PHN-313: Signals and Systems

Assignment 1 (2023-24)
INSTRUCTED BY: Prof. Peayush Kumar Choubey

Name: Aarav Ratra

Enrolment No: 21122002
Batch: S5
Date of Submission: 2023.08.08
 Note: I am learning LATEX, and hence for practice, I am submitting this assignment in LATEX
format as well as attaching handwritten solutions for some of the questions which require a diagram
for explanation.

1
 Figure 1: Q1(a) : Handwritten Solution

1 Question 1: Periodicity and Fundamental Period

Determine whether or not each of the following signals is periodic and find the fundamental period
if it is.

1.1 x(t) = ei(πt−1)

Answer: The signal is periodic and has a fundamental period = 2.
Explanation:
We know that eit has a fundamental period of 2π (thanks to the Euler Identity eit = cos t+i sin t).
If we have f (t) = eit , we can write x(t) as a transformation by:

1. Right shift by 1 unit: f1 (t) = f (t − 1) = ei(t−1)

2. Scale along the independent axis by a factor of π: x(t) = f1 (πt) = ei(/pit−1)

While shifting the function left or right does not change the periodicity, scaling it along the
independent axis does. The new period is:
T0 2π
T0 = = =2
π π

2
 Figure 2: Q1(b) : Handwritten Solution

P∞ −(2t−k)
1.2 x(t) = k=−∞ e u(2t − k)
Answer: The signal is periodic and has a fundamental period = 0.5.
Explanation:
If we judge the function f (t) = et u(t), we get a function which is 0 for all values from -∞ to 0,
then sharply increasing to 1 at t=0 and following a regular exponential decay curve.

e , if t ≥ 0
 t 
f (t) =
0, if t < 0
We can also define g(t) = f (2t)

e2t , if t ≥ 0
 
g(t) =
0, if t < 0

The curve of g(t) would look similar, except the decay would occur much faster.
We can easily observe that our signal x(t) can be written as:
∞
X
x(t) = g(t − k/2)
k=−∞

The signal is a linear combination of the same signal g(t) superimposed on itself infinite times after
each being shifted by t=1/2 units.
Hence, on sketching the curve, we would get a periodic curve with a period of 1/2.

3
 Figure 3: Q1(c) : Handwritten Solution

1.3 x[n] = cos ( π8 n2 )

Answer: The signal is periodic and has a fundamental period = 8.
Explanation:
If our signal is periodic, let us say that for some N0 ∈ Z :

x[n + N0 ] = x[n]
We can attempt to solve for N0 , but we can only know that the function is periodic if we obtain N0
for which n is unrestricted. Hence, effectively we are supposed to prove:

∀n, ∃N0 : x[n + N0 ] = x[n]

On solving, we get:
π π
cos( (n + N0 )2 ) = cos( n2 )
8 8
π π
=⇒ (n + N0 )2 = 2kπ ± n2 : k ∈ Z
8 8
=⇒ (n + N + 0) = 16k ± n2
2

=⇒ N0 (N0 ± 2n) = 16k

With all the variables being integers, this implies that the number N0 is an even number. Hence, we
substitute N0 = 2m:
4m(m + n) = 16k =⇒ m(m + n) = 4k
Keeping n unrestricted, the only possible answer would be to have m be a multiple of 4. If we look
for the smallest value of m, we have m = 4 =⇒ N0 = 8.
Hence, we can conclude that our function x[n] is periodic with a fundamental period of 8.

Note: This problem is pretty interesting as it shows that a function which is non-periodic in the
case of a continuous-time signal can be periodic in the case of a discrete-time signal. It is obvious
that the function x(t) = cos( π8 t2 ) is not periodic.

4
 Figure 4: Q1(d) : Handwritten Solution

P∞
1.4 x[n] = k=−∞ (δ[n − 4k] − δ[n − 1 − 4k])
Answer: The signal is periodic and has a fundamental period = 4.
Explanation:
First, let us assess the following function:

g[n] = δ[n] − δ[n − 1]

It is easy to observe that the function can be written as:

;n=0 
 
 1,
g[n] = −1, ;n=1
0, otherwise
 

And hence the signal waveform can be understood as two peaks at n = 0 and n = 1, reaching values
+1 and -1, respectively.
It is easy to judge that the signal x[n] is a superposition of infinitely many g[n] signals, each
shifted by 4 units from the other. On sketching the graph and observing, we can redefine x[n] as

; n = 4k 
 
X∞  1,
x[n] = g[n − 4k] = −1, ; n = 4k + 1
0, otherwise
 
k=−∞

From the definition and the sketch of the waveform, since k ∈ Z, it is easy to conclude that the
signal x[n] is periodic and has a fundamental period of 4.

5
 Figure 5: Plots of signals x(t) (a) and y(t) (b).

Figure 6: Q2: Handwritten Solution

2 Question 2: Signal Manipulation by Shifting and Scaling

Consider signals x(t) and y(t) as shown in the accompanying figure. Express y(t) in terms of x(t).
Answer:
It is easy to observe that y(t) is the same shape as x(t) except it is flipped twice, scaled and
shifted. We can use the following procedure to obtain y:

1. Scale x(t) about the independent axis such that it compresses by a factor of 2 ⇒ x(t) → x(2t)

2. Scale x(t) about the dependent axis such that it expands by a factor of 2 ⇒ x(2t) → 2x(2t)

3. Flip x(t) about both the independent and dependent axes as it would effectively cause a 180º
rotation about (0,0) ⇒ 2x(2t) → −2x(−2t)

4. Shift x(t) 1 unit upwards ⇒ −2x(−2t) → 1 − 2x(−2t)

5. Shift x(t) 1 unit rightwards ⇒ 1 − 2x(−2t) → 1 − 2x(−2(t − 1)) = 1 − 2x(2 − 2t)

With this series of manipulations, we have finally obtained y(t) = 1 − 2x(2 − 2t)

6
 Figure 7: Q3: Handwritten Solution

3 Question 3: Unit Step Function

Given x(t) = 5u(t + 2)u(t) + 3u(t2)7u(t4), find and plot x(2t4).
Answer: A good first step would be to first plot x(t). We can analyse and arrive to a piece-wise
definition for x(t):  

 0; t < −2 

 5; −2 ≤ t < 0 

 

x(t) = 4; 0 ≤ t < +2
 7; +2 ≤ t < +4 

 

 
0; t > +4
 

And the curve of x(t) can be drawn very easily. We can use this piece-wise definition as well as
the graphical method to find x(2t − 4). I shall use the graphical method. We can obtain x(2t − 4)
from x(t) by first scaling along the t-axis by a factor of 2 (compressing) followed by a right shift of
2 points. The analysis would arrive to the following definition for x(2t − 4):
 

 0; t < 1 

5; 1 ≤ t < 2

 

 
x(2t − 4) = 4; 2 ≤ t < 3 = 5u(t − 1) − u(t − 2) + 3u(t − 3)
7; 3 ≤ t < 4

 


 

0; t > 4
 

7
 Figure 8: Q4(a): Handwritten Solution

4 Question 4: Even and Odd Signals

4.1 Continuous Signal Average
The average value x of a signal x(t) is given by
Z T
1
lim x(t)dt
n→∞ 2T −T

Let xe (t) and xo (t) be the even and the odd part of x(t), respectively. Show that xo = 0 and xe = x.

Answer:
While it is easy to intuitively understand from the graphical nature of the even and odd parts
that the mean of the odd part of any function would be 0, leaving only the even part to contribute
to the mean, we shall do it formally.
It is easy to establish that the function x(−t) also has an average value x. We can say so due to
the symmetry of the expression for the average.
Hence, on calculating xo and xe :
Z T
1 T (x(t) + x(−t))
Z
1 (x(t) + x(−t)) x+x
xe = lim dt = lim dt = =x
n→∞ 2T −T 2 n→∞ 2 −T 2T 2
Z T
(x(t) − x(−t)) 1 T (x(t) − x(−t)) x−x
Z
1
x0 = lim dt = lim dt = =0
n→∞ 2T −T 2 n→∞ 2 −T 2T 2
Hence, Proven.

8
 Figure 9: Q4(b): Handwritten Solution

4.2 Discrete Signals

Let xe [n] and xo [n] be the even and the odd part of x[n], respectively. Show that
∞
X ∞
X ∞
X
x2 [n] = x2e [n] + x2o [n]
n=−∞ n=−∞ n=−∞

Answer:
We know that

xe [n] = 0.5(x[n] + x[−n]); xo [n] = 0.5(x[n] − x[−n])

Hence, the quantity x2e [n] + x2o [n] can be expanded as:
1 1
x2e [n] + x2o [n] = (x2 [n] + x2 [−n] + 2x[n]x[−n]) + (x2 [n] + x2 [−n] − 2x[n]x[−n])
4 4
1
= (x2 [n] + x2 [−n])
2
Hence, the quantity on the RHS can be written as:
∞
X ∞
X ∞
X
x2e [n] + x2o [n] = x2e [n] + x2o [n]
n=−∞ n=−∞ n=−∞
∞ ∞∞
X 1 2 2 1 2
X X
= (x [n] + x [−n]) = ( x [n] + x2 [−n])
n=−∞
2 2 n=−∞ n=−∞

In the second summation, we can substite n → −n and obtain the summation as

∞ ∞ ∞
1 X 2 X
2
X
( x [n] + x [n]) = x2 [n]
2 n=−∞ n=−∞ n=−∞

which is equal to the LHS.

Hence Proven.

9
 Figure 10: Q5: Handwritten Solution

5 Question 5: Sum of Periodic Signals

Consider two periodic signals x1 (t) and x2 (t) with periods T1 and T2, respectively.

1. Show that the sum x1 (t) + x2 (t) is periodic only if the ratio T1 /T2 is a rational number.

2. Find the period T0 of x(t) for T1 /T2 = k2 /k1 , where k1 and k2 are positive integers.

Answer:
For the sum of two signals to be periodic, they need to have a least common multiple, which will
be equal to the signal. We can understand it via the following logic:

x1 (t) + x2 (t) = x1 (t + k1 T1 ) + x2 (t + k2 T2 ); (k1 , k2 ) ∈ Z 2

If the function is to be periodic, it is necessary that k1 T1 = k2 T2 .
Thus, the ratio T1 /T2 = k2 /k1 , indicates that the ratio is a rational number. The period, hence,
would be equal to the least common integral multiple of the individual periods.

T0 = k1 T1 = k2 T2

It is important to note that the period obtained from the LCM need not always be the fundamental
period. It is essential to do a sketch of the curve and check for the nature of the periodic signals
being added.

10
6 Q6: The Impulse Function
Prove the following relations.

6.1 δ(t) = δ(t)

Answer:
We can come to the following statements, which come from the definition of the impulse function,
which is described by its action rather than by its mathematical composition.

δ(t) = 0∀t ̸= 0
Z ∞
δ(t) = 1
−∞

Now, if we try to define δ(−t), we get:

δ(−t) = 0∀(−t) ̸= 0
Z ∞ Z −∞
t→(−t)
δ(−t)dt −−−−→ δ(t)d(−t)
−∞ ∞
Z ∞
= δ(t) = 1
−∞

Clearly, the definition of δ(t) and δ(−t) match with each other. Hence, we can say that they are
equivalent.

δ(t)
6.2 δ(at) = |a|
Answer: R∞
We can use the integral −∞ δ(at)dt to obtain it in terms of δ(t)
On substituting v = at, our integral transforms to:
Z ∞ Z ∞ Z ∞
dv 1
δ(at)dt = sgn(a) δ(v) = δ(v)dv
−∞ −∞ a |a| −∞
Note that the factor sgn(a) comes as a result of flipping of the limits of the integral in the case where
a < 0. Hence, this causes the term |a| to appear.
We can further conclude:
Z ∞ Z ∞
1 δ(t)
δ(at)dt = δ(t)dt =⇒ δ(at) =
−∞ |a| −∞ |a|
Hence proving the statement.

11
 1
6.3 δ(t2 − t20 ) = 2|t0 | (δ(t + t0 ) + δ(t − t0 ))
Answer:
R∞
We shall be using the sifting property of the impulse function. , −∞ δ(t2 t20 )dt. The substitution
√
we shall be using
√ is t = v =⇒ t = ± v.
2
√
For t = + v , i.e. t > 0 , we get dt = dv/2 v
Z ∞ Z ∞ Z ∞ Z ∞
dv 1 1 1
δ(t2 − t20 )dt = δ(vt20 ) √ = √ δ(vt20 )dv = p = δ(tt0 )dt
0 0 2 v 0 2 v 2 t20 2|t0 | 0
√ √
Similarly, for t = − v , i.e. t < 0 , we get dt = −dv/2 v
Z 0 Z 0 Z 0 Z 0
2 2 2 −dv −1 2 1 1
δ(t − t0 )dt = δ(vt0 ) √ = √ δ(vt0 )dv = p = δ(t + t0 )dt
−∞ −∞ 2 v −∞ 2 v 2 t20 2|t0 | −∞

Hence, on addition, we get:

Z ∞ Z ∞
2 1
δ(t − t20 )dt = δ(t − t0 ) + δ(t + t0 )dt
−∞ 2|t0 | −∞

Hence, concluding:
1
δ(t2 ˘t20 ) = (δ(t + t0 ) + δ(t − t0 ))
2|t0 |
Hence proven.

du(t)
6.4 δ(t) = dt
Answer:

We can express the integral of δ(t) from −∞ to arbitrary time t as:

0, ; t < 0
Z t  
δ(t)dt = = u(t)
−∞ 1, ; t ≥ 0
Hence, differentiating both sides gives us the following : https://www.overleaf.com/project/64cbf260ab862739a7e1ba

d t
Z
du d
= δ(t)dt = δ(t) (t) − 0 = δ(t)
dt dt −∞ dt

Hence, proven.

12
7 Question 7: Classification of Systems
A system may or may not be (i) memoryless, (ii) time-invariant, (iii) linear, (iv) causal, and (v)
stable. Determine which of these properties hold or do not hold for the systems described by following
input-output relations.
R 2t
7.1 y(t) = −∞ x(τ )dτ
1. Memory: This system is an integral, which by definition means the area under the curve from
−∞ to time 2t. This clearly depends on the value of the function for values of x(τ ) for τ < 2t
. Hence, the system is not memory-less.

2. Time Variance: To check for time variance, we check if shifting the signal x(t) by time t0
also does the same to y(t).

Z 2t
x(τ ) → x(τ − t0 ) ⇒ y1 (t) = x(τ − t0 )dτ
−∞

Substituting τ ′ = τ − t0 , we get:
Z 2(t−t0 )
y1 (t) = x(τ ′ )dτ ′ = y(t − t0 )
−∞

Hence, the given system is time invariant.

3. Linearity: To check for linearity, we check that for the input x1 (t) + x2 (t), the output is equal
to y1 (t) + y2 (t), where y1 (t) and y2 (t) correspond to the outputs corresponding to x1 (t) and
x2 (t). R 2t R 2t
For y1 (t) = −∞ x1 (τ )dτ and y2 (t) = −∞ x2 (τ )dτ :
Z 2t Z 2t Z 2t
x1 (τ ) + x2 (τ )dτ = x1 (τ )dτ + x2 (τ )dτ = y1 (t) + y2 (t)
−∞ −∞ −∞

Hence, the given system is linear.

4. Causality: From the expression, it is easy to say that y(t0 ) is equal to the area under the
curve of x(t0 ) from time −∞ to 2t0 . This clearly shows that y(t0 ) is dependant on values of
x(t) for values t > t0 . This makes the system non-causal. For a system to be causal, the
output y(t0 ) should only be dependant on values of x(t) for t ≤ t0 .

5. Stability: For a system should be stable, the output signal needs to be bounded when the
input is also a bounded signal.
Before trying to formally prove for stability, we can check a test case where the input is x(t) = 1
i.e. constant +ve input. The signal is obviously bounded.
Z 2t Z 2t
y(t) = x(τ )dτ = 1dτ = [2t − (−∞)] = −∞
−∞ −∞

This is clearly an unbounded solution. Hence, the system is unstable.

13
  
0, if t < 0
7.2 y(t) =
x(t) + x(t − 2), if t ≥ 0
1. Memory: From the definition of this signal, we can determine that output y(t) is dependant
on x(t) as well as x(t − 2). The dependence on x(t − 2) makes this system non memory-less.

2. Time Variance: To check for time variance, we check if shifting the input signal x(t) by time
t0 also does the same to y(t).

if t < 0
 
0,
y1 (t) =
x(t − t0 ) + x(t − t0 − 2), if t ≥ 0

However,

if t − t0 < 0
 
0,
y(t − t0 ) =
x(t − t0 ) + x(t − t0 − 2), if t − t0 ≥ 0
This shows that the signals y1 (t) and y(t − t0 ) are not equal (they show inequality in the region
(0, t0 ) for t0 > 0 and in region (t0 , 0) for t0 < 0). This makes the system time-variant.

3. Linearity: We define y1 (t) and y2 (t) as output signals corresponding to input signals x1 (t)
and x2 (t). We compute the signal output y1+2 (t) corresponding to x1 (t) + x2 (t)

if t < 0
 
0,
y1+2 (t) =
x1 (t) + x2 (t) + x1 (t − 2) + x2 (t − 2), if t ≥ 0

if t < 0
 
0,
= = y1 (t) + y2 (t)
(x1 (t) + x1 (t − 2)) + (x2 (t) + x2 (t − 2)), if t ≥ 0

Hence, the given system is linear.

4. Causality: From the expression, it is easy to say that for a time t0 , the output y(t0 ) is
dependant on x(t0 ) and x(t0 − 2). There is no dependence of y(t0 ) on x(t); t > t0 . This is
enough to say that the system is causal.

5. Stability: Let us say that the input signal x(t) is bounded. This can be mathematically
expressed as

|x(t)| ≤ Mx < ∞
For t < 0, our output signal y(t) is obviously bounded at a single value y = 0 . Hence, we only
have to analyse for t ≥ 0:

y(t) = x(t) + x(t − 2)

=⇒ |y(t)| = |x(t) + x(t − 2)| ≤ |x(t)| + |x(t − 2)| ≤ Mx + Mx = 2Mx
This is enough to say that y(t) is bounded. This proves that our system is stable.

14
7.3 y[n] = nx[n]
1. Memory: The signal y[n = n0 ] is solely dependant on n0 and x[n = n0 ]. There is no
dependence of the system on x[n < n0 ]. Hence, the system is memory-less.

2. Time Variance: To check for time variance, we check if shifting signal x[n] by n0 samples
does the same for output y[n].

y1 [n] = nx[n − n0 ], but y[n − n0 ] = (n − n0 )x[n − n0 ] ̸= y1 [n]

This clearly proves that the signal is time variant.

3. Linearity: We define y1 [n] and y2 [n] as output signals corresponding to inputs x1 [n] and x2 [n]
respectively. We compute signal output y1+2 [n] for input x1 [n] + x2 [n].

y1+2 [n] = n(x1 [n] + x2 [n]) = nx1 [n] + nx2 [n] = y1 [n] + y2 [n]
Hence, proving that our system is linear.

4. Causality: Like observed before, the system output y[n = n0 ] is solely dependent on n and
x[n0 ]. It is not dependent on the inputs for x[n > n0 ], hence the system is causal.

5. Stability: Let our input signal x[n] be bounded such that |x[n]| ≤ Mx < ∞. Then, we can
say the following about our output signal:

|y[n]| = |n(x[n])| = |n||x[n]| ≤ |n|Mx

This clearly shows that y[n] is not bounded, proving our system to be unstable. We could have
also proven the same using a counter-example (like substituting x[n] = 1 would give y[n] = n
which is obviously not bounded).
 
 x[n], if n ≥ 1 
7.4 y[n] = 0, if n = 0
x[n + 1], if n ≤ −1
 

1. Memory: From the definition of the function, we can see that output y[n0 ] is dependant on
input x[n0 ] for n > 0 and on x[n0 + 1] for n < 0. Since there is no dependence of y[n0 ] for any
x[n > n0 ], we can conclude that the signal is memory-less.

2. Time-Variance: To check for time variance, we check if shifting signal x[n] by n0 samples
does the same for output y[n].

if n > 0 
 
 x[n − n0 ],
y1 [n] = 0, if n = 0
x[n − n0 + 1], if n < 0
 

However,

if n > n0 
 
 x[n − n0 ],
y[n − n0 ] = 0, if n = n0
x[n − n0 + 1], if n < n0
 

This clearly shows that the definitions here are very different. Hence, the system is time-
variant.

15
3. Linearity: We define y1 [n] and y2 [n] as output signals corresponding to inputs x1 [n] and x2 [n]
respectively. We compute signal output y1+2 [n] for input x1 [n] + x2 [n]

if n > 0 
 
 x1 [n] + x2 [n],
y1+2 [n] = 0, if n = 0
x1 [n + 1] + x2 [n + 1], if n < 0
 

if n > 0   x2 [n], if n > 0 

   
 x1 [n],
= 0, if n = 0 + 0, if n = 0 = y1 [n] + y2 [n]
x1 [n + 1], if n < 0 x2 [n + 1], if n < 0
   

Hence, our given system is linear.

4. Causality: From the definition of the function, it is clear that for n0 < 0, y[n0 ] depends on
x[n0 +1]. This dependence on the future value of the signal proves that the signal is non-causal
in nature.

5. Stability: Let our input signal x[n] be bounded such that |x[n]| ≤ Mx < ∞∀x . Then, we can
say the following about our output signal:

if n ≥ 1   Mx , if n ≥ 1 
   
 |x[n]|,
|y[n]| = 0, if n = 0 ≤ 0, if n = 0 =⇒ |y[n]| ≤ Mx
|x[n + 1]|, if n ≤ −1 Mx , if n ≤ −1
   

This proves that our output signal y[n] is bounded for a bounded input signal. Hence, given
system is stable.

16
 Figure 11: Q8: Given Figure

8 Question 8: Feedback System

Consider a feedback system shown in the figure. Here, S 1 represents a unit-delay operator.

8.1 Find the input-output relation for the system.

Answer:
Let us say that the input at time n = 0 is x[0] and let y[0] = 0. The result of the Σ operation would
give v[0] = x[0] + 0
We can say that at time n = 1, the value of the signal becomes x[1] and thanks to the delay operator,
we get y[1] = v[0] = x[0], and hence v[1] = x[1] + y[1] = x[1] + x[0].
Similarly, for n = 2, we get y[2] = v[1] = x[1] + x[0] and v[2] = x[2] + y[1] = x[2] + x[1] + x[0].
On repeating this till arbitrary n, we can make the following conclusions:
n
X
v[n] = x[n − k]
k=0
n−1
X
y[n] = v[n − 1] = x[n − k − 1] = x[n − 1] + x[n − 2] + ... + x[0]
k=0

8.2 Is this system causal, linear and time-invariant?

Answer:
The system is causal since it does not have a dependence on future values of x[n]. It is however not
memory-less since it is a summation of the same signal delayed again and again.
The system is linear since for input x1 [n] + x2 [n] the output comes out to be:

n−1
X n−1
X n−1
X
y1+2 [n] = x1 [n − k − 1] + x2 [n − k − 1] = x1 [n − k − 1] + x2 [n − k − 1] = y1 [x] + y2 [x]
k=0 k=0 k=0

where y1 [n] and y2 [n] are the outputs corresponding to inputs x1 [n] and x2 [n].

8.3 Assuming y[n] = 0 for n < 0, sketch the output of the system for
x[n] = δ[n].
Answer:
We can describe the output y[n] for input x[n] = δ[n] as:
n−1
X
y[n] = δ[n − k − 1] = δ[n − 1] + δ[n − 2] + ... = u[n − 1]
k=0

The curve matches that of the unit-step function, hence we can say so. The sketch of the curve is
attached.

17
 Figure 12: Q8(c): Handwritten Figure

It is important to note that this system functions like a delayed integrator as it calculates the
sum of all the values of the signal till x[n − 1].

THE END

18