GEORGIA INSTITUTE OF TECHNOLOGY

SCHOOL of ELECTRICAL & COMPUTER ENGINEERING

FINAL EXAM
DATE: 29-APR-15 COURSE: ECE 2026A

NAME: Solutions v1 STUDENT #:

LAST, FIRST

2 points 2 points 2 points

Recitation Section: Circle the date & time when your Recitation Section meets (not Lab):
L00:Tue-9:30am (Zhang) L01:Mon-3:00pm (Casinovi)
L03:Mon-4:30pm (Casinovi) L05:Tue-12:00pm (Zhang)
L06:Thu-12:00pm (Walkenhorst) L07:Tue-1:30pm (Zajic)
L08:Thu-1:30pm (Walkenhorst) L09:Tue-3:00pm (Zajic)
L10:Thu-3:00pm (Fekri) L12:Thu-4:30pm (Fekri)

• Write your name on the front page ONLY. DO NOT unstaple the test.
00
• Closed book, but a calculator is permitted. However, one page (8 12 ⇥ 1100 ) of HAND-WRITTEN
notes permitted. OK to write on both sides.

• Unless stated otherwise, JUSTIFY your reasoning clearly to receive any partial credit.
Showing your work is required to receive any partial credit.

• You must write your answer in the space provided on the exam paper itself.
Only these answers will be graded. Circle your answers, or write them in the boxes provided.
If space is needed for scratch work, use the backs of previous pages.

Problem Value Score

1 14
2 14
3 16
4 12
5 14
6 14
7 14
Rec 2
Total 100
Problem F.2:
DFT Properties

(a) Suppose the DFT, X[k], of a sequence x[n] below, is real. That is, X ⇤ [k] = X[k] for k = 0, . . . , 7.
Can the unknown values of x[n] be determined? If yes, give the missing values. If no, then justify
your answer.
{1, 7, ?, ?, 1, 6, 5, 7}

(b) Given an unknown length-5 real sequence x[n] with a corresponding 5-point DFT coefficient
sequence X[k] = {0, 1, j, j, 1}, determine x[0].

(c) If X8 [k] is the 8-point DFT of a sequence {a, b, c, d, 0, 0, 0, 0}, Express X4 [k], the 4-point DFT,
of the sequence {a, b, c, d} in terms of X8 [k].
Version 1 for Problem 2

(a) Since X*[k]=X[k], i.e. being real for all k, the IDFT of X[k] will also be conjugate symmetric,
i.e., x[N-n]=x*[n], but x[n] is real, so we have x[N-n]=x*[n]=x[n]. This can also be shown:
N −1 N −1
1 j 2π kn / N 1 *
x[ n] = ∑ X [k ]e = ∑X [ k ]e j 2π kn / N ;
N k =0 N k =0
N −1 N −1
1 1
x[ N − n] =
N
∑ X [k ]e j 2π k ( N −n )/ N =
k =0 N
∑ X [k ]e
k =0
− j 2π kn / N
;
N −1
1
so x*[ N − n] = ∑X *
[k ]e j 2π kn / N = x[n] = x[ N − n].
N k =0

Therefor if x[n] = {1; 7; ?; ?; 1; 6; 5; 7}, then the missing x[2]=x[8-2]=5, and x[3]=x[8-3]=6.

(b) The IDFT of an N-point DFT sequence, X[k], can be evaluated as:

N −1
1 j 2π kn / N
x[ n] = ∑ X [k ]e ;
N k =0
N −1
1
so x[0] = ∑ X [k ].
N k =0

Therefor if x[n] = {0; 1; j; -j; 1}, with N=5, then x[0]=2/5.

(c) Recall that zero-padding of a sequence, x[n] = {a; b; c; d}, to y[n] = {a; b; c; d; 0; 0; 0; 0},
produces the same DTFT for both x[n] and y[n]. Therefore, with 8-point DFT that samples the
DTFT of y[n], we end up with twice the number of points than that for a 4-point DFT of x[n].
Since there are four sample falling on top of each other, we can deduct that X[k] = Y[2k], for
k=0, 1, 2, and 3. This can also be proved as follows:

7
Y [k ] = X 8 [k ] = ∑ y[n]e j 2π kn /8 ;
n =0
3 7
and X [k ] = X 4 [k ] = ∑ x[n]e j 2π kn / 4 = ∑ y[n]e j 2π kn / 4 ;
n =0 n =0

therefore X 4 [k ] = X 8 [2k ].
Problem F.3:
Frequency Response

Below are the pole-zero plots of the z-transforms (H(z)) of four discrete-time systems. On the
following pages are plots of magnitude frequency responses ( H(ej ! ˆ ) ) and impulse responses (h[n]).
The numbers on the pole-zero plots represent the multiplicity of the poles and zeros. For each pole-zero
plot, enter the letter of the matching frequency response and impulse response respectively. If it is
helpful, you can tear the next two pages out of the exam to facilitate comparison, but turn them in with
your exam.

ℑ(z)

Frequency Response Impulse Response

ℜ(z)
ANS = A ANS = j

ℑ(z)

Frequency Response Impulse Response

2 3 ℜ(z)
ANS = F ANS = k

ℑ(z)

Frequency Response Impulse Response

ℜ(z)
ANS = B ANS = l

ℑ(z)

Frequency Response Impulse Response

ℜ(z)
ANS = C ANS = i
 (Magnitude) Frequency Responses

-1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1

Normalized Frequency (ω/π) Normalized Frequency (ω/π) Normalized Frequency (ω/π)

A B C

-1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1

Normalized Frequency (ω/π) Normalized Frequency (ω/π) Normalized Frequency (ω/π)

D E F

-1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1

Normalized Frequency (ω/π) Normalized Frequency (ω/π)

G H
 Impulse Responses

2 2 2

1.5 1.5 1.5

1 1 1

0.5 0.5 0.5

0 0 0

-0.5 -0.5 -0.5

-1 -1 -1
-5 0 5 10 15 -5 0 5 10 15 -5 0 5 10 15

i j k

2 2 2
1.5 1.5
1.5
1 1
1
0.5 0.5
0 0.5 0
-0.5 -0.5
0
-1 -1
-0.5
-1.5 -1.5
-1
-5 0 5 10 15 -5 0 5 10 15 -5 0 5 10 15

l m n

2 2

1.5 1.5

1 1

0.5 0.5

0 0

-0.5 -0.5

-1 -1
-5 0 5 10 15 -5 0 5 10 15

o p
Problem F.5:
Frequency Response

y1[n] = x2[n]
x1[n] System 1 System 2 y2[n]

Two systems are connected in cascade, as shown in the figure above. The first system is described by
the following di↵erence equation:

y1 [n] = x1 [n] + b1 x1 [n 1] + x1 [n 2]

The second system is described by the following di↵erence equation:

y2 [n] = a2 y2 [n 1] + x2 [n]

(a) Write an expression for H(z), the system function of the overall cascade system.

(b) When the input to the system is the following:

x1 [n] = 2( 1)n cos(⇡n/5 + ⇡/4)

the corresponding output of the overall system is:

y2 [n] = 4( 1)n

Determine the numerical values of b1 and a2 . Show enough work to make it clear how you arrived
at your final answer. (Hint: Start by computing b1 based on the fact that one of the input
components is missing from the output.)
Problem X

y1[n] = x2[n]
x1[n] System 1 System 2 y2[n]

Two systems are connected in cascade, as shown in the figure above. The first system is described by the
following di↵erence equation:

y1 [n] = x1 [n] + b1 x1 [n 1] + x1 [n 2]

The second system is described by the following di↵erence equation:

y2 [n] = a2 y2 [n 1] + x2 [n]

(a) Write an expression for H(z), the system function of the overall cascade system.
(b) When the input to the system is the following:

x1 [n] = 2( 1)n cos(⇡n/5 + ⇡/4)

the corresponding output of the overall system is:

y2 [n] = 3.5( 1)n

Determine the numerical values of b1 and a2 . Show enough work to make it clear how you arrived at
your final answer. (Hint: Start by computing b1 based on the fact that one of the input components
is missing from the output.)

Solution
(a)
1 + b1 z 1 + z 2
H(z) =
1 a2 z 1
(b)

0 = 1 + b1 e j⇡/5 + e j2⇡/5
1 + e j2⇡/5 ⇣ ⌘
b1 = = ej⇡/5 + e j⇡/5
e j⇡/5
= 2 cos(⇡/5) = 1.618
1 b1 + 1 3.5
H( 1) = = = 1.75
1 + a2 2
1 + a2 = (2 b1 )/1.75 = 2.067
a2 = 1.067