Signals Basics

Tapas Kumar Bhattacharya

Department of Electrical Engineering

I.I.T Kharagpur
July 27, 2019

S
PA
TA

1
Tapas K Bhattacharya Introducing Signals 2

1 Introduction
We shall in this section consider signals which vary with time. Value of the signal at different time
may be shown graphically or if we are lucky the signal can be represented mathematically in terms
of an equation.

1.1 Time Reversal: To get x(−t) from x(t)

If we know, x(t) then mathematically x(−t) can be obtained by replacing t with −t in x(t). Nat-
urally, both x(t) and x(−t) can be plotted. Essentially x(−t) is nothing but the mirror image of
x(t) about the vertical axis. Also it can be interpreted as folding x(t) about the vertical axis to get
x(−t). The idea is illustrated in self explained figure 1(a) and (b).

+1 2

f(−t) f(t)
f(−t) f(t)

t T/2 t
o O T
−1 1
(a) (b)

Figure 1: Getting x(−t) from x(t)

S
PA
1.2 Types of signals
Usually these type of signals can be nicely expressed in terms of equations as shown in figure 2
TA

v(t) Vm
x(t)

3T/2 t t
o 2T o
T/2 T

(b) Exponentially decaying sinusoid

(a) Sinusoidal

Figure 2: Continuous time signals

Tapas K Bhattacharya Introducing Signals 3

Some more functions or signals, called the singularity functions having discontinuities at certain
points are shown in figure 3.

δ( t)
u(t) r(t)
1
0
0
1
0
1
0
1 1 1
0
1 __
_1
0
0
1 h
0o+
o 1
t
0
1
o o
t=0 t t

h Unit ramp at origin

Unit step at origin
Unit impulse at origin

δ( t−τ) u(t −τ ) r(t −τ )

1
0
0
1
0
1
0
1 1
0
1
0
1
0
1
0
1 t
0
1
ο τ o
τ o
t τ t
Shifted unit impulse
Shifted unit step Shifted ramp at t= τ
at t = τ
S at t = τ

Figure 3: Singularity functions

PA
TA

1.3 Properties of signals

2 Singularity functions
In this section we shall take up most useful singularity functions one at a time and discuss about
its properties.

2.1 Impulse or Delta function

In figure 4(a) & (b), an unit impulse function is shown. The function is shown as thin rectangular
pulse, centered around the the origin. The width of the pulse is shown to be h and it’s height to
be 1/h. The area enclosed by the pulse is unity, no matter whatever is the value of h you choose.
This general rectangular pulse will become an unit impulse function as h → 0 i.e., h = 0+ − 0− ,
where 0+ and 0− are small perturbations in time in the negative and positive direction of time axis
around t = 0. So the the height of the pulse tends to infinity and width tends to zero. Thus no
Tapas K Bhattacharya Introducing Signals 4

δ( t)

δ( t)
1
__
_ h
o o+ t t
t=0

h
(a) Unit impulse at origin (b) Simplified of representing δ( t)

δ( t−τ) δ( t−τ)

_ +
τ τ
ο τ ο τ t
t

(c) Shifted unit impulse at t = τ S (d) Representation of shifted impulse

Figure 4: Impulse function

PA
point in talking about the functional value of this function at t = 0 as it is infinitely large. However
TA

area under the impulse is finite and equal to unity and this can be mathematically described as:
Z ∞ Z 0+
δ(t) dt = δ(t) dt = 1 since elsewhere, δ(t) = 0
−∞ 0−

The area enclosed is the strength of the impulse. For example 2δ(t) represents an impulse of strength
2. Similarly for a shifted delta function δ(t − τ ) as shown in figure 4(c) & (d), following relation
holds good.
Z ∞ Z τ+
δ(t − τ ) dt = δ(t − τ ) dt = 1 since, δ(t − τ ) = 0 for, t 6= τ
−∞ τ−

The consequence of this is that when you multiply a continuous function x(t) with δ(t), following
relations are true.
x(t)δ(t) = x(0)δ(t) and x(t)δ(t − τ ) = x(τ )δ(t − τ )
Tapas K Bhattacharya Introducing Signals 5

We can generate any continuous arbitrary function x(t), using delta function as follows:
Z ∞
Consider the integral x(τ )δ(t − τ ) dτ
−∞
Z ∞
= x(t)δ(t − τ ) dτ
−∞
Z ∞
= x(t) δ(t − τ ) dτ
−∞
= x(t)

Thus finally, Z ∞
x(t) = x(τ )δ(t − τ ) dτ
−∞

Representation of unit impulse function is shown with a arrowed line as depicted in figure 4(c)

2.1.1 Time scaling of impulse function : δ(at)

1
It will be shown here that δ(at) = |a|
δ(t) where a could be +ve or -ve real numbers. Let us calculate
the integral: S Z +∞
I= x(t)δ(at) dt
PA
−∞

Now,
TA

Z +∞
I = x(t)δ(at) dt
−∞
substitute at = τ and assume a > 0
Z +∞ Z +∞
δ(τ ) δ(τ )
or, I = x(τ /a) dτ = x(0) dτ
−∞ a −∞ a
Z +∞ Z +∞
δ(τ )
thus, I = x(t)δ(at) dt = x(τ ) dτ
−∞ −∞ a
1
Comparing we get, δ(at) = δ(t)
a
1
We can repeat the above steps when a is -ve and will arrive at δ(at) = |a|
δ(t) which will be true for
both +ve and -ve values of a.
Tapas K Bhattacharya Introducing Signals 6

2.1.2 What about : δ(at + b)?

Now     
b 1 b
δ(at + b) = δ a t + = δ t+
a |a| a
This can be easily proved starting from basic integral as follows:
Z +∞ +∞   
b
Z
I = x(t)δ(at + b) dt = x(t)δ a t + dt
−∞ −∞ a
b
Substitute t + = τ and proceed to get the result shown below
a     
b 1 b
δ(at + b) = δ a t + = δ t+
a |a| a
1
which naturally means, δ [a(t + c)] = δ (t + c)
|a|

2.1.3 What about : x(t) d{δ(t)}

dt
?

It will be shown here that x(t) d{δ(t)}

dt
= − dx
dt
(0) δ(t) = −ẋ(0)δ(t) We start with the integral :
S Z +∞
[d(δ(t)]
PA
I= x(t) dt
−∞ dt

Integrating the above by parts considering the first function to be x(t) and the second function to
TA

be d{δ(t)}
dt
.
+∞
{d(δ(t)}
Z
I = x(t) dt
−∞ dt
Z +∞ Z +∞
+∞ dx dx
or, I = x(t)δ(t)|−∞ − δ(t) dt = − (0)δ(t) dt
−∞ dt −∞ dt
Z +∞ Z +∞
{d(δ(t)} dx dx
∴ x(t) dt = − (0)δ(t) dt = − (0)
−∞ dt −∞ dt dt
R +∞
2.1.4 What about : −∞
x(t)δ{f (t)} dt?
If an impulse function is described as δ{f (t)} where f (t) is a function having several real roots, say
a1 , a2 · · · ai etc. Therefore, it is expected we shall have a number of impulses occurring at those
roots. However the strengths of those impulses may not be the same.
Tapas K Bhattacharya Introducing Signals 7

Consider a general root ai which makes f (ai ) = 0. For this general root ai , we shall first show
that Z +∞ Z +∞
δ(t − ai ) x(ai )
x(t)δ{f (t)} dt = x(t) =
−∞ −∞ |f˙(ai )| |f˙(ai )|
Which essentially means that
δ(t − ai )
δ{f (t)} =
|f˙(ai )|
Now the value of the function f (t), very close to ai i.e., (t − ai ) → 0 can be obtained from Taylor
series expansion as follows:

f (t) ≈ f (ai ) + f˙(ai )(t − ai ) neglecting higher order terms

or, f (t) = f˙(ai )(t − ai ) since f (ai ) = 0

Therefore for the ith root we can write

Z +∞ Z +∞
x(t)δ{f (t)} dt = x(t)δ{f˙(ai )(t − ai )} dt
−∞ −∞
δ(t − b)
Recall that δ{a(t − b)} = use this in above equation to get
S |b|
Z +∞ Z +∞
δ(t − ai ) x(ai )
PA
˙
x(t)δ{f (ai )(t − ai )} dt = x(t) dt = for a single root ai
−∞ −∞ |f˙(ai )| |f˙(ai )|
Z +∞ X x(ai )
For all the roots x(t)δ{f (t)} dt =
TA

−∞ i
|f˙(ai )|

Thus δ{f (t)} is nothing but a collection of as many impulses as the number of real roots with
different strengths. That is
X δ(t − ai )
δ{f (t)} =
i
|f˙(ai )|

2.2 Unit step function

Unit step function shown in figure 5(a), can be described mathematically as follows:
(
0, if t < 0;
u(t) =
1, if t > 0;

Note that u(t) is not defined at t = 0. Similarly mathematical description of the shifted unit step
Tapas K Bhattacharya Introducing Signals 8

u(t) 5 u(t) u(t −τ )

1 5 1

o o o
t t τ t
(c) Shifted unit step
(a) Unit step at origin (b) Step function of value 5
at t = τ

Figure 5: Step function

function shown in figure 5(c) is as follows:

(
0, if t < τ ;
u(t − τ ) =
1, if t > τ ;

Here also the step function is not defined at t = τ . Figure 5(b), shows a step function of strength
5, i.e., it shows the plot of 5u(t).

2.3 Relationship between impulse & step function

Here we shall show that by integrating a delta function we shall get the step function and by
S
differentiating a step function, delta function can be obtained.
PA
Z t
consider the integral δ(τ ) dτ
TA

−∞
Z t
and note δ(τ ) dτ = 0 if t < 0
−∞
= 1 if t > 0
Z t
∴ δ(τ ) dτ = u(t)
−∞
du
We can also conclude from this, = δ(t)
dt
Tapas K Bhattacharya Introducing Signals 9

2.4 Ramp function

r(t) r(t −τ )
2 r(t)
1 1 2

o 1 o o
t t 1 t
1
(a) Unit ramp at origin
(b) Shifted ramp at t= τ (c) A ramp of 2 r(t)

Figure 6: Ramp functions

An unit ramp function is shown in figure 6(a) and it is mathematically described as :

(
0, if t ≤ 0;
r(t) =
t, if t ≥ 0;
In the same way, a shifted unit ramp, shown in figure 6(b) can be described as follows:
(
0, if t ≤ τ ;
r(t) =
t − τ, if t ≥ τ ;
S
Finally a ramp with a higher slope 2r(t) is shown in figure 6(c) and it’s mathematical description
PA
is: (
0, if t ≤ 0;
r(t) =
2t, if t ≥ 0;
TA

2.5 Relationship between step & ramp function

The relationships are obvious and mentioned below. If an unit step function is integrated, unit
ramp will be obtained and if a ramp is differentiated, step function will be obtained.
Z t
r(t) = u(t) dt
−∞
dr
u(t) =
dt
Z t
also earlier, we have already got δ(τ ) dτ = u(t)
−∞
du
= δ(t)
dt
The above relations can be visually more appealing as shown in figure 7
Tapas K Bhattacharya Introducing Signals 10

r(t)
1
Unit ramp at origin

o 1 t
u(t)
dr
1 u(t) = __
dt
Unit step at origin
o
t

du
δ( t) = __
dt
δ( t) Unit impulse at origin
t
o

Figure 7: Relationship among Ramp,unit & impulse functions

2.6 Expressing a given signal in terms of singularity functions

Suppose a given signal x(t) is multiplied with u(t), then it means a signal whose value will be zero
S
for t < 0, since u(t) = 0 for t < 0 and it will be x(t) for t > 0 as u(t) = 1 for t > 0. In figure 8(a),
PA
we have plotted x(t) sin ωt and its range is −∞ < t < ∞. If now this x(t) is multiplied by u(t),
the resulting function will become as shown in figure 8(c). The usefulness of the step and ramp
functions are shown in the following example.
TA

Example
Write down the description of x1 (t) and x2 (t) shown in figure 9(a) and (b).
The expressions are written by inspection of the functions as:

x1 (t) = u(t) + u(t − 1) − u(t − 2) − u(t − 3)

1
x2 (t) = r(t) − r(t − 1) − 2u(t − 3) + r(t − 3)
2
It should be noted that both the functions x1 (t) and x2 (t) can be expressed, if we like, elaborately
Tapas K Bhattacharya Introducing Signals 11

x(t) = 2 sin ω t

(a)
T/2 t
O
−T/2 T

u(t)
1
(b)

O t
x (t) = 2 sin ωt u(t)
1
2

(c)

T/2 t
O T

Figure 8: Sinusoidal signal multiplied with step function

S
PA
in different time zones. For f1 (t), it will look like:

 0, if t ≤ 0;
TA



1, if 0 < t < 1;



x1 (t) = 2, if 1 < t < 2;

1, if 2 < t < 3;





0, if t > 3;
We can easily see the advantage of using step and ramp functions for expressing a given function
x(t).

3 Even & odd functions

Any function may be either even or odd or neither even or odd. If x(t) = x(−t), then the function
is said to be even and if x(t) = −x(−t), then the function is said to be odd. If function x(t), does
not satisfies any of the above condition, it is said to be neither odd nor even. Look at figure 10,
where an even, odd and a general functions are shown.
Tapas K Bhattacharya Introducing Signals 12

x (t)
1 x (t)
2 2

1 1 1

t 4 5 t
o 1 2 3 o 1 2 3
−1
(a)
(b)

Figure 9: Examples with step & ramp functions.

x(t)
+1

x(t) x(t)
+1

t T t −T o t
o o
−T T −T T

−1

(a) Even function (b) Odd function (c) Neither odd nor even function

Figure 10: Even, odd & a general function

S
One can easily verify that cosine function (cos t) is even while a sine function (sin t) is odd. It
PA
is interesting to note that a general function x(t) can be shown to be a sum of an even function and
an odd function. The even part and the odd part can be easily calculated as follows.
TA

Let, x(t) = xeven (t) + xodd (t)

Then, x(−t) = xeven (−t) + xodd (−t)
Now applying properties of even & odd functions:
or, x(−t) = xeven (t) − xodd (t)
Now manipulating the first & third equations:
x(t) + x(−t)
xeven (t) =
2
x(t) − x(−t)
xodd (t) =
2

Example
Is the function x(t) = 2t + 3 is even, odd or neither (general)?, If the answer is neither, find out the
even and odd part of the function.
Tapas K Bhattacharya Introducing Signals 13

Since x(t) 6= x(−t) or x(t) 6= −x(−t), we conclude that the function is general.

x(t) + x(−t) 1
xeven (t) = = (2t + 3 − 2t + 3) = 3
2 2
x(t) − x(−t) 1
xodd (t) = = (2t + 3 + 2t − 3) = 2t
2 2
The problem just solved, a bit trivial in the sense that from the original function itself we could
make out the answer.

3.1 Time scaling of a function

If a function x(t) is known, then we shall be able to know x(at) where a is a factor by which the
time axis has been scaled. In essence, we have to replace t in x(t), by at in order to get x(at).
Visual effect of time scaling of a signal x(t) is shown in 11.

x(t)
+1
x(2t)
x(t/2)

S t
PA
o 1 2
−2 −1 −1/2 1/2

Figure 11: Time scaling of a signal

TA

The triangular pulse x(t) is shown by firm line and its mathematical expression is

x(t) = t + 1 for, − 1 ≤ t ≤ 0
x(t) = 1 − t for, 0 ≤ t ≤ 1

Now let us now try to find out x(2t). Replace t by 2t in the above equation.

x(2t) = 2t + 1 for, − 1/2 ≤ t ≤ 0

x(t) = 1 − 2t for, 0 ≤ t ≤ 1/2

The effect of time scaling with a = 2, is to make the triangular pulse x(2t) compressed between
-1/2 to +1/2 as shown in figure 11 by dashed line. In the same way, if we choose scaling factor to
be a = 1/2, the resulting waveform of x(t/2) will be expanded between -2 to +2 as shown by the
chain-dotted line in figure 11.
Tapas K Bhattacharya Introducing Signals 14

Example-1
A function x(t) is shown in figure 12(a), plot x(−2t − 2).
It may be noted that all the three operations namely time shift,time scaling and time reversal
operations to be executed on x(t). Follow the following steps to get x(−2t − 2) from x(t).
1. Do right shift operation to get x(t − 2) from x(t) as shown in figure 12(b).
2. Now do time time scaling on x(t − 2) to get x(2t − 2) as shown in figure 12(c).
3. Finally do time reversal ( or folding ) operation on x(2t − 2) to get x(−2t − 2) as shown in
12(d)

x(t)

(a) original signal

t
o
1 2 3 4 5 6 7 8

x(t−2)
(b) Do time shifting on x(t)
and get x(t−2)
S t
PA
o
1 2 3 4 5 6 7 8

x(2t − 2)
TA

(c) Now time scale on x(t−2)

to get x(2t − 2)

t
o
1 2 3 4 5 6 7 8

x(−2t − 2)
(d) Now do time reversal
or folding operation on x(2t − 2)
to get finally x(−2t − 2)

t
o
−4 −3 −2 −1 1 2 3 4 5 6 7 8

Figure 12: Time scaling of a signal

As discussed in the class, we could also get x(−2t − 2) from x(t) following some other rules.
End result will of course be same. I prefer the rule : Shift → Scale → Reversal.
Tapas K Bhattacharya Introducing Signals 15

4 Energy and Power of a signal

A signal may be also classified as energy signal or power signal or neither energy or power signal.
Here we shall discuss how the energy and power associated with a given signal are calculated.

4.1 Energy of a signal

For a signal x(t), total energy of the signal is defined as:
Z T
E = lim |x(t)|2 dt
T →∞ −T

If E is found to be finite and non-zero, then the signal is said to be energy signal.

4.2 Power of a signal

For a signal x(t), average power of the signal is defined as:
T
1
Z
P = lim |x(t)|2 dt
ST →∞ 2T −T
PA
If P is found to be finite and non-zero, then the signal is said to be power signal.
Example-1 Is the unit step function u(t) an energy or a power signal?
Solution: Let us first calculate energy associated with u(t).
TA

Z T
E = lim |x(t)|2 dt
T →∞ −T
ZT
= lim |u(t)|2 dt
T →∞ −T
ZT
= lim dt
T →∞ −T
lim E = (2T )
T →∞
∴ E → ∞
so, u(t) is not an energy signal
Tapas K Bhattacharya Introducing Signals 16

Now let us calculate power associated with u(t).

Z T
1
P = lim |x(t)|2 dt
T →∞ 2T −T
Z T
1
= lim |u(t)|2 dt
T →∞ 2T −T
Z T
1
= lim dt
T →∞ 2T −T
1
lim P = (2T ) = 1
T →∞ 2T
∴ P is finite
so, u(t) is a power signal

Example-2 Is the function sin 5t an energy or a power signal?

Solution: Let us first calculate energy associated with sin ωt.
Z T
S E = lim sin2 5t dt
T →∞ −T
PA
Z T
= lim (1 − cos 10t) dt
T →∞ −T
Z T
TA

= lim dt
T →∞ −T
= lim (2T )
T →∞
Z T
∴ lim sin2 5t dt → ∞
T →∞ −T
so, sin 5t is not an energy signal
Tapas K Bhattacharya Introducing Signals 17

Let us now calculate the power associated with sin 5t.

Z T
1
P = lim sin2 5t dt
T →∞ 2T −T
Z T
1
= lim (1 − cos 10t) dt
T →∞ 4T −T
Z T
1
= lim dt
T →∞ 4T −T
1 1
= lim (2T ) = finite
T →∞ 4T 2
T
1
Z
∴ lim sin2 5t dt =
T →∞ −T 2
so, sin 5t is a power signal

S
PA
TA