δ ( f ) = U ( f ) . Consider the pulse p ( t ) shown in Fig. with x ( t ) = δ ( t ) and X ( f ) = 1 .25(a).25: (a) A signal with discontinuities (b) Derivative of the signal at (a) p ( t ) can be written as p ( t ) = u ( t + 2 ) − 2 u ( t + 1) + u ( t − 3 ) 1. Fig. 1.34(a) can also be established by working in the frequency domain.33. 1.Principles of Communication Prof. From Eq. 1. 2 j 2πf = u (t ) −∞ Eq.

Example 1. V. x ( t ) = t e − 2 t u ( t ) and h ( t ) = e − 4 t u ( t ) .12). Find a) Y ( f ) 1. From this result.17 Let x ( t ) .16 Let x ( t ) be the doublet pulse of Example 1. its derivative will have an impulse of weight A at t = t1 . h ( t ) and y ( t ) denote the input.75 Indian Institute of Technology Madras .25(b).1. we note that if there is a step discontinuity of size A at t = t1 in the signal. d x (t ) dt = δ (t + T ) − 2 δ (t ) + δ (t − T ) Taking Fourier transform on both the sides.7 (Fig. Venkata Rao Hence. d p (t ) dt = δ ( t + 2 ) − 2 δ ( t + 1) + δ ( t − 3 ) which is shown in Fig. It is given that.Principles of Communication Prof. 1. j 2 π f X ( f ) = e j 2 π f T − 2 + e− = e j π f T − e− X (f ) = 2 j T j πf T j 2 πf T ( (e j πf T − e− j πf T ) ) (e 2 j πf T − e− j πf T ) 2 j πf T 2j = 2 j T sin c ( f T ) sin ( π f T ) Example 1. impulse response and the output respectively of an LTI system. We shall find X (f ) from d x (t ) dt .

we obtain Y (f ) = X (f ) H (f ) If z ( t ) = e − 2 t u ( t ) . V.22. As u ( t − τ ) = 0 for τ > t .Principles of Communication Prof. 1. u ( t − τ ) is 1 only for τ negative. Y (f ) = ( − 14 ) 2 + j 2πf + 1 1 2 4 + 2 4 2πf j + (2 + j 2 π f ) 1 1 − 4t ⎛ 1⎞ e u (t ) Hence. 2 + j 2πf ⎛ j ⎞ d Z (f ) 1 As x ( t ) = t z ( t ) .76 Indian Institute of Technology Madras . we have X ( f ) = ⎜ = ⎟ ⎝ 2 π ⎠ df ( 2 + j 2 π f )2 H (f ) = 1 4 + j 2πf 1 Hence. we have y (t ) = ∫τ e 0 t − 2τ e − 4( t − τ ) dτ 1. Y ( f ) = (2 + j 2πf ) 2 1 4 + j 2πf (b) Using partial fraction expansion. but then u ( τ ) = 0 . then Z ( f ) = 1 . Venkata Rao b) y ( t ) = F − 1 ⎡ ⎣Y ( f ) ⎤ ⎦ c) y ( t ) = x ( t ) ∗ h ( t ) a) From Eq. y ( t ) = ⎜ − ⎟ e − 2 t u ( t ) + t e − 2 t u ( t ) + 2 4 ⎝ 4⎠ ∞ (c) y ( t ) = −∞ ∞ ∫ x ( τ) h (t − τ) d τ ∫ τ e − 2τ u ( τ ) e − 4( t − τ ) = u (t − τ) d τ −∞ y ( t ) = 0 for t < 0 because for t negative.

t < T0 2 T t > 0 2 where T0 is the period of the cosine signal and f0 = (a) Show that d 3 x (t ) d t3 1 . T0 d x (t ) ⎫ T ⎞ T ⎞ ⎛ 2 ⎧ ⎪ ⎛ ⎪ = ( 2 π f0 ) ⎨δ ⎜ t + 0 ⎟ − δ ⎜ t − 0 ⎟ − ⎬ 2⎠ 2⎠ dt ⎪ ⎝ ⎪ ⎩ ⎝ ⎭ (b) Taking the FT of the equation in (a) above. show that X (f ) = f0 sin c ( f T0 ) f0 − f 2 2 1.6 ⎧ 1 + cos ( 2 π f0 t ) ⎪ ⎪ Let x ( t ) = ⎨ ⎪ 0 ⎪ ⎩ .Principles of Communication Prof. V. t ≥ 0 4 ⎣ ⎦ = 0 for t < 0 Exercise 1. . Venkata Rao = e− 4 t ∫τ e 0 t 2τ dτ ⎤ e2 τ τ d ∫ 2 ⎥ ⎦0 t = e − 4t ⎡ e2 τ − ⎢τ ⎣ 2 = e − 4t t ⎫ ⎧ ⎡ e2 τ ⎤ ⎪ ⎪ e2 t − ⎢ ⎨t ⎥ ⎬ 2 4 ⎦0 ⎪ ⎣ ⎪ ⎩ ⎭ = t e− 2t − e− 4 t 2 ⎡ e2 t − 1⎤ ⎢ ⎥.77 Indian Institute of Technology Madras .

Rx y ( τ ) and Ry x ( τ ) . autocorrelation function. Our interest is in electrical signals. 1.6(b): The cross-correlation function Ry x ( τ ) is given by 1. let us develop the required familiarity with the correlation operation. We would like to quantify this intuitive notion and come up with a formal definition for correlation so that we have a mathematically consistent and physically meaningful measure for the correlation of the objects of interest to us. In the context of correlation functions. say. the similarity between a transmitted signal and the received signal or between two different transmitted or received signals etc.6(a): The cross-correlation function Rx y ( τ ) is given by ∞ ∗ ∫ x (t ) y (t − τ) d t Rx y ( τ ) = (1. 1.6 Correlation Functions Two basic operations that arise in the study of communication theory are: (i) convolution and (ii) correlation. When we say that there is some correlation between two objects.35a) −∞ Def. We shall first introduce the cross correlation functions. we have to distinguish between the energy signals and power signals.Principles of Communication Prof. As we have some feel for the convolution operation by now. namely. Def. we make the following definitions. 1. this will be followed by the special case. Venkata Rao 1.1 Cross-correlation functions (CCF) Let x ( t ) and y ( t ) be signals of the energy type. Accordingly. We may like to quantify. we imply that there is some similarity between them.78 Indian Institute of Technology Madras . We now define their cross-correlation functions. V.6.

Rx y ( τ ) is given by T Rx y 1 ( τ ) = Tlim →∞ T − x ( t ) y ∗ ( t − τ ) dt ∫ T 2 2 (1. y ∗ ( t − τ ) is a conjugated and shifted version of y ( t ) .35(a). 1. hence Rx y ( ) as well as Ry x ( ) is a function of τ .36b) The power signals that we have to deal with most often are of the periodic variety. the shift parameter ( τ is also called the scanning parameter or the search parameter).35(a). Venkata Rao Ry x ( τ ) = ∞ −∞ ∗ ∫ y (t ) x (t − τ) d t (1.Principles of Communication Prof. Def. τ accounting for the shift in y ∗ ( t ) .37a) 0 2 T0 Def.7(a): The cross-correlation function.79 Indian Institute of Technology Madras . 1. 1.7(b): The cross-correlation function. we have the following definitions: T0 Def.35b) In Eq. Rx y ( τ ) = ∞ ∗ ∫ x (λ + τ) y (λ ) dλ −∞ ∗ = Ry x ( − τ) (1. 1.36a) Def. Hence.35 is t .38) 1. 1. V.37b) 0 2 Let t − τ = λ in Eq. Note that the variable of integration in Eq. t = τ + λ and dt = d λ . Let x ( t ) and y ( t ) be the signals of the power type. Ry x ( τ ) is given by T 1 Ry x ( τ ) = lim T →∞ T − y ( t ) x ∗ ( t − τ ) dt ∫ T 2 2 (1.8(a): 1 Rx p y p ( τ ) = T0 − ∫ T 2 x p ( t ) y p∗ ( t − τ ) dt (1. Then.8(b): 1 Ry p x p ( τ ) = T0 − ∫ T 2 y p ( t ) x p∗ ( t − τ ) dt (1. 1. 1. For periodic signals.

1. commutative. we have nonzero value for the 2 2. cross-correlation. consider the situation shown in Fig.Principles of Communication Prof. V. unlike convolution is not in general. For the value of τ 1 . we will have a positive value for 1. To understand the significance of the parameter τ . However.26: Waveforms used to compute Ry x ( τ ) and Rz x ( τ ) ∞ If we compute −∞ ∫ x ( t ) y ( t ) dt . we find that x ⎜ t − ⎟ 2⎠ ⎝ and y ( t ) start overlapping and for τ > integral. Fig. we find it to be zero as x ( t ) and y ( t ) do not 1⎞ ⎛ overlap.26. if we delay x ( t ) by half a unit of time. Venkata Rao As Rx y ( τ ) ≠ Ry x ( τ ) .75 . 1.80 Indian Institute of Technology Madras .

involves the following steps: (i) Shift y ∗ ( t ) by τ (ii) Take the product of x ( t ) and y ∗ ( t − τ ) (iii) Integrate the product with respect to t .Principles of Communication Prof. It is not difficult to see that Rx y ( τ ) = x ( τ ) ∗ y ∗ ( − τ ) . it would be difficult for us to accept that x ( t ) could have been the transmitted signal. Thus the parameter τ helps us to find time-shifted similarities present between the two signals.75 .39a) Let E x y ( f ) = F ⎡ ⎣R x y ( τ ) ⎤ ⎦. we see that computing Rx y ( τ ) for a given τ . V.39b) 1. From Eq. we find that Rz x ( τ ) than Ry x ( τ ) max max would be much smaller (Note that for τ 2. Venkata Rao ∞ −∞ ∫ y ( t ) x ( t − τ ) dt . ∗ ∗ ⎤ Then. If y ( t ) is the received signal of a communication system. y ( t ) x ( t − τ ) . Similarly. which would be about the maximum of Ry x ( τ ) for any τ . because x ( τ) ∗ y ∗ ( − τ) ∞ = −∞ ∞ ∗ ⎣− ( τ − t )⎤ ⎦ dt ∫ x (t ) y ⎡ = −∞ ∗ ∫ x ( t ) y ( t − τ ) dt = Rx y ( τ ) (1. becoming zero for τ > 5 . is essentially positive for all t ). then we are willing to accept x ( t ) as the likely transmitted signal (we can treat the received signal as a delayed and distorted version of the transmitted signal) whereas if z ( t ) is received.75 . if we compute Rz x ( τ ) .81 Indian Institute of Technology Madras . Ry x ( τ ) keeps decreasing.35(a). E x y ( f ) = F ⎡ ⎣ x ( τ ) ∗ y ( − τ )⎦ = X ( f ) Y ( f ) (1. the product quantity. The above steps closely resemble the operations involved in convolution. 1. For values of τ > 2.

18 (i) τ < − 1: x ( t ) and y ( t − τ ) do not overlap and the product is zero. Rx y ( τ ) = 0 for τ < − 1. ii) − 1 ≤ τ < 1: Rx y ( τ ) = 1+ τ ∫ 0 − 1+ τ e − t dt = 1 − e ( ) 1. From the results of (a) and (b) above.82 Indian Institute of Technology Madras . V. (a) R x y ( τ ) : x ( t ) and y ( t ) are sketched below (Fig. 1. Venkata Rao Example 1.38. 1.Principles of Communication Prof.27).18 ⎛t⎞ Let x ( t ) = exp1 ( t ) and y ( t ) = ga ⎜ ⎟ . That is. 1. Fig.27: Waveforms of Example 1. ⎝ 2⎠ Let us find (a) Rx y ( τ ) and (b) Ry x ( τ ) . let us verify Eq.

28. 1. Ry x ( τ ) = −1 ∫e 1 − (t − τ) dt 1⎞ ⎛ = eτ ⎜ e − ⎟ e⎠ ⎝ Rx y ( τ ) and Ry x ( τ ) are plotted in Fig. Hence. V. Ry x ( τ ) = ∫e τ 1 − (t − τ) dt −τ − 1⎤ = eτ ⎡ ⎣e − e ⎦ − 1− τ = 1− e ( ) (iii) For τ ≤ − 1. y ( t ) and x ( t − τ ) do not overlap.83 Indian Institute of Technology Madras .Principles of Communication Prof. (ii) For − 1 < τ ≤ 1. 1. Venkata Rao (iii) τ ≥ 1: Rx y ( τ ) = τ+1 τ−1 ∫ − τ−1 − τ+1 e − t dt = e ( ) − e ( ) 1⎞ ⎛ = e− τ ⎜ e − ⎟ e⎠ ⎝ (b) Ry x ( τ ) : (i) For τ > 1 . Ry x ( τ ) = 0 for τ > 1 .

1. (1.Principles of Communication Prof.41 implies that for orthogonal signals. namely Ry x ( 0 ) = 0 . 1. from Eq. 1.28: Cross correlation functions of Example 1.37(a). say x ( t ) and y ( t ) Rx y ( τ ) τ=0 = 0 (1. Venkata Rao Fig.9: ∞ Two signals x ( t ) and y ( t ) are said to be orthogonal if = 0 −∞ ∗ ∫ x ( t ) y ( t ) dt (1.18 From the plots of Rx y ( τ ) and Ry x ( τ ) .84 Indian Institute of Technology Madras .40) Eq. it is easy to see that Ry x ( τ ) = R x y ( − τ ) Def.41(a). V.41a) We have the companion relation to Eq. for any integer n . 1. 1. 1. Then.41b) Let x p ( t ) and y p ( t ) be periodic with period T0 . if x ( t ) and y ( t ) are orthogonal.

6. 1. In Eq.Principles of Communication Prof. Rx ( τ ) = ∞ ∗ ∫ x (τ + λ ) x (λ ) d λ −∞ ∞ = −∞ ∗ ∫ x ( t + τ ) x ( t ) dt (1. Derivation of the FT of Rx p y p ( τ ) is given in appendix A1. 1. we find Ry p x p ( τ ) . This is quite meaningful because. Similarly. 1. assuming τ positive. we use somewhat simplified notation. whereas as a value of Rx ( τ ) close to zero implies the orthogonality of the two signals.42(b) gives another relation for Rx ( τ ) .42b) Eq. let x ( t ) = y ( t ) . Hence Rx ( τ ) can provide some information about the time variations of the signal. we have Rx x ( τ ) = ∞ ∗ ∫ x ( t ) x ( t − τ ) dt (1.35(a). If x ( t ) and x ∗ ( t − τ ) are quite similar. Then.85 Indian Institute of Technology Madras . V. Rx p y p ( τ ) is also periodic with the same period as x p ( t ) and y p ( t ) . x ∗ ( t − τ ) is a right shifted version of x ∗ ( t ) . 1. 1. We then have.42a) −∞ Instead of Rx x ( τ ) . namely. Venkata Rao Rx p y p ( τ + nT0 ) = Rxp y p ( τ ) That is. we can expect large value for Rx ( τ ) . Let ( t − τ ) = λ in Eq.2. ACF compares x ( t ) with a shifted and conjugated version of itself. Rx ( τ ) which is called the auto correlation function of x ( t ) .2 Autocorrelation function (ACF) ACF can be treated as a special case of CCF. In Eq.42(a).

42(a). 1. 1.45a) 1. for the power signals. the product of Eq.44b) − 0 2 Properties of ACF (energy signals) P1) ACF exhibits conjugate symmetry. This does not change the integral of Eq. That is.Principles of Communication Prof. then the ACF is special case of Eq.44a) 0 2 T0 1 = T0 2 ∗ xp ( t ) x p ( t + τ ) dt (1. we keep x ( t ) fixed and move x ∗ ( t ) to the right by τ and take the product. because if we let t = t1 and τ = τ1 . we keep x ∗ ( t ) fixed and move x ( t ) to the left by τ . 1.) If x ( t ) is a power signal. This product is obtained from Eq. (Note that shifting a function does not change its area. ∗ Rx ( − τ ) = Rx ( τ) (1. For a given τ = τ1 . T Rx ( τ ) = lim T →∞ − x ( t + τ ) x ∗ ( t ) dt ∫ T 2 2 (1. 1. Venkata Rao 1.43a) It can easily be shown that. 1.42(b).42(b) for t = t1 − τ1 .86 Indian Institute of Technology Madras .42(a) is x ( t1 ) x ∗ ( t1 − τ1 ) . as t is varied.36(a). in Eq. V. all the product quantities are obtained and hence the integral for a given τ1 remains the same. we have T0 Rx ( τ ) = 1 T0 − ∫ T ∫ T 2 ∗ xp (t ) xp ( t − τ ) dt (1.43b) For power signals that are periodic. we have T Rx ( τ ) = lim T →∞ − x ( t ) x ∗ ( t − τ ) dt ∫ T 2 2 (1.42(a). That is.

Principles of Communication Prof. 1. P2) Rx ( 0 ) = ∞ −∞ ∫ x (t ) 2 d t = Ex (1.39(b). Proof: The proof of the above property follows from Schwarz’s inequality. Rx ( τ ) ←⎯ Proof From Eq. ∞ That is. −∞ ∫ g1 ( t ) g2 ( t ) dt ∞ ≤ −∞ ∫ g1 ( t ) dt 2 ∞ −∞ ∫ g 2 ( t ) dt .45c) Let E x ( f ) denote the Energy Spectral Density (ESD) of the signal x ( t ) .45(a) implies that the real part of Rx ( τ ) is an even function of τ where as the imaginary part is an odd function of τ . That is. V. 1. Let g1 ( t ) and g 2 ( t ) be two energy signals.10). ∞ −∞ 2 ∫ x ( t ) x ( t − τ ) dt ∗ ∞ ≤ −∞ ∫ x ( t ) dt 2 ∞ −∞ ∫ x ∗ ( t − τ ) dt 2 Rx ( τ ) 2 ≤ ⎡ ⎣R x ( 0 ) ⎤ ⎦ or 2 Rx ( τ ) ≤ Rx ( 0 ) P4) (1. P3) Maximum value of Rx ( τ ) occurs at the origin. 1. ∞ 2 Then. Venkata Rao Proof: Exercise Eq. 2 Let g1 ( t ) = x ( t ) and g 2 ( t ) = x ∗ ( t − τ ) . which is stated below. Rx ( τ ) ≤ Rx ( 0 ) . −∞ ∫ Ex ( f ) df = Ex → Ex (f ) Then.45b) where E x is the energy of the signal x ( t ) (Eq. Ex y (f ) = X (f ) Y ∗ (f ) 1. From the Schwarz’s Inequality.87 Indian Institute of Technology Madras .

That is. then.45d) It is to be noted that E x ( f ) depends only on the magnitude spectrum. That is. its ACF is unique. x (t ) may not have any resemblance to y ( t ) . 2 Rx ( τ ) ←⎯ → X (f ) 2 2 But X ( f ) is the ESD of x ( t ) . X ( f ) . Proofs of these properties are left as an exercise. P5) Let x ( t ) = y ( t ) + v ( t ) .45g) Properties of ACF (periodic signals) We list below the properties of the ACF of periodic signals. Then. V. Ex (f ) = Ex x (f ) = X (f ) 2 Hence. y ( t ) and v ∗ ( t − τ ) are orthogonal for all τ ). then E x x ( f ) = X ( f ) . but given some ACF. Rx ( τ ) ←⎯ → Ex (f ) (1. Let x ( t ) and y ( t ) be two signals such that X ( f ) = Y ( f ) . that is E x ( f ) = E y ( f ) + Ev ( f ) (1.45f) In such a situation. Rx ( τ ) = Ry ( τ ) + Rv ( τ ) (1. Given x ( t ) . Note that if θ x ( f ) ≠ θy ( f ) . In other words. but their ACFs will be the same. ACF does not provide a unique description of the signal. Rx ( τ ) is the superposition of the ACFs of the components of x ( t ) . Rx ( τ ) = Ry ( τ ) + Rv ( τ ) + Ry v ( τ ) + Rv y ( τ ) (1. This also leads to the superposition of the ESDs.88 Indian Institute of Technology Madras . 1. Rx ( τ ) = Ry ( τ ) .45e) Proof: Exercise If Ry v ( τ ) = Rv y ( τ ) ≡ 0 (that is. Venkata Rao Let x ( t ) = y ( t ) . we can find many signals that have the given ACF. Then.Principles of Communication Prof.

16).46a) x p ( t ) d t = Px p 2 P2) Rx p ( τ ) τ = 0 1 = T0 − T0 ∫ 2 (1. 1. we expect the PSD to be purely impulsive. That is. P3) Rx p ( τ ) ≤ Rx ( 0 ) (1. . P4) Rx p ( τ ± nT0 ) = Rx p ( τ ) . 2 n = −∞ ∑ ∞ ⎛n⎞ ⎛ n⎞ X ⎜ ⎟ δ⎜f − ⎟ T0 ⎠ ⎝ T0 ⎠ ⎝ Example 1. outside 1 Show that F ⎡Rx p ( τ ) ⎤ = Px p ( f ) = 2 ⎣ ⎦ T0 and x ( t ) ←⎯ → X (f ) . P5) Let Px p ( f ) denote the Power Spectral Density (PSD) of x p ( t ) . 2.19 a) Let x ( t ) be the signal shown in Fig.46c) That is. Then. 1. Rx p ( τ ) ←⎯ → Px p ( f ) (1. Exercise 1. . Venkata Rao P1) ACF exhibits conjugate symmetry ∗ Rx p ( − τ ) = Rx ( τ) p T0 (1. the ACF of a periodic signal is also periodic with the same period as that of the signal. Pg.46e) As Rx p ( τ ) is periodic. Compute and sketch Rx ( τ ) .89 Indian Institute of Technology Madras . 1. − 0 < t < 0 Let x ( t ) = ⎨ 2 2 ⎪ ⎩ 0 .46d) where T0 is the period of x p ( t ) . 3.2.46b) 2 where Px p denotes the average power of x p ( t ) (Sec. That is.Principles of Communication Prof.29.2. n = 1. V.7 T T ⎧ ⎪xp (t ) . (1. ∞ −∞ ∫ Px ( f ) df p = Px p . the maximum value of Rx p ( τ ) occurs at the origin. 1. .

the positive part ⎝ 4⎠ of x ( t − τ ) fully overlaps with the negative part of x ( t ) . In addition. As τ increases in this range. V.90 Indian Institute of Technology Madras . this makes Rx ( τ ) further negative and the ACF reaches its minimum value at τ = 2 .Principles of Communication Prof. there by further reducing the positive value of the ∫ x (t ) x (t − τ) . 2 = 1. the overlap between the positive parts of the pulses x ( t ) and x ( t − τ ) (and also between the negative parts and these pulses) decreases. 1 + 1 . a part of x ( t − τ ) that is positive overlaps with the negative part of x ( t ) . (The student is advised to make a sketch of x ( t ) and x ( t − τ ) . Fig.) This decrease is linear (with a constant ⎛ 1⎞ slope) until τ = 1 . at τ = 0 and Rx ( 0 ) = E x . 1. If the output y ( t ) of the system is Rx ( t − 3 ) . which implies a decrease in the positive value for the integral of the product.19 a) Computation of Rx ( τ ) : We know that the maximum value of the ACF occurs at the origin.29: x ( t ) of Example 1. Venkata Rao b) x ( t ) of part (a) is given as the input to an LTI system with the impulse response h ( t ) . The value of Rx (1) is ⎜ − ⎟ . As can easily be 1.5 4 Consider the product x ( t ) x ( t − τ ) for 0 < τ ≤ 1. that is. Rx ( 0 ) = 1. find h ( t ) . For 1 < τ < 2 .

h ( t ) = x ⎡ ⎣− ( t − 3 )⎤ ⎦ which is sketched in Fig. 1. Rx ( t ) = x ( t ) ∗ x ( − t ) .91 Indian Institute of Technology Madras . 1. Rx ( 2 ) = ⎜ − ⎟ . As τ increases beyond 2. 1. 1. Fig.30: ACF of the signal of Example 1. we have all the information necessary to sketch Rx ( τ ) .39(a). which is shown in Fig. As Rx ( − τ ) = Rx ( τ ) . the Rx ( τ ) becomes less ⎝ 2⎠ and less negative and becomes zero at τ = 3 . y ( t ) = x ( t ) ∗ ⎡ ⎣ x ( − t ) ∗ δ ( t − 3 )⎤ ⎦ = x (t ) ∗ x ⎡ ⎣− ( t − 3 )⎤ ⎦ That is. 1.Principles of Communication Prof. Venkata Rao ⎛ 1⎞ checked.30. Hence.19 b) Calculating h ( t ) : We have y ( t ) = x ( t ) ∗ h ( t ) = Rx ( t − 3 ) = Rx ( t ) ∗ δ ( t − 3 ) But from Eq. V.31.

1. V.20 Let x ( t ) = A cos ( ω0 t + θ ) .31 Impulse response of the LTI system of Example 1. We will find Rx ( τ ) . 2 (iv) Rx ( τ ) is independent of θ . Method 1: T0 Rx ( τ ) = 1 T0 − ∫ T T0 2 A2 cos ( ω0 t + θ ) cos ⎡ ⎣ω0 ( t − τ ) + θ ⎤ ⎦ dt 0 2 2 A2 = T0 = − ∫ T 0 1 {cos ( 2 ω0 t − ω0 τ + 2 θ ) + cos ω0τ} dt 2 2 A2 cos ω0 τ 2 We find that: (i) Rx ( τ ) is periodic with the same period as x ( t ) .92 Indian Institute of Technology Madras . Venkata Rao Fig.19 Example 1. 1. (ii) Its maximum value occurs at τ = 0 (iii) The maximum value is A2 which is the average power of the signal.Principles of Communication Prof.

we show x ( t ) and x ( t − τ ) for 0 < τ < 2 . In Fig. x ( t ) x ( t − τ ) = 0 which implies Rx ( τ ) = 0 for τ > 2 . Hence. Note that if τ > 2 .32. 0 ≤ t ≤ 2 . Let us find its ACF and sketch it.93 Indian Institute of Technology Madras . Rx ( τ ) = 2 Re ⎡ ⎣Rv ( τ ) ⎤ ⎦ ⎡ A ⎢1 Rv ( τ ) = 4 ⎢T0 ⎢ ⎣ 2 T0 − ∫ T 2 j ω t + θ) − e ( 0 e ⎡ω0 ( t − τ ) + θ ⎦ ⎤ j⎣ 0 2 ⎤ A2 = dt ⎥ ⎥ 4T0 ⎥ ⎦ T0 − ∫ T 2 e j ω0 τ dt 0 2 = A2 j ω0 τ e 4 A2 cos ( ω0 τ ) 2 Hence Rx ( τ ) = Example 1. 1. Rx ( τ ) = Rv ( τ ) + Rw ( τ ) ∗ But Rw ( τ ) = Rv ( τ ) . Venkata Rao Method 2: ⎡ A j ω t + θ ) A − j ( ω0 t + θ ) ⎤ A cos ( ω0 t + θ ) = ⎢ e ( 0 + e ⎥ 2 ⎣2 ⎦ = v (t ) + w (t ) As x ( t ) = v ( t ) + w ( t ) . we have Rx ( τ ) = Rv ( τ ) + Rw ( τ ) + Rv w ( τ ) + Rw v ( τ ) . Hence.Principles of Communication Prof. 1. V. It is not difficult to see that Rv w ( τ ) = Rw v ( τ ) = 0 for all τ .21 Let x ( t ) = sin ( 2 π t ) .

Principles of Communication Prof.33. 1. V. 1. we have ⎧ τ cos ( 2 π τ ) sin ( 2 π τ ) . τ ≤ 2 + ⎪cos ( 2 π τ ) − Rx ( τ ) = ⎨ 2 2π ⎪ . otherwise 0 ⎩ Rx ( τ ) is plotted in Fig.21 and (b) its shifted version Rx ( τ ) = = ∫ sin ( 2 π t ) sin ⎡ ⎣2 π ( t − τ ) ⎤ ⎦ dt τ 2 ∫ τ 2 cos ( 2 π τ ) − cos ( 4 π t − 2 π τ ) 2 dt 2 2 ⎡ sin ( 4 π t − 2 π τ ) ⎤ ⎡ t cos 2 π τ ⎤ = ⎢ − ⎢ ⎥ ⎥ 2 4π ⎣ ⎦τ ⎣ ⎦τ = cos ( 2 π τ ) − = cos ( 2 π τ ) − τ cos ( 2 π τ ) 2 − sin ( 8 π − 2 π τ ) − sin ( 2 π τ ) 4π τ cos 2 π τ sin 2 π τ + . 1.32: (a) Sinusoidal pulse of Example 1.94 Indian Institute of Technology Madras . Venkata Rao Fig. 0 ≤ τ ≤2 2 2π As Rx ( − τ ) = Rx ( τ ) .

21 1. V. both the magnitude spectrum and the phase spectrum of the input signal. Y ( f ) = X ( f ) H ( f ) θ y ( f ) = θ x ( f ) + θh ( f ) (1. 1. in general.47b) From Eq. the output y ( t ) is y (t ) = x (t ) ∗ h (t ) or Y (f ) = X (f ) ∗ H (f ) That is.95 Indian Institute of Technology Madras . there are certain networks. which would alter only the (input) phase spectrum.33: Rx ( τ ) of Example 1. That is.7 Hilbert Transform Let x ( t ) be the input of an LTI system with the impulse response h ( t ) . if H ( f ) is the frequency response of an all pass network. Then. Venkata Rao Fig. then 1.47 we see that an LTI system alters. 1.47a) (1. However.Principles of Communication Prof. called all pass networks.

f > 0 ⎪ = ⎨ 0 .48b) ˆ (t ) When x ( t ) is the input to a Hilbert transformer.96 Indian Institute of Technology Madras .Principles of Communication Prof. with the impulse response h ( t ) = δ ( t − td ) . phase shift imparted to each input spectral component is proportional to the frequency. we define the Hilbert transformer in the frequency domain. f > 0 ⎪ ⎪ 2 = ⎨ ⎪ π + θ (f ) . ⎝ 2⎠ Hence. the proportionality constant being 2 π td . a Hilbert transform is essentially a ⎜ ± ⎟ phase shifter. Though θ y ( f ) ≠ θ x ( f ) . with the frequency response function H ( f ) = − j sgn ( f ) (1. Venkata Rao Y (f ) = X (f ) θ y ( f ) = θ x ( f ) + θh ( f ) An interesting case of all-pass network is the ideal delay. we denote its output as x where ˆ (t ) = x (t ) ∗ x 1 πt 1.48a) where sgn ( f ) ⎧ 1 . f < 0 ⎩ 1 → As ⎡ . f ≠ 0 and ⎧ π − + θx (f ) . Its output is characterized by: (i) Y ( f ) = X ( f ) . f = 0 ⎪ −1 . ⎣ − j sgn ( f ) ⎤ ⎦ ←⎯ πt h (t ) = 1 πt (1. Another interesting network is the Hilbert transformer. V. f < 0 x ⎪ ⎩ 2 (ii) θy ( f ) ⎛ π⎞ That is.

Let x ( t ) = δ ( t ) .Principles of Communication Prof. Let x ( t ) = cos ( 2 π f0 t ) . ⎢ δ ( t ) ∗ ⎥ = !! πt ⎦ πt ⎣ Example 1.23 HT of a cosine signal. V. let us look at some examples of HT. For the present. Hilbert Transform (HT) will prove quite useful later on in the study of bandpass signals and single sideband signals. 1. Venkata Rao = 1 π ∫ (t − τ) d τ −∞ ∞ x ( τ) 1 (1. Let us find x → 1.97 Indian Institute of Technology Madras . both x ( t ) and x variable ( t in our case). x ˆ ( t ) are functions of the same Note: Unlike other transforms.22 Hilbert transform δ ( t ) . we have As δ ( t ) ←⎯ ˆ ⎤ F⎡ ⎣δ ( t ) ⎦ = − j sgn ( f ) ⇒ ˆ (t ) = 1 δ πt ⎡ 1⎤ 1 This also establishes the relation. ˆ (t ) . Let us find x 1 This integral is actually Cauchy’s principal value. Example 1. ˆ (t ) .49b) and ˆ ( f ) = − j sgn ( f ) X ( f ) X ˆ ( t ) is called the Hilbert transform of x ( t ) .49a) (1.

we can show that if x ( t ) = sin ( 2 π f0 t ) . That is. Venkata Rao X (f ) = 1 ⎡ δ ( f − f0 ) + δ ( f + f0 ) ⎤ ⎦ 2⎣ ˆ ( f ) = − j sgn ( f ) X ( f ) . V.98 Indian Institute of Technology Madras .24 Let x ( t ) = 1 ˆ (t ) . x Alternatively. X ˆ ( f ) = 1 ⎡δ ( f − f ) − δ ( f + f )⎤ X 0 0 ⎦ 2j ⎣ ˆ ( t ) = sin ( 2 π f0 t ) . then x − j 2 π f0 t − π ( ) π⎞ 1 ˆ ( t ) = cos ⎛ ⎡ has x x1 ( t ) + x2 ( t ) ⎤ ⎜ ω0 t − 2 ⎟ ⎣ ⎦ 2 ⎝ ⎠ Example 1. Similarly. we have t −τ −∞ ∞ ∫ 1.Principles of Communication Prof. Let us find x 1 + t2 ˆ (t ) = x (t ) ∗ x 1 = π ∞ 1 πt 1 dτ ∫ 2 − ∞ (1 + τ ) ( t − τ ) ⎡∞ t +τ ⎢∫ dτ + 2 ⎢ ⎣− ∞ 1 + τ dτ ⎤ ∫ t − τ) ⎥ ⎥ −∞( ⎦ ∞ 1 1 = π 1 + t2 τ dτ = As ∫ 2 − ∞1+ τ ∞ 1 d τ = 0 . then x j 2 π f0 t j 2 π f0 t − π ( 2 ) 2 ˆ2 ( t ) = e . . That is. ˆ1 ( t ) = e if x1 ( t ) = e j 2 π f0 t . x ( t ) = cos ( 2 π f0 t ) = = sin ( ω0 t ) ˆ ( t ) = − cos ω0 t . then x and if x2 ( t ) = e − Hence.

we have ˆ (t ) = x t 1 + t2 Exercise 1. V. f > 0 M ( f − fc ) e ⎪ ⎪2 = ⎨ π ⎪ 1 M (f + f ) e j 2 . We will show that ˆ ( t ) = m ( t ) sin ( 2 π fc t ) = m ( t ) cos ( 2 π fc t ) . Hence. f < 0 c ⎪ ⎩2 1.8 (a) Let x1 ( t ) = sin t 1 − cos t ˆ1 ( t ) = . ˆ (f ) X − j π ⎧1 2 . x X (f ) = 1 ⎡M ( f − fc ) + M ( f + fc ) ⎤ ⎦ 2⎣ ˆ ( f ) = 1 ⎡M ( f − f ) + M ( f + f ) ⎤ ⎡ − j sgn ( f ) ⎤ X c c ⎦ ⎣ ⎦ 2⎣ But ⎡ ⎣M ( f − fc ) ⎤ ⎦ is nonzero only for f > 0 and ⎡ ⎣M ( f + fc ) ⎤ ⎦ is nonzero only for f < 0 . Show that x 1 π t + 2 Example 1.Principles of Communication Prof. Venkata Rao 1 t ˆ (t ) = x π 1 + t2 ⎡∞ 1 ⎤ ⎢∫ ⎥ d τ 2 1 + τ ⎢ ⎥ − ∞ ⎣ ⎦ As the bracketed integral is equal to π .25 Let x ( t ) = m ( t ) cos 2 π fc t where m ( t ) is a lowpass signal with M ( f ) = 0 for f > W and fc > W . Show that x t t 1 t − 1 2 ˆ 2 ( t ) = − ln (b) Let x2 ( t ) = ga ( t ) .99 Indian Institute of Technology Madras .

We assume that the signals under consideration have no impulses in their spectra at f = 0 .100 Indian Institute of Technology Madras . Then ˆ (t ) = m (t ) v ˆ ( t ) . f > 0 ⎪2 F⎡ ⎣m ( t ) sin ( 2 π fc t ) ⎤ ⎦ = ⎨ π j ⎪1 2 . Hence. V. Let x ( t ) = m ( t ) v ( t ) where m ( t ) is a lowpass signal with M ( f ) = 0 for f > W and v ( t ) is a high-pass signal with V ( f ) = 0 for f < W . π ⎧1 − j ⎪ M ( f − fc ) e 2 . 1. which is stated below. we have ˆ ( t ) = m ( t ) sin ( 2 π fc t ) = m ( t ) cos ( 2 π fc t ) x Note: It is possible to establish even a stronger result. f < 0 + M f f e ( c) ⎪ ⎩2 ˆ As F ⎡ ⎣m ( t ) sin 2 π fc t ⎤ ⎦ = X ( f ) . [We assume that there are no impulses in either M ( f ) or x V (f ) . P1) ˆ ( t ) .7. have the same energy. Venkata Rao Consider m ( t ) sin ( 2 π fc t ) ←⎯ → M (f ) ∗ ←⎯ → M (f ) ∗ 1 ⎡ δ ( f − fc ) − δ ( f + fc ) ⎤ ⎦ 2j⎣ − 1⎡ ⎢δ ( f − fc ) e 2⎣ ⎢ j π 2 + ejπ e π 2 − j π 2 ⎤ δ ( f + fc ) ⎥ ⎥ ⎦ − 1⎡ ←⎯ → M ( f ) ∗ ⎢ δ ( f − fc ) e 2⎣ ⎢ j π 2 +e j ⎤ δ ( f + fc ) ⎥ ⎥ ⎦ That is. we develop the properties of HT as applied to real signals. x ∞ Proof: E x = −∞ ∫ X (f ) d f 2 1.1 Properties of Hilbert transform Our area of application of HT is real signals. A signal x ( t ) and its HT.Principles of Communication Prof.

x P3) 1. A signal x ( t ) and its HT. this is satisfying. Venkata Rao ˆ ( f ) = − j sgn ( f ) X ( f ) As X ˆ ( f ) = X ( f ) . and P2) That is.Principles of Communication Prof.22. are orthogonal. That is. ˆ ( t )⎤ {HT ⎡ ⎣x ⎦} = − x (t ) f = 0 is zero. X x x ˆ (f ) Note: Though X hence the energy. We shall find x t ˆ (t ) = 1 . we change the phase of a spectral component in X ( f ) by 900 . HT ⎡ δ ⎢ πt ⎥ ⎣ ⎦ ⎣ ⎦ ⎡ 1⎤ HT ⎢ ⎥ = − π δ ( t ) ⎣t ⎦ ˆ ( t ) . we have δ πt Hence.26 Let x ( t ) = 1 ˆ (t ) . ˆ ( t ) ←⎯ Proof: x → − j sgn ( f ) X ( f ) ˆ ( t )⎦ ⎡x ⎤ ←⎯ ⎡ − j sgn ( f ) X ( f ) ⎦ ⎤ HT ⎣ → − j sgn ( f ) ⎣ = ( − j sgn ( f ) ) X ( f ) = − X ( f ) 2 Example 1. Hence E = E ˆ . applying the HT twice on a given signal x ( t ) changes the sign of the signal. V. ˆ ( t ) ⎤ = − δ ( t ) = HT ⎡ 1 ⎤ . From Example 1. Hence.101 Indian Institute of Technology Madras . Each time we perform HT. Intuitively. it will not change the value of the integral. . performing the transformation twice results in a phase shift of 1800 .

1. Hence. X ( − f ) = X ∗ ( f ) . ∞ −∞ ∫ ˆ (t ) d t = x (t ) x ∞ −∞ ∫ − j sgn ( − f ) X ( f ) d f 2 = 0 (Note that the integrand on the RHS is odd) ˆ ( α t )⎤ P4) HT ⎡ ⎣ sgn ( α ) x ⎦.Principles of Communication Prof. where α is a nonzero constant is ⎡ Proof: We will first establish that if y (t ) = x (t ) ∗ h (t ) and z ( t ) = x ( α t ) ∗ h ( α t ) . ∞ −∞ ∫ ˆ (t ) d t = x (t ) x ∞ −∞ ⎣− ∫ X (f ) ⎡ j sgn ( − f ) X ( − f ) ⎤ ⎦ df As x ( t ) is real. α If z ( t ) = x ( α t ) ∗ h ( α t ) . y ( α t ) ←⎯ → ⎛f ⎞ ⎛f ⎞ ⎛f ⎞ Y ⎜ ⎟ = X⎜ ⎟ H⎜ ⎟ ⎝α⎠ ⎝α⎠ ⎝α⎠ = α Z (f ) 2 Hence. y ( α t ) ←⎯ → That is. Venkata Rao Proof: We know that ∞ −∞ ∫ ˆ (t ) d t = x (t ) x ∞ −∞ ˆ (− f ) d f ∫ X (f ) X (See the note. ⎣ x ( α t )⎤ ⎦ . property P10. Sec. then z ( t ) = 1 y (α t ) . where α ⎜ ⎝α⎠ Also. Therefore. V.102 Indian Institute of Technology Madras . then Z (f ) = 1 α 2 ⎛f ⎞ ⎛f ⎞ X⎜ ⎟ H⎜ ⎟ ⎝a⎠ ⎝a⎠ 1 ⎛f ⎞ Y ⎟ . z ( t ) = 1 2 α Z (f ) = α Z (f ) α y (α t ) α 1.4).

Then 1 + t2 ⎡ 1 ⎤ 1 using P4. Let us obtain HT ⎢ 2⎥ 1 4 t + 1 + 4t2 ⎣ ⎦ ˆ (t ) = As x ⎡ 1 ⎤ t HT . ⎟ t ⎝ t ⎠ find the HT ⎡ ⎣sin c ( t ) ⎤ ⎦. let x ( t ) = x (α t ) = 1 and α = 2 .9 1 − cos t ⎛ sin t ⎞ = Using the result HT ⎜ . ⎢x (α t ) ∗ π α t ⎥ = α ⎣ ⎦ α ˆ ( α t ) = sgn ( α ) x ˆ (α t ) . sgn ( α ) x 2⎟ ⎝1 + 4t ⎠ = 2t . Venkata Rao HT ⎡ ⎣ x ( α t )⎤ ⎦ = x (α t ) ∗ 1 πt ⎛ 1 ⎞ = x (α t ) ∗ α ⎜ ⎟ ⎝ α πt ⎠ ⎡ 1 ⎤ = α ⎢x (α t ) ∗ α πt ⎥ ⎣ ⎦ ⎡ 1⎤ ˆ ( t ) . 1 + 4t2 ⎛ − 2t ⎞ ˆ (α t ) = − ⎜ With α = − 2 . .Principles of Communication Prof. we have As ⎢ x ( t ) ∗ ⎥ = x πt ⎦ ⎣ ˆ (α t ) x ⎡ 1 ⎤ . = sgn ( 2 ) ⎢ 2⎥ 1 + t2 ⎣1 + 4 t ⎦ ⎡ 2t ⎤ 2t ⎢ ⎥ = 2 1 + 4t2 ⎢1 + ( 2 t ) ⎦ ⎥ ⎣ If α = − 2 . Hence HT ⎡ ⎣ x ( α t )⎤ ⎦ = α x As a simple illustration of the property. as required 1 + 4t2 Exercise 1. V. 1.103 Indian Institute of Technology Madras . x ( α t ) = 1 which is the same as with α = 2 .

Then the spectrum S ( f ) of the transmitted signal is S ( f ) = 1 ⎡M ( f − fc ) + M ( f + fc ) ⎤ ⎦ . That is. If M ( f ) is as shown in 2⎣ Fig 1. S ( f ) will be as shown in Fig 1. Rx x P5) The cross correlation function of x ( t ) and x ˆ ( τ ) is the negative of ˆ ( τ ) .34(a). 1.34(b).104 Indian Institute of Technology Madras .Principles of Communication Prof. V.8 Bandpass Signals Consider a communication system that transmits the signal s ( t ) = m ( t ) cos 2π fc t . Venkata Rao ˆ ( t ) . 1. ˆ Rx x ˆ ( τ ) = − Rx ( τ ) Exercise 1. the HT of R x ˆ Rx x ˆ ( τ ) = − Rx ( τ ) ˆ ( − τ )⎤ Proof: F ⎡ ˆ ( τ)⎤ ⎣ x ( τ) ∗ x ⎦ ⎣R x x ⎦ = F⎡ = X (f ) ⎡ ⎣ − j sgn ( − f ) X ( − f ) ⎤ ⎦ = X (f ) ⎡ ⎣ − j sgn ( − f ) ⎤ ⎦ = X (f ) ⎡ ⎣ j sgn ( f ) ⎤ ⎦ 2 2 That is.10 ˆ Show that Rx ˆ x ( τ ) = Rx ( τ ) . then for a carrier frequency fc = 100 k Hz . where m ( t ) is a (lowpass) message signal and cos ( 2π fc t ) is the (high-frequency) carrier term.

Moreover s ( t ) is a narrowband.35 shows some more spectra that represent NBBP signals. is quite small in comparison with the carrier frequency fc of 100 k Hz . bandpass spectrum We see that the spectrum of s ( t ) is confined to the frequency interval 95 ≤ f ≤ 105 k Hz . Bandpass (NBBP) signal. (105 − 95 ) = 10 k Hz . s ( t ) is a bandpass signal.105 Indian Institute of Technology Madras . 1. bandpass signal because the spectral width of S ( f ) . Fig 1. Venkata Rao Fig. Hence. we call s ( t ) as a Narrowband.Principles of Communication Prof.34: A typical narrowband. Where as m ( t ) is a lowpass signal. V. 1.

) The mathematical concepts of preenvelope and complex envelope have been developed for this purpose. 1. V.106 Indian Institute of Technology Madras . bandpass signals and bandpass systems are studied in terms of their lowpass equivalents. (That is. the analysis of the communications schemes would become somewhat simplified.) is also of the bandpass variety. The study of such transmissionreception schemes becomes a little complicated because of the presence of the carrier term in some form or the other (The term cos ωc t in m ( t ) cos ωc t is meant only to “carry” the information m ( t ) and is not part of the information) If we develop tools to study bandpass signals and bandpass systems. independent of the carrier.Principles of Communication Prof. Venkata Rao Fig. 1. Let us assume that the rest of the communication system (channel. a part of the receiver etc. We shall make use of these concepts in our studies on linear modulation and angle modulation.35: A few more examples of NBBP spectra NBBP signals play an important role in the communication process.

f = 0 ⎪ 0 . 1.10: Let x ( t ) be any real signal with the FT. even though X ( f ) is two sided (As x p e ( t ) has spectral components only for f ≥ 0 . X ( f ) . 1. X p e (f ) ⎧2 X ( f ) .107 Indian Institute of Technology Madras .1 Pre-envelope Def.Principles of Communication Prof. Venkata Rao 1. 1.50a) Taking the Fourier transform of Eq. f < 0 ⎩ (1.50b) (We assume that X ( f ) has no impulse at f = 0 ). 1.8. f > 0 ⎪ = ⎨ X (0) . Of course. we have X p e (f ) = X (f ) + j ⎡ ⎣ − j sgn ( f ) X ( f ) ⎤ ⎦ = X ( f ) + sgn ( f ) X ( f ) That is. x p e ( t ) has spectrum only for f ≥ 0 .36. That is. x p e ( t ) will have spectrum only for f ≤ 0 ). some authors use the symbol x+ ( t ) to denote the ∗ pre-envelope of x ( t ) ). Consider the signals x1 ( t ) and x2 ( t ) whose spectra are shown in Fig.54(a). ˆ (t ) xp e (t ) = x (t ) + j x (1. V. We define its pre- envelope as.

p e ( t ) of the signals in 1. Fig. Venkata Rao Fig. p e ( t ) and (b) x2. 1. V.36: Typical two-sided spectra The corresponding X1. 1. 1.108 Indian Institute of Technology Madras .37(a) and (b) respectively. p e ( f ) are as shown in Fig.37: Fourier transform of (a) x1.Principles of Communication Prof. p e ( f ) and X 2.

51) Example 1. x pe ( t ) = = ˆ (t ) = Then.109 Indian Institute of Technology Madras . → As ex1( t ) ←⎯ ex1( 2 π t ) ←⎯ → 1 1 + j 2πf 1 1 . we know that First. x 1 1+ jt = 1− jt 1 + t2 1 t + j .50(b). 1. f > 0 ⎪1 ⎪ = ⎨ . Let us find X pe ( f ) and x pe ( t ) .5. 2 1+ t 1 + t2 t .27 Let x ( t ) = 1 . 1 + t2 X ( f ) . 1 + t2 1. From Example 1. 2π 1 + j f 1 . Venkata Rao Because of Eq. Hence X pe ( f ) = 2 π e − 2 π f u ( f ) . V. where u ( f ) ⎧1 . we can write xp e (t ) = 2 ∞ ∫ X (f ) e 0 j 2πf t df (1.Principles of Communication Prof. 1− jt From duality. let us compute 1 − ←⎯ → πe 2 1+ t 2 πf . ⎪2 ⎪ ⎩0 . f < 0 We require F − 1 ⎡ ⎣ X pe ( f ) ⎤ ⎦. which is a known result. 2 π ex1( 2 π f ) = 2 π e − 2 π f u ( f ) ←⎯ → That is. f = 0.

Let x ( t ) be a bandpass signal with X p e ( f ) “centered” with respect to fc = 102 k Hz as shown in Fig. Venkata Rao As can be seen from the above discussion. Fig 1. 1. nonzero in the frequency range ( − 2 ) to 8 k Hz .38(b). the spectrum of x p e ( t ) is still bandpass (though one-sided) if x ( t ) is a bandpass signal. we see that X p e ( f + fc ) is a lowpass spectrum.38: A typical X p e ( f ) and shifted version (We can treat fc to be the center frequency in Fig. V. 1.110 Indian Institute of Technology Madras .38(a).Principles of Communication Prof. 1.) From Fig 1.38(a). though the spectrum is zero for the frequency range 94 to100 k Hz . by taking the bandpass spectrum from 94 to 110 k Hz .

111 Indian Institute of Technology Madras . j 2 π fc t xp e (t ) = ⎡ ⎣ xc ( t ) + j xs ( t ) ⎤ ⎦e As the real bandpass signal x ( t ) is the real part of x p e ( t ) .52(a) also implies x p e ( t ) = xc e ( t ) e j 2 π fc t (1. 1. x c e ( t ) = xc ( t ) + j xs ( t ) (1. we obtain. Note that sin ( 2 π fc t ) is in phase quadrature to cos ( 2 π fc t ) .52b) (We assume that we know the center frequency fc and it is such that X p e ( f + fc ) is lowpass in nature).53.Principles of Communication Prof. Venkata Rao 1.52(a) implies j 2 π fc t (1.54) We will show a little later that.2 Complex envelope Def. (It is also common in the literature to use the symbol xI ( t ) for the in-phase component and xQ ( t ) for the quadrature component.11: We now define the complex envelope of x ( t ) . denoted xc e ( t ) as xc e ( t ) = x p e ( t ) e − Eq. 1. xc ( t ) .53) xc e ( t ) is also referred to as the equivalent lowpass signal of the bandpass signal x ( t ) . Then. 1. Using Eq. the coefficient of the sine term.54 in Eq. which is the coefficient of the cosine term.55) Eq.) 1. 1. is usually referred to as the in- phase component and xs ( t ) . we have x ( t ) = xc ( t ) cos ( 2 π fc t ) − xs ( t ) sin ( 2 π fc t ) (1. both xc ( t ) and xs ( t ) are lowpass in nature.55 is referred to as the canonical representation of the bandpass signal. as the quadrature component. In general xc e ( t ) is complex. V.8. 1. Let xc ( t ) be the real part and xs ( t ) its imaginary part.52a) X c e ( f ) = X p e ( f + fc ) (1. Equation 1.

56c) (1. Venkata Rao To express Eq.112 Indian Institute of Technology Madras . 1. we find that they have a close resemblance. j 2 π fc t ⎤ jϕ A cos ( 2 π fc t + ϕ ) = Re ⎡ ⎣A e e ⎦ (1.39(a). 1.56a) (1.58 resembles phasor representation of a sinusoid. 1.58) jϕ t = Re ⎡ A ( t ) e ( ) e j 2 π fc t ⎤ ⎣ ⎦ Eq. Phasor of the monochromatic (single frequency) signal has constant amplitude A and a fixed phase ϕ . (slowly) time-varying. 1.57) (1.39(b).58 with Eq.59. (Note that a single frequency sinusoid is the extreme case of a narrowband signal with zero spectral width!) 1. (The phasor is a complex number providing information about the amplitude and phase (at t = 0 ) of the sinusoid. In the case of the complex envelope (of a narrowband signal) both A ( t ) and ϕ ( t ) are. ⎣ ⎦ Comparing Eq. let A (t ) = 2 xc ( t ) + xs2 ( t ) (1. 1.54 in polar form.56b) ⎡ x (t ) ⎤ ϕ ( t ) = tan−1 ⎢ s ⎥ ⎢ ⎣ xc ( t ) ⎥ ⎦ Then.) The quantity ⎡ A e j ϕ e j 2 π fc t ⎤ can be treated as a rotating vector as shown in Fig. V.59) ⎡ A e j ϕ ⎤ is generally referred to the as the phasor associated with the sinusoidal ⎣ ⎦ signal A cos ( 2 π fc t + ϕ ) . jϕ t xce ( t ) = A ( t ) e ( ) j 2 π fc t ⎤ ⎡ x ( t ) = Re ⎡ ⎣ xp e ( t )⎤ ⎦ = Re ⎣ xc e ( t ) e ⎦ (1.Principles of Communication Prof. This is shown in Fig 1. We know that.

different modulation schemes are basically different methods of controlling either A ( t ) or ϕ ( t ) (or both) as a function of the message signal m ( t ) . jϕ t that for a given time t = t1 . V. As we shall see later.Principles of Communication Prof. then xc ( t ) and xs ( t ) are lowpass signals with spectrum confined 1.113 Indian Institute of Technology Madras . Note. having the necessary information about the narrowband signal. Venkata Rao j 2 π f t + ϕ) Fig. the generalization permits the amplitude and phase to change as a function of time.39: (a) Phasor representation of A e ( c (b) Complex envelope as a (slowly) varying amplitude and phase In other words. 1. however. If x ( t ) is a NBBP signal with spectrum confined to the frequency range f ± fc ≤ W . A ( t1 ) e ( 1 ) is a complex number. complex envelope can be treated as a generalization of the phasor representation used for single frequency sinusoids.

Similarly. The scheme shown below (Fig 1.40: Scheme for the recovery of xc ( t ) and xs ( t ) from x ( t ) In Fig 1.40. v1 ( t ) = 2 x ( t ) cos ωc t = 2 ( xc ( t ) cos ωc t − xs ( t ) sin ωc t ) cos ωc t = 2 xc ( t ) cos2 ωc t − xs ( t ) sin 2 ωc t 1. x ( t ) a NBBP signal. with the spectrum confined to the interval f ± f c ≤ W . we have ∗ ⎡ xc e ( t ) + xc ⎤ e ( t )⎦ ⎣ xc ( t ) = 2 ( ) is nonzero only for f ≤ W . Venkata Rao to f ≤ W . X c ( f ) is also zero for f > W . 1. This is because X c e ( f ) = X p e f + f c As xc ( t ) is the real part of xc e ( t ) . or F⎡ ⎣ xc ( t ) ⎤ ⎦ = Xc (f ) = ∗ Xc e (f ) + Xc e (− f ) 2 ∗ As X c e ( f ) and X c e ( − f ) are zero for f > W . where W << fc . V.Principles of Communication Prof. xs ( t ) is also a lowpass signal with X s ( f ) = 0 for f > W . Fig.40) enables us to obtain xc ( t ) and xs ( t ) from x ( t ) .114 Indian Institute of Technology Madras .

(1. band limited to W Hz. ϕ ( t )⎤ ⎦ and these are lowpass in nature. its phase. outside xc ( t ) and xs ( t ) are lowpass signals.61 is referred to as the envelope and phase representation of x ( t ) .62) 1. V.61) Eq. As we assume that fc is known. Hl p ( f ) is an Ideal Lowpass Filter (ILPF) with the frequency response. we have. the information about x ( t ) is contained in either of the quantities ( xc ( t ) . From Eq. xs ( t ) ) or ⎡ ⎣ A ( t ).Principles of Communication Prof. A ( t ) is called the natural envelope (or simply the envelope) of x ( t ) and ϕ ( t ) . f ≤ W Hl p ( f ) = ⎨ ⎪ ⎩0 . ⎧ ⎪1 . we obtain xc ( t ) . Similar analysis will show that the output of the bottom channel is xs ( t ) . 1. Note that A ( t ) = xce ( t ) = x pe ( t ) and is always non-negative.58.60) In Fig. We shall now illustrate the concepts of x pe ( t ) . 1. A ( t ) and ϕ ( t ) with the help of a few examples. xce ( t ) . xc ( t ) cos ( 2 ωc t ) and xs ( t ) sin ( 2 ωc t ) have bandpass spectra centered at ± fc . Venkata Rao ⎡1 + cos 2 ωc t ⎤ = 2 xc ( t ) ⎢ ⎥ − xs ( t ) sin 2 ωc t 2 ⎣ ⎦ = xc ( t ) + xc ( t ) cos ( 2 ωc t ) − xs ( t ) sin 2 ωc t (1. x ( t ) = A ( t ) cos ⎡ ⎣ωc t + ϕ ( t ) ⎤ ⎦ (1.40.115 Indian Institute of Technology Madras . These quantities will be filtered out by the ILPF and at the output of the top channel. 1.

50(b). xce ( t ) and A ( t ) . ⎛t x pe ( t ) = ga ⎜ ⎝T ⎞ j 2 π fc t ⎟e ⎠ 1. Example 1. Method 1 (Frequency domain): Because of the assumption fc T >> 1 . A ( t ) and ϕ ( t ) . We shall find x pe ( t ) . we can take X ( f ) approximately as X (f ) ⎧T sin c ⎡ ⎣( f − fc )T ⎤ ⎦ . ϕ ( t ) = 0 and A ( t ) = xce ( t ) = 1. 1. Assume that fc T >> 1 so that x ( t ) can be ⎠ taken as a NBBP signal. otherwise ⎪ ⎩ Hence. we have Hence X pe ( f ) = δ ( f − fc ) ⇒ x pe ( t ) = e j 2 π fc t As X ce ( f ) = δ ( f ) . we obtain xce ( t ) = 1 As xce ( t ) is real and positive. xce ( t ) . X (f ) = 1 ⎡ δ ( f − fc ) + δ ( f + fc ) ⎤ ⎦ 2⎣ From Eq.Principles of Communication Prof. Venkata Rao Example 1. V.28 Let x ( t ) = cos [ 2 π fc t ] .50(b).116 Indian Institute of Technology Madras . f > 0 ⎪ ⎪2 ⎨ ⎪T sin c ⎡( f + f )T ⎤ . ⎧ ⎪T sin c ⎡ ⎣( f − fc )T ⎤ ⎦ . Let us find x pe ( t ) . 1.29 ⎛t Let x ( t ) = ga ⎜ ⎝T ⎞ ⎟ cos ωc t . f > 0 X pe ( f ) = ⎨ 0 . f < 0 c ⎣ ⎦ ⎪ ⎩2 From Eq.

the complex envelope is real valued and is equal to the envelope.28 and 1. Venkata Rao xce ( t ) = x pe ( t ) e − j 2 π fc t ⎛t ⎞ = ga ⎜ ⎟ ⎝T ⎠ As xce ( t ) is real. ⎛t ⎞ A ( t ) = xce ( t ) = g a ⎜ ⎟ ⎝T ⎠ Note that for Examples 1.55. Let us find xc ( t ) and xs ( t ) . Method 2 (Time domain): Comparing the given x ( t ) with Eq.29.117 Indian Institute of Technology Madras . we find that it is already in the ⎛t ⎞ canonic form with xc ( t ) = ga ⎜ ⎟ . xs ( t ) = 0 ⎝T ⎠ ⎛t ⎞ Hence. xce ( t ) = ga ⎜ ⎟ ⎝T ⎠ And x pe ( t ) = xce ( t ) e j 2 π fc t ⎛t ⎞ = ga ⎜ ⎟ e j 2 π fc t ⎝T ⎠ ⎛t ⎞ A ( t ) = ga ⎜ ⎟ and ϕ ( t ) = 0 ⎝T ⎠ Example 1. we have xs ( t ) = 0 ⇒ ϕ ( t ) = 0 .Principles of Communication Prof.30 x ( t ) is a NBBP signal with X ( f ) as shown in Fig 1. A ( t ) .41. 1. V. 1.

1. 1. Venkata Rao Fig.Principles of Communication Prof. Fig. V.41: Spectrum of the NBBP signal of Example 1. we can obtain X ce ( f ) .42: Spectrum of the complex envelope of the signal of Example 1.118 Indian Institute of Technology Madras . which is shown below in Fig.42.30 As shown in the Fig 1.42.30 From the given spectrum. 1. Inverse Fourier transform of A : 200 sin c (100 t ) e − j 100 π t Inverse Fourier transform of B : 50 sin c ( 50 t ) e j 50 π t Hence xce ( t ) = 200 sin c (100 t ) e − j 100 π t + 50 sin c ( 50 t ) e j 50 π t xc ( t ) = 200 sin c (100 t ) cos (100 π t ) + 50 sin c ( 50 t ) cos ( 50 π t ) xs ( t ) = 50 sin c ( 50 t ) sin ( 50π t ) − 200 sin c (100 t ) sin (100 π t ) 1. we can take X ce ( f ) as the sum of A and B.

If y ( t ) is the output. Exercise 1. Ry ( τ ) = Rx ( τ ) ∗ Rh ( τ ) ∗ ⎤ = Rx ( τ ) ∗ ⎡ ⎣h ( τ ) ∗ h ( − τ )⎦ (1. Exercise 1. 1. then Ey ( f ) = Ex ( f ) H ( f ) 2 (1.63b) 1.5 .5 and 1. Then.Principles of Communication Prof.9.13 Find the expression for the envelope of x (t ) = ⎡ ⎣1 + k cos ( ωm t ) ⎤ ⎦ cos ( ωc t ) where k is a real constant and fc >> fm .12 Let x ( t ) = ex1( t ) sin ⎡ ⎣( ωc + ∆ ω) t ⎤ ⎦ where ωc >> ∆ ω . Find xce ( t ) .63a) Let H ( f ) 2 ←⎯ → Rh ( τ ) be the Fourier transform pair where Rh ( τ ) is the ACF of the impulse response h ( t ) .9 Bandpass (BP) Systems Let a signal x ( t ) be input to an LTI system with impulse response h ( t ) . Hint: use the result of Exercise 1.119 Indian Institute of Technology Madras .11 Find the pre-envelope of x ( t ) = sin c ( t ) . then Y ( f ) = X ( f ) H ( f ) and Y (f ) 2 = X (f ) 2 H (f ) 2 If E x ( f ) is the energy spectral density of x ( t ) . Sketch it for k = 0. Venkata Rao Exercise 1. V.

Venkata Rao Let x ( t ) be a bandpass signal with X ( f ) zero for f ± fc > W . outside ⎩ x ( t ) is the input to an LTI system with impulse response h ( t ) = x (T − t ) . we can treat h ( t ) as a NBBP signal.Principles of Communication Prof. 0 ≤ t ≤ 1 m sec ⎣ ⎦ { ( ) ( ) ( ) } ( ) Again. a system with passband in the interval f ± fc < B where B ≤ W . x ( t ) can be taken as a NBBP signal.64. ( ) where T = 1 m sec . Find y ce ( t ) and y ( t ) . Its complex envelope. In appendix A1.3 (Eq A1.64) We shall now give an example to illustrate the use of Eq. 1. that is. V.120 Indian Institute of Technology Madras .31 Let x ( t ) be a sinusoidal pulse given by ⎧ 2 π 106 t ⎤ . This study is again greatly facilitated if we were to use complex envelopes of the signals involved. 0 ≤ t ≤ 1 m sec ⎪2 cos ⎡ ⎣ ⎦ x (t ) = ⎨ ⎪ 0 . xce ( t ) = 2 .8).3. with 1. A bandpass signal is usually processed a bandpass system. We would like to study the effect of a BP system on a BP input. Example 1. it is shown that y ce ( t ) = 1 ⎡ xce ( t ) ∗ hce ( t ) ⎤ ⎦ 2⎣ (1. 0 ≤ t ≤ 1 m sec h ( t ) = 2 cos ⎡2 π 106 ⎣ ( ) (T − t )⎤ ⎦ ) ( ) ( ) ( )} = 2 cos 2 π 106 T cos 2 π 106 t + sin 2 π 106 T sin 2 π 106 t { ( = 2 cos 2 π × 103 cos ⎡2 π 106 t ⎤ + sin 2 π × 103 sin 2 π × 106 t ⎣ ⎦ = 2 cos ⎡2 π 106 t ⎤ .

Principles of Communication

Prof. V. Venkata Rao

⎧2 , 0 ≤ t ≤ 1 m sec hce ( t ) = ⎨ ⎩0 , outside ⎛t −T ⎞ 2⎟ That is, xce ( t ) = hce ( t ) = 2 ga ⎜ ⎜ T ⎟ ⎝ ⎠ From Eq.1.64, y ce ( t ) = 1 ⎛t −T ⎞ ⎡ xce ( t ) ∗ hce ( t ) ⎦ ⎤ = 2 × 10− 3 tri ⎜ ⎟ ⎣ 2 ⎝ T ⎠

(

)

y ( t ) = y ce ( t ) cos ( 2 π fc t )

Obtaining y ( t ) directly as ⎡ ⎣ x ( t ) ∗ h ( t )⎤ ⎦ would be quite cumbersome.

Exercise 1.14

Let x ( t ) = sin c ( 200 t ) cos ⎡ 2 π 106 t ⎤ be the input to an NBBP ⎣ ⎦ system with H ( f ) = 1 ⎡ H1 f − 106 − H1 f + 106 ⎤ ⎦ 2j ⎣

( )

(

)

(

)

(1.65)

where H1 ( f ) is as shown in Fig. 1.43.

Fig. 1.43: H1 ( f ) of Eq. 1.65 Find y ce ( t ) , the complex envelope of the output.

1.121
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Appendix A1.1 Tabulation of sinc ( λ )
λ 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 1.60 1.65

sinc ( λ ) 1.000000 0.995893 0.983631 0.963397 0.935489 0.900316 0.858393 0.810331 0.756826 0.698645 0.636619 0.571619 0.504550 0.436331 0.367882 0.300104 0.233871 0.170010 0.109291 0.052414 -0.000001 -0.047424 -0.089422 -0.125661 -0.155915 -0.180064 -0.198091 -0.210086 -0.216236 -0.216821 -0.212203 -0.202833 -0.189207 -0.171888

λ 1.70 1.75 1.80 1.85 1.90 1.95 2.00 2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.45 2.50 2.55 2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05 3.10 3.15 3.20 3.25 3.30 3.35

sinc ( λ ) -0.151481 -0.128616 -0.103943 -0.078113 -0.051770 -0.025536 0.000000 0.024290 0.046840 0.067214 0.085045 0.100035 0.111964 0.120688 0.126138 0.128323 0.127324 0.123291 0.116435 0.107025 0.095377 0.081847 0.066821 0.050705 0.033919 0.016880 0.000000 -0.016326 -0.031730 -0.045876 -0.058468 -0.069255 -0.078036 -0.084661

λ

sinc ( λ )

3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.05 4.10 4.15 4.20 4.25 4.30 4.35 4.40 4.45 4.50 4.55 4.60 4.65 4.70 4.75 4.80 4.85 4.90 4.95 5.00

-0.089038 -0.091128 -0.090946 -0.088561 -0.084092 -0.077703 -0.069600 -0.060021 -0.049237 -0.037535 -0.025222 -0.012607 -0.000000 0.012295 0.023991 0.034821 0.044547 0.052960 0.059888 0.065199 0.068802 0.070650 0.070736 0.069097 0.065811 0.060993 0.054791 0.047385 0.038979 0.029796 0.020074 0.010059 0.000000

1.122
Indian Institute of Technology Madras

Principles of Communication

Prof. V. Venkata Rao

Fig. A1.1: Plot of sin c ( λ ) for λ ≤ 5

1.123
Indian Institute of Technology Madras

∗ T0 1 Rx p y p ( τ ) = T0 ∞ − T0 ∫ 2 2 ⎡ x (t ) ⎢ ⎢ ⎣n T0 / 2 −T0 / 2 ⎤ * − τ − y t nT ∑ ( 0 ) ⎥ dt ⎥ = −∞ ⎦ ∞ = n = −∞ ∑ 1 T0 ∫ x ( t ) y * ( t − τ − nT0 ) dt As x ( t ) = 0 for t > T0 . − T0 T < t < 0 2 2 outside then. 2 1. − 0 < t < 0 where y ( t ) = ⎨ 2 2 ⎪ outside ⎩ 0 . T0 1 Rx p y p ( τ ) = T0 − T0 ∫ 2 ∗ xp (t ) y p ( t − τ ) dt 2 Let ⎧ ⎪x (t ) . x (t ) = ⎨ p ⎪ ⎩ 0 . V.Principles of Communication Prof. Venkata Rao Appendix A1. x p ( t ) = n = −∞ ∑ ∞ x ( t − nT0 ) Similarly.124 Indian Institute of Technology Madras . We have only to decide the weights of these impulses.2 Fourier transform of Rxp y p ( τ ) As Rx p y p ( τ ) is periodic. y ∗ p (t − τ) = n = −∞ ∑ ∞ y ∗ ( t − τ − nT0 ) T T ⎧ ∗ ⎪y p (t ) . we expect the spectrum to be purely impulsive.

125 Indian Institute of Technology Madras . we get the dual relation m = −∞ ∑ ∞ δ ( f − m f0 ) = 1 ⎡ ⎢ f0 ⎢ m ⎣ ∑ = ∞ ⎤ e j 2 π m f T0 ⎥ ⎥ −∞ ⎦ As f0 = 1 . Venkata Rao Rx p y p ( τ ) = = 1 T0 1 T0 n = − ∞ −∞ ∑ ∞ ∞ ∗ ∫ x ( t ) y ( t − τ − nT0 ) dt n = −∞ ∑ ∞ Rx y ( τ + nT0 ) Taking the Fourier transform on both sides. 1 ⎡ F⎢ F ⎡R x p y p ( τ ) ⎤ = ⎣ ⎦ T0 ⎢ n ⎣ ⎤ R nT τ + ( ) ∑ xy 0 ⎥ ⎥ = −∞ ⎦ ∞ As Rx y ( τ ) ←⎯ → X (f ) Y ∗ (f ) .2.Principles of Communication Prof.2. T0 by f0 .2.1) But from Example 1. 1 ⎡ F ⎡R x p y p ( τ ) ⎤ = ⎢ ⎣ ⎦ T0 ⎢ n ⎣ ⎤ j 2 π n f T0 ∗ X f Y f e ⎥ ( ) ( ) ∑ ⎥ = −∞ ⎦ ∞ = 1 ⎡ ⎢ X (f ) Y ∗ (f ) T0 ⎢ ⎣ n ⎤ j 2 π n f T0 e ⎥ ∑ ⎥ = −∞ ⎦ ∞ (A1. V.2) = 1 ⎡ ⎢ T0 ⎢ m ⎣ ∑ = ∞ ⎤ e j 2 π m f0 t ⎥ ⎥ −∞ ⎦ Replacing t by f . we have T0 ⎛ m⎞ ∑ δ⎜f − T ⎟ = 0⎠ = −∞ ⎝ ∞ 1 T0 m m = −∞ ∑ ∞ e j 2 π m f T0 (A1.3) 1.12. we have ⎡ −1 1 t mT F δ − = ⎢ ( ) ∑ 0 T ⎢ = −∞ ⎣ 0 ∞ m m ⎤ f n f δ − ( ) ∑ 0 ⎥ ⎥ = −∞ ⎦ ∞ (A1.

1 ⎡ F ⎡R x p y p ( τ ) ⎤ = 2 ⎢ ⎣ ⎦ T0 ⎢ ⎣m ⎛m⎞ ⎛m⎞ ∑ X ⎜T ⎟ Y∗ ⎜T ⎟ ⎝ 0⎠ ⎝ 0⎠ = −∞ ∞ ⎛ m ⎞⎤ δ⎜f − ⎟⎥ T0 ⎠ ⎥ ⎝ ⎦ (A1. V.2. Venkata Rao Using Eq. A1.2.3 in Eq. 1. we obtain.4) ⎛m⎞ where X ⎜ ⎟ = X ( f ) f ⎝ T0 ⎠ ⎛m⎞ and Y ∗ ⎜ ⎟ = Y ∗ ( f ) f ⎝ T0 ⎠ = m T0 = m T0 . A1.Principles of Communication Prof.126 Indian Institute of Technology Madras .2.1.

3 Complex envelope of the output of a BP system Let x ( t ) .3.1) hce ( t ) = hc ( t ) + j hs ( t ) .6 becomes X (f ) H (f ) = 1 ∗ ∗ ⎡Hce ( f − fc ) + Hce ⎤ ⎡ X ce ( f − fc ) + X ce ⎤ ⎡ ⎡ ⎣ − ( f + fc ) ⎤ ⎦⎦ ⎣ − ( f + fc ) ⎤ ⎦⎦ ⎣ 4 ⎣ { } Consider the product term 1.Principles of Communication Prof. Venkata Rao Appendix A1. which is also a BP signal. X (f ) = Similarly.2) (A1. We shall derive a relation for y c e ( t ) in terms of xc e ( t ) and hc e ( t ) .3.6) Because of Eq.4. A1.4) Taking the FT of Eq. V.3.5) But j 2 π fc t ⎤ y ( t ) = Re ⎡ ⎣ y ce ( t ) e ⎦. We know that. a BP signal. h ( t ) = hc ( t ) cos ωc t − hs ( t ) sin ωc t .3.3. Therefore.5. (A1. let 1 ∗ ⎡ X ce ( f − fc ) + X ce ⎣ − ( f + fc ) ⎤ ⎦ .3.3. we have ∗ ⎡ 2 H ( f ) = Hce ( f − fc ) + Hce ⎣ − ( f + fc ) ⎤ ⎦ (A1. j 2 π fc t ⎤ h ( t ) = Re ⎡ ⎣hce ( t ) e ⎦. Y (f ) = X (f ) H (f ) = ∗ ⎡ Yce ( f − fc ) + Yce ⎣ − ( f + fc ) ⎤ ⎦ 2 (A1. ∗ 2 h ( t ) = hce e j 2 π fc t + hce e− j 2 π fc t (A1. be applied as input to a BP system with impulse response h ( t ) . A1.3.127 Indian Institute of Technology Madras . A1. j 2 π fc t ⎤ x ( t ) = Re ⎡ ⎣ xce ( t ) e ⎦. 2 { } (A1.3) . xce ( t ) = xc ( t ) + j xs ( t ) . x ( t ) = xc ( t ) cos ωc t − xs ( t ) sin ωc t . Let the resulting output be denoted y ( t ) .3. Eq.3.1 and A1.

3. 2 4 ⎣ and 1 ∗ 1 ∗ ∗ ⎡Hce Yce ⎡ − ( f + fc ) ⎤ = − ( f + fc ) ) X ce ( ( − ( f + fc ) )⎤ ⎣ ⎦ ⎣ ⎦. 1 ⎡ xc ( t ) + j xs ( t ) ⎤ ⎦∗⎡ ⎣hc ( t ) + j hs ( t ) ⎤ ⎦ 2 ⎣ 1 {xc ( t ) ∗ hc ( t ) − xs ( t ) ∗ hs ( t )} 2 1 {xc ( t ) ∗ hs ( t ) + xs ( t ) ∗ hc ( t )} 2 { } (A1. Hce ⎣ − ( f + fc ) ⎤ ⎦ X ce ( f − fc ) = 0 .3.3. ∗ ⎡ That is. y ce ( t ) = From Eq. A1.128 Indian Institute of Technology Madras . fc + B ) . Hence. y ce ( t ) = Therefore. ∗ ⎡ Yce ( f − fc ) + Yce ⎣ − ( f + fc ) ⎤ ⎦ = 1 H f −f X f −f ce ( c) ce ( c) 2 4 1 ∗ ∗ ⎡ + − ( f + fc ) ⎤ Hce ⎡ X ce ⎣ ⎦ ⎣ − ( f + fc ) ⎤ ⎦ 4 Yce ( f − fc ) has nonzero spectral components only in the range ( fc − B .3. 1 1 ⎡Hce ( f − fc ) X ce ( f − fc ) ⎤ Yce ( f − fc ) = ⎦.7) (A1. 1.9) (A1.Principles of Communication Prof. the spectra Hce ( f − fc ) and X ce ⎣ − ( f + fc ) ⎤ ⎦ do not overlap.3. Venkata Rao ∗ ⎡ Hce ( f − fc ) X ce ⎣ − ( f + fc ) ⎤ ⎦ ∗ ⎡ Hce ( f − fc ) has spectrum confined to the range ( fc − B . − ( fc − W )} . That is. V.10) y ce ( t ) = y c ( t ) + j y s ( t ) . fc + B ) . Similarly.8) and ys (t ) . yc (t ) = ys (t ) = and. Yce ( f ) = Therefore. 1 ⎡ xce ( t ) ∗ hce ( t ) ⎤ ⎦ 2⎣ we obtain the equations for yc (t ) 1 X ce ( f ) Hce ( f ) 2 (A1.zero spectral components only in the range {− ( fc + W ) . 2 4 In other words. X ce ⎣ − ( f + fc ) ⎤ ⎦ has non.8. hence the ∗ ⎡ product is zero.

Venkata Rao Exercise A1. 1.3.Principles of Communication Prof. hs ( t ) ⎤ ⎦ suggest a scheme to recover y c ( t ) and y s ( t ) .1 Given the pairs ( xc ( t ) . V.129 Indian Institute of Technology Madras . xs ( t ) ) and ⎡ ⎣ hc ( t ) .

Thomson Asia Pvt. Modern Digital and Analog Communication Systems (3rd Edition). Carlson. Brooks/Cole Publishing Company. Couch II. Oxford University Press.. S. V. Lathi. B. A. V. Oppenheim. Mc Graw-Hill. PHI. B. S. P. with Hamid Nawab. 1998 Note: The above three books have a large collection of problems. 1997 2. Communication Systems (4th Edition). The student is advised to try to solve them.. Singapore. 1998 1. Lathi.. B. Ashok Ambardar. W. Other Suggested Books 1.. A. 1999 3. Signals and Systems (2nd Edition). Signal Processing and Linear systems. Willisky. Ltd. Analog and Digital Signal Processing (2nd Edition). Venkata Rao References 1.Principles of Communication Prof. Berkeley-Cambridge Press. A. P. 2003 3... 2001 2. L.130 Indian Institute of Technology Madras . Digital and Analog Communication Systems (6th Edition) Pearson Asia.