Middle East Technical University

Department of Electrical and Electronics Engineering

EE301: Signals and Systems I

Solutions for Homework 1

October 20, 2019
If you face any problem or mistake please contact Ömer Çayır, ocayir@metu.edu.tr, DZ-10.

1. i. x(t)

−1
−6 −5 −4 −3 −2 1 2 3 4 5 6 t

−1

x(t + 1)

−2
−6 −5 −4 −3 −1 1 2 3 4 5 6 t

−1

ii. x(2t)

−0.5
−6 −5 −4 −3 −2 −1 0.5 1 2 3 4 5 6 t

−1

 
iii. t−2
x
3

−3 −2 −1 1 2 3 4 5 6 7 8 9 t

−1
 iv. 1
xe (t) = Eν {x(t)} = (x(t) + x(−t))
2
xe (t)

0.5

−6 −5 −4 −3 −2 −1 1 2 3 4 5 6 t

1
xo (t) = Od {x(t)} = (x(t) − x(−t))
2
xo (t)

1
0.5
−2 −1
−6 −5 −4 −3 1 2 3 4 5 6 t
−0.5
−1

2. • If a CT signal x(t) is periodic, there exists a positive number T such that x(t + T ) = x(t)
and its fundamental period T0 is the smallest nonzero period T satisfying this relation.

• If a DT signal x[n] is periodic, there exists a positive integer N such that x[n + N ] = x[n]
and its fundamental period N0 is the smallest nonzero period N satisfying this relation.
 π
i. x(t) = 3 cos 4t +
3
2π π
is periodic with the fundamental period T0 = = .
4 2
h  i   
ii. π 2 1 2π
x(t) = cos 2t − = 1 + cos 4t −
3 2 3
2π π
is periodic and its fundamental period is T0 = = .
4 2
iii. x(t) = 2 cos (10t + 1) − sin (4t − 1)
2π π 2π π
cos (10t + 1) is periodic with period = and sin (4t − 1) is periodic with period = .
10 5 π π  4 2
Hence, x(t) is periodic and its fundamental period is T0 = LCM , = π.
5 2
iv. 1 1
x(t) = Eν {cos (4π t) u(t)} = [cos (4π t) u(t) + cos (−4π t) u(−t)] = cos (4π t)
2 2
(Remember that cos (−θ) = cos (θ).)
2π 1
The signal x(t) is periodic and its fundamental period is T0 = = .
4π 2
  π 
v. x[n] = j n = exp j n
2
 π 
x[n + N ] = x[n] =⇒ exp j N = 1 = exp (j2πk) for k ∈ Z
2
Then, we obtain N = 4k. Hence, x[n] is periodic and its fundamental period is N0 = 4 (for
k = 1).
√  π n √ n  π 
vi. x[n] = (1 + j)n = 2 exp j = 2 exp j n
4 4
√
n
Owing to 2 , we cannot find a nonzero N satisfying x[n + N ] = x[n], and the signal x[n]
is not periodic.
 
vii. 25
x[n] = exp j πn
4
 
25
x[n + N ] = x[n] =⇒ exp j πN = 1 = exp (j2πk) for k ∈ Z
4
8
Then, we obtain N = k. Hence, x[n] is periodic and its fundamental period is N0 = 8
25
(for k = 25).

3. (1) A system is memoryless (instantaneous) if the output at any time instant depends only on
the value of the input at that particular instant.

(2) A system is causal if its output at any time depends only on the values of the input at present
time and in the past.

(3) A system is stable if bounded inputs lead to bounded outputs (BIBO stable).

(4) A system is time-invariant if a time shift in the input signal causes same amount of time shift
in the output signal.

(5) A system is linear if it possesses the superposition property:

CT: a x1 (t) + b x2 (t) → a y1 (t) + b y2 (t) ⇐= x1 (t) → y1 (t) and x2 (t) → y2 (t)
DT: a x1 [n] + b x2 [n] → a y1 [n] + b y2 [n] ⇐= x1 [n] → y1 [n] and x2 [n] → y2 [n]

System (1) Memoryless (2) Causal (3) Stable (4) Time-Invariant (5) Linear

i. y[n] = x[n − 2]+x[2 − n] 7 7 3 7 3

ii. y(t) = cos (3t) x(t) 3 3 3 7 3
2n
X
iii. y[n] = x[k] 7 7 7 7 3
k=−∞
dx(t)
iv. y(t) = 7 3 7 3 3
dt

i. y[n] = x[n − 2] + x[2 − n]

The system is not causal. For instance, when n = 0, we see that y[0] = x[−2] + x[2], and
the current value of the output depends on the future value of the input.
 The system is not memoryless since it is not causal.
The system is not time-invariant owing to x[2−n]. Let x2 [n] = x1 [n−n0 ] for x1 [n] → y1 [n]
and x2 [n] → y2 [n], then check whether y2 [n] = y1 [n − n0 ] or not.
)
y2 [n] = x2 [n − 2] + x2 [2 − n] = x1 [n − 2 − n0 ] + x1 [2 − n − n0 ]
=⇒ y2 [n] 6= y1 [n − n0 ]
y1 [n − n0 ] = x1 [n − n0 − 2] + x2 [2 − n + n0 ]

ii. y(t) = cos (3t) x(t)

The system is not time-invariant owing to cos (3t). Let x2 (t) = x1 (t − t0 ) for x1 (t) → y1 (t)
and x2 (t) → y2 (t), then check whether y2 (t) = y1 (t − t0 ) or not.
)
y2 (t) = cos (3t) x2 (t) = cos (3t) x1 (t − t0 )
=⇒ y2 (t) 6= y1 (t − t0 )
y1 (t − t0 ) = cos (3t − 3t0 ) x1 (t − t0 )

iii. 2n
X
y[n] = x[k]
k=−∞

The system is not causal. For instance, when n = 1, we see that y[1] = x[2]+x[1]+x[0]+. . .,
and the current value of the output depends on the future value of the input.
The system is not memoryless since it is not causal.
The system is not time-invariant owing to the upper limit of the summation 2n. Assuming
x2 [n] = x1 [n − n0 ] for x1 [n] → y1 [n] and x2 [n] → y2 [n], we obtain y2 [n] 6= y1 [n − n0 ].
2n 2n 2n−n

X X `,k−n0
X0 

y2 [n] = x2 [k] = x1 [k − n0 ] = x2 [`]



k=−∞ k=−∞ `=−∞
=⇒ y2 [n] 6= y1 [n − n0 ]
2n−2n
X0 


y1 [n − n0 ] = x1 [k]


k=−∞

The system is not stable. Let x[n] = u[n]. Then, we have

2n
X 2n
X
y[n] = u[k] = u[k] = (2n + 1) u[n],
k=−∞ k=0

and y[n] → ∞ as n → ∞.
iv. dx(t)
y(t) =
dt
The derivative of x(t) can be expressed as

dx(t) x(t) − x(t − ∆)

= lim ,
dt ∆→0 ∆
see Section 2.5.3 from Oppenheim. Thus, the system is causal, but it is not memoryless
since the current value of the output depends on the past value of the input.
√
The system is not stable. Let x(t) = 4 − t2 if |t| < 2, and x(t) = 0 otherwise. Then, we
have
dx(t) t
y(t) = =√
dt 4 − t2
if |t| < 2, and it is zero otherwise. As t → 2, y(t) → ∞, while x(t) 6 2 for all t.
 Z ∞ Z ∞
4. i. yi (t) = x1 (t) ∗ x2 (t) = x1 (t − τ ) x2 (τ ) dτ = [u(t − τ − 3) − u(t − τ − 5)] e−3τ dτ
−∞ 0 | {z }
nonzero only for t−5<τ <t−3

For t 6 3, we have yi (t) = 0.

For 3 < t 6 5, we have
Z   t−3
t−3
−1 −3τ  1 − e−3(t−3)
yi (t) = e −3τ
dτ = e  = .
3  3
0 τ =0

For t > 5, we have

Z   t−3

t−3
−3τ −1 −3τ 
 e−3(t−5) − e−3(t−3) 1 − e−6 e−3(t−5)
yi (t) = e dτ = e  = = .
t−5 3 τ =t−5 3 3


0, t63



 1 − e−3(t−3)
, 3<t65
yi (t) = 3




 −6 e−3(t−5)
 1−e

, t>5
3
ii. dx1 (t)
yii (t) =
dt
∗ x2 (t) = [δ(t − 3) − δ(t − 5)] ∗ x2 (t) = x2 (t − 3) − x2 (t − 5)
yii (t) = e−3(t−3) u(t − 3) − e−3(t−5) u(t − 5)

iii. 
0,
 t63
dyi (t) e−3(t−3) , 3<t65
yii (t) = =
dt 

 e−6 − 1
e−3(t−5) , t > 5


5. We find the output signals for the given input signals and h[n] = αn u[n], where α ∈ C, as follows:
∞
X ∞
X ∞
X 1
y1 [n] = h[n] ∗ x1 [n] = x1 [n − k] h[k] = αk u[k] = αk = , if |α| < 1
1−α
k=−∞ k=−∞ k=0

∞ ∞ n 
(n + 1) u[n], α=1
X X X
k k
y2 [n] = x2 [n − k] h[k] = u[n − k] α u[k] = u[n] α = n+1
k=−∞ k=−∞ k=0 1 − α

u[n], α 6= 1
1−α
∞
X ∞
X −1
X
n−k
y3 [n] = x3 [k] h[n − k] = u[−k − 1] α u[n − k] = αn−k u[n − k]
k=−∞ k=−∞ k=−∞

This summation is bounded only for |α| < 1, and it can be computed separately for n < −1 and
n > −1, since u[n − k] is nonzero only when k 6 n.
 n ∞
 X X 1

 α n−k `,n−k
= α `
= , n < −1 and |α| < 1
−1 
 1−α
X k=−∞ `=0
y3 [n] = αn−k u[n − k] = −1 ∞

 X `,−k−1
X αn+1
k=−∞ 
 α n−k
= α n+`+1
= , n > −1 and |α| < 1
 1−α
k=−∞ `=0
 As x3 [n] = x1 [n] − x2 [n], we have y3 [n] = y1 [n] − y2 [n] owing to the distributive property of LTI
systems shown below.

y3 [n] = h[n] ∗ x3 [n] = h[n] ∗ (x1 [n] − x2 [n]) = h[n] ∗ x1 [n] − h[n] ∗ x2 [n] = y1 [n] − y2 [n]


 1
 , n < 0 and |α| < 1
1−α
y3 [n] = y1 [n] − y2 [n] =

 αn+1
 , n > 0 and |α| < 1
1−α

 
 1
n+1   , n < −1 and |α| < 1
α  1 1−α
 = =⇒ y3 [n] = y1 [n] − y2 [n] =
1 − α 1−α 
 αn+1
n=−1  , n > −1 and |α| < 1
1−α

(
6. i. 1, 1 6 n 6 5
x1 [n] = x2 [n] =
0, otherwise

∞
X 5
X 
n − 1, 1<n66
y[n] = x1 [n] ∗ x2 [n] = x1 [k] x2 [n − k] = x2 [n − k] = 11 − n, 6 < n 6 10


k=−∞ k=1 0, otherwise
1

0
0 2 4 6 8 10 12

0
0 2 4 6 8 10 12

0
0 2 4 6 8 10 12

To verify our result x1 [n] ∗ x2 [n] in MATLAB, the sample code is given below.
x = @ ( n ) (n >=1 & n <=5)*1; % x [ n ]=1 if 1 <=n <=5 , x [ n ]=0 otherwise
nx = 0:6; % sample index for x1 and x2
ny = 2* min ( nx ):2* max ( nx ); % sample index for the convolution

% obtain x1 and x2
xOne = x ( nx );
xTwo = x ( nx );
 % compute the convolution of x1 and x2 by using ’ conv ’ function
y = conv ( xOne , xTwo );

% visualization of the DT signals

figure (1); subplot (211); % x1
stem ( nx , xOne , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ min ( ny ) max ( ny )] , ’ YTick ’ ,[0 1])
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$x_1 [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)

subplot (212); % x2
stem ( nx , xTwo , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ min ( ny ) max ( ny )] , ’ YTick ’ ,[0 1])
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$x_2 [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)

figure (2); % y
stem ( ny ,y , ’ filled ’ , ’ LineWidth ’ ,2);
title ( ’ The convolution of $$x_1 [ n ] $$ and $$x_2 [ n ] $$ ’ , ...
’ Interpreter ’ , ’ latex ’ );
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$y [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)
ii. Assuming x1 [n] = 0 for n 6 0 and n > N1 , and x2 [n] = 0 for n 6 0 and n > N2 , we can use
the MATLAB code given below to compute the convolution of x1 [n] and x2 [n].
function [y , ny ] = ee301hw1q6partii ( xOne , xTwo )
% get the length of each input signal
NOne = length ( xOne );
NTwo = length ( xTwo );

% memory allocation for the convolution of the input signals

y = zeros (1 , NOne + NTwo -1);

% select the loop with less repetitions to run faster

if NOne < NTwo
% use x1 [1]* x2 [n -1]+ x1 [2]* x2 [n -2]+...+ x1 [ N1 ]* x2 [n - N1 ]
for k =1: NOne
ind = k +(0: NTwo -1);
y ( ind ) = y ( ind )+ xOne ( k )* xTwo ;
end
else
% use x2 [1]* x1 [n -1]+ x2 [2]* x1 [n -2]+...+ x2 [ N2 ]* x1 [n - N2 ]
for k =1: NTwo
ind = k +(0: NOne -1);
y ( ind ) = y ( ind )+ xTwo ( k )* xOne ;
end
end

% get the sample index for the convolution of the input signals
ny = 2:( NOne + NTwo );
 ( (
iii. n, 1 6 n 6 20 11 − n, 1 6 n 6 10
x1 [n] = x2 [n] =
0, otherwise 0, otherwise
20

10

0
5 10 15 20 25 30

10

0
5 10 15 20 25 30

1000

500

0
5 10 15 20 25 30

To compute x1 [n] ∗ x2 [n] in MATLAB, the sample code is given below.

NOne = 20;
nxOne = 1: NOne ; % sample index for x1
NTwo = 10;
nxTwo = 1: NTwo ; % sample index for x2

% obtain x1 and x2
% x1 [ n ]= n if 1 <=n <=20 , x1 [ n ]=0 otherwise
xOne = nxOne ;
% x2 [ n ]=11 - n if 1 <=n <=10 , x2 [ n ]=0 otherwise
xTwo = 11 - nxTwo ;

% compute the convolution of x1 and x2 by using ’ conv ’ function

[y , ny ] = ee301hw1q6partii ( xOne , xTwo );

% get lower and upper limits of time axis for alignment

nMin = min (1 , min ( ny ));
nMax = max ( max ( NOne , NTwo ) , max ( ny ));

% visualization of the DT signals

figure (1); subplot (211); % x1
stem ( nxOne , xOne , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ nMin nMax ])
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$x_1 [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)

subplot (212); % x2
 stem ( nxTwo , xTwo , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ nMin nMax ])
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$x_2 [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)

figure (2); % y
stem ( ny ,y , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ nMin nMax ])
title ( ’ The convolution of $$x_1 [ n ] $$ and $$x_2 [ n ] $$ ’ , ...
’ Interpreter ’ , ’ latex ’ );
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$y [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)
iv. x1 [n − 4] ∗ x2 [n + 5] = (x1 [n] ∗ δ[n − 4]) ∗ (x2 [n] ∗ δ[n + 5])
= (x1 [n] ∗ x2 [n]) ∗ (δ[n − 4] ∗ δ[n + 5])
| {z } | {z }
y[n] from part (iii) δ[n+1]

= y[n + 1]
Using commutative After running the code given in part (iii), we use the following lines in
MATLAB to obtain the desired result.
figure (2); % y
stem ( ny -1 ,y , ’ filled ’ , ’ LineWidth ’ ,2);
set ( gca , ’ XLim ’ ,[ nMin nMax ])
title ( ’ The convolution of $$x_1 [n -4] $$ and $$x_2 [ n +5] $$ ’ , ...
’ Interpreter ’ , ’ latex ’ );
xlabel ( ’ $$n$$ ’ , ’ Interpreter ’ , ’ latex ’ );
ylabel ( ’ $$y [ n ] $$ ’ , ’ Interpreter ’ , ’ latex ’)

1000

500

0
5 10 15 20 25 30