1

DIGITAL SIGNAL PROCESSING

Lecture 1 - Chapter 1

Classification of Signals: Continuous-Time verses Discrete-Time Signals

Continuous time or analog signals are signals that are defined for every value of a < t <
-|t|
b, where (a, b) can be ( - ¥, + ¥ ), i.e., x (t) = e or x(t) = cos(pt ). .

Discrete-time signals are defined at discrete-time instants and between the two discrete
time instants are undefined but are not zero. They can be obtained either by sampling
analog signals or they can be discrete in nature like discrete measurement signals.

A discrete-time signal having a set of discrete values is called a digital signal. Note that
sampling an analog signal produces a discrete-time signal. Then quantization of its values
produces a digital signal.

Deterministic versus Random Signals

Any signal that can be uniquely described by an explicit mathematical expression or a
well-defined rule is called “deterministic”. The past, present and future of a deterministic
signal are known with certainty. Otherwise, it is called “Random” and its properties is
explained by statistical techniques.

Review of Sinusoids in Continuous and Discrete Time

xa (t )= Acos(Wt +q ) - ¥ < t < ¥
= A cos(2pFt +q ) W = 2pF
xa (t + Tp )= xa (t ), Tp
= 1 : fundamental period. Increasing F means increasing

F
oscillation in time domain. F = 0 corresponds to Tp = ¥ .
 2

j ( W t +q )
Also, for complex exponential signals, xa (t ) = Ae . A sinusoidal signal then can
also be expressed as
1 j(W t +q ) 1 - j (W t +q )
xa (t )= A cos(W t +q )= Ae + Ae
2 2
Discrete-Time Sinusoid Signals
x(n)= A cos(w0 n +q ) w = 2p f
x(n)= A cos(2p f 0 n +q ) - ¥ < n < ¥

A few important differences between continuous sinusoid and discrete sinusoids:

1) A discrete-time sinusoid is periodic only if its frequency is a rational number.
By default, x(n + N )= x(n) for all n if x(n) is periodic. The smallest N is called
Fundamental Period.
x(n + N )= Acos(2p f 0 (n + N )+q ) = cos(2p f 0 n + q )

This relationship is true if and only if 2pf 0 N = 2Kp

K
Þ f0 = : a rational number.
N
To determine the period N of a periodic discrete time sinusoid, we express f as two
relatively prime numbers. Observe that a small change in frequency can result in a
è ø
large change in period. For example, x1 (n)=

period is N1 =K = 2 = 30 .
f1 60
æ æ 31 ö ö
Now consider x2 (n) = A cos ç 2p ç
ç
÷n + q ÷, ÷

è 60 ø
 æ æ 1ö
Acos ç2p ç
ö
÷n + q ÷ = Acos p n +q
( ) its
ç ÷

è è2 ø ø 31
f2= = 0.517 ® N 2 = 60
60

2. An analog F (- ¥,+¥) maps to - 1 £ f £ 1 or equivalently to - p £ w £ p or in

2 2
other words, the highest rate of oscillation occurs at f = + 1 or w = +p . To see

2
what happens for p £ w £ 2p consider w1 = w0 and w2 = 2p - w0 . When w1 varies
between p to 2p, then w2 varies between p and 0. Now
x1 (n)= Acosw1 n = A cosw0 n
 3

x2 (n) = Acos(w2 )n = A cos(2p - w0 )n =* Acos(-wo n)= Acoswon = x1

(n) * this is only true because n is an integer value, i.e., x(n) is a discrete signal.
Hence, x2 (n) = x1 (n) is an alias because - ¥ < F < +¥ maps only to - 1£f£ 1
2 2
by sampling.
Analog to Digital Conversion (A/D)
- sampling (sampling rate)
- quantization
x(n) = xa (nTs ), where - ¥ < n < ¥
Sampling:
x (t )
= a t=nTs

t = nT = n , Fs = Sampling rate (Frequency) (Hz)

s
Fs
Consider an analog sinusoid: x (t )= A cos(2p Ft +f )
a
x(n)= xa (nTs )= A cos(2p F nTs + f) = æ F ö
A cosç 2pç
n+f÷ ÷

è F ø
s
Now recall that - ¥ < F < ¥ maps to - 1 £f£ 1
2 2
-1 £f= F £ 1
® ® FF³³22FF Nyquist rate
ss maxmax
2 Fs 2
Example 1: x (t )= cos 20p t Þ F1 = 10Hz
a1
xa2 (t )= cos100p t Þ F2 = 50Hz
If we digitize both of these signals with Fs = 40Hz, then

x (n)= x (nT )= x
æ n ö= cos 20p n = cos p
ç ÷ n
1 a1 s a1 è 40 ø 40 2
(n)= x æ n ö 100p 5p æ pö
x2 a ç ÷ = cos n = cos n = cos 2p +
ç ÷n
2 4 40 2 2
è0 ø è ø
= cos p n = x (n)!!
2 1

Therefore, by this sampling rate x2(n) has become same as x1(n), which is an aliasing
error. Equivalently, in this case, the frequency of 50 Hz is an alias of 10 Hz by sampling
 4

rate of 40 Hz. Furthermore, all frequencies (F1 + 40K) are aliases of F1. Hence, do not use
the Nyquist rate blindly.
Example 2: x a (t )= 3 cos 50p t + 10 sin 300p t - cos 100p t
14243 142434 14243
x1 x2 x3

Nyquist rate? f1 = 25, f2 = 150, f3 = 50 Hz

® Fmax = 150 Hz ® Fs = 300 Hz?!
æ n ÷ = 10 sin p n = 0 all the time!
ö

ç
÷
Problem: with Fx = 300 Hz, x2 (n)= xa2 ç F

è sø

If it had a phase shift o <q < n , then it would have been fine, but it is best to choose a
higher sampling rate.

Sampling
x a (t ) P(t)
xa(t) x(n)

Ts

x p (t ) = xa (t )× p(t )
x(n)

+¥

p(t ) = åd (t - nTs )
-¥
+¥

x(n) = x p (t )|t=nTx = å x (nTs )d (t - nTs )

-¥

X p (w )= 1 [X a (w)* P (w )]
2p
2p +¥
P(w) = å d ( w - kw s )
T
s Ks-¥
+¥
1
X p (w) =
T
å X a (w - kws )
-¥
s

Therefore, Xp(w) is a periodic function of shifted the X a(w).

 5

X (w )

Obviously, if ws ³ 2wM there is no

Xp(w) aliasing and the signal can be

reconstructed accurately.
· · ·
wM ws

In practice however, generating very narrow impulse is very difficult. Therefore, the
practical way for sampling is zero-order hold. Such a system samples xa(t) at a given
sampling instant and holds that value until the succeeding sampling instant.
Reconstruction of xr(t) from the output of this system requires a cascade of low-pass
filters or a non-constant gain of LPF.

P(t) H0(t) xo (t )

It is as if xa (t)® x xp(t)
|Hr(w)|
-ws /2 ws /2

<Hr(w) xr(t) hr (t )
p H r (w)
-ws /2 2 w
ws /2