"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1

Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 1

Q.1.1 DMS with source probabilities : {0.30 0.25 0.20 0.15 0.10}
1
Entropy H(X) = ∑ pi log
i pi
= 0.30 log 1/0.30 + 0.25 log 1/0.25 + ……………
= 2.228 bits

qi
Q.1.2 Define D(p ⎜⎜q) = ∑p i
i log
pi
(1)

pi, qi – probability distributions of discrete source X.

qi ⎛ qi ⎞
D(p ⎜⎜q) = ∑p i log ≤ ∑P⎜⎜ − 1⎟⎟ [using identity ln x ≤ x – 1]
i
i pi i ⎝ pi ⎠
= ∑ (qi − pi ) = 0
i

∴ D(p ⎜⎜q) ≥ 0

Put qi = 1/n in (1) where n = cardinality of the distance source.

D(p ⎜⎜q) = ∑ p log p

i
i i + ∑p
i
i log n

∑ p log p
i i + log n = − H ( X ) + log n ≥ 0
= i

− H ( X ) ≤ log n

H(X) = log n for uniform probability distribution. Hence proved that entropy of a
discrete source is maximum when output symbols are equally probable. The
quantity D(p ⎜⎜q) is called the Kullback-Leibler Distance.

Q. 1.3 The plots are given below:

1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

2
y = x -1
1
y = ln(x)
0

-1

-2

-3
0 0.5 1 1.5 2 2.5 3 3.5

Q 1.4 Consider two probability distributions: {p0, p1, , pK-1} and {q0, q1, , qK-1}.
K −1
⎛q ⎞ 1 K −1 ⎛q ⎞
We have ∑ pk log 2 ⎜⎜ k ⎟⎟ = ∑ pk ln⎜⎜ k ⎟⎟ . Use ln x ≤ 1 – x,
k =0 ⎝ pk ⎠ ln 2 k =0 ⎝ pk ⎠
K −1
⎛q ⎞ 1 K −1 ⎛ qk ⎞
∑ pk log 2 ⎜⎜ k ⎟⎟ ≤ ∑ pk ⎜⎜ − 1⎟⎟
k =0 ⎝ pk ⎠ ln 2 k =0 ⎝ pk ⎠
1 K −1
≤ ∑ (qk − pk )
ln 2 k =0
1 ⎛ K −1 K −1
⎞
≤ ⎜ ∑ qk − ∑ pk ⎟ = 0
ln 2 ⎝ k =0 k =0 ⎠
K −1
⎛q ⎞
Thus, ∑p k log 2 ⎜⎜ k ⎟⎟ ≤ 0 . (1)
k =0 ⎝ pk ⎠
n m P ( xi , y j )
Now, I(X; Y) = ∑∑ P( xi , y j )log (2)
i =1 j =1 P( xi ) P( y j )

From (1) and (2) we can conclude (after basic manipulations) that I(X;Y) ≥ 0. The
equality holds if and only if P ( xi , x j ) = P( xi ) P( x j ) , i.e., when the input and output
symbols of the channel are statistically independent.

2
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q.1.5 Source X has infinitely large set of outputs P(xi) = 2-i, i = 1, 2, 3, ………
∞ ∞
1
H ( X ) = ∑ p ( xi ) log = ∑ 2 −i log 2 −i
i =1 p ( xi ) i =1
∞
= ∑ i.2
i =1
−i
= 2 bits

Q.1.6 Given: P(xi) = p(1-p)i-1 i = 1, 2, 3, ……..

H ( X ) = − ∑ p (1 − p ) i −1 log {p (1 − p ) i −1 }
i

= − ∑ p (1 − p ) i −1 {log p + (i − 1) log (1 − p )}
i

= − p log p ∑ p (1 − p ) i −1 − p log (1 − p ) ∑ (i − 1) (1 − p) i −1

i =1 i
1 1− p
= − p log p × − p log(1 − p) × 2
p p
⎛ 1− p ⎞ p log p − (1 − p) log(1 − p) 1
= − log p − ⎜⎜ ⎟⎟ log(1 − p) = = H ( p) bits
⎝ p ⎠ p p

Q 1.7 Hint: Same approach as the previous two problems.

Q 1.8 Yes it is uniquely decodable code because each symbol is coded uniquely.

Q 1.9 The relative entropy or Kullback Leibler distance between two probability mass
functions p(x) and q(x) is defined as
⎛ p ( x) ⎞
D( p || q ) = ∑ p ( x) log⎜⎜ ⎟⎟ . (1.76)
x∈X ⎝ q ( x) ⎠
(i) Show that D ( p || q ) is non negative.

⎛ p ( x) ⎞ ⎛ q ( x) ⎞
Solution: − D( p || q ) = − ∑ p ( x) log⎜⎜ ⎟⎟ = ∑ p ( x) log⎜⎜ ⎟⎟
x∈X ⎝ q( x) ⎠ x∈X ⎝ p ( x) ⎠
q( x)
≤ log ∑ p ( x) (from Jensen’s Inequality: Ef(X)≥ f(EX))
x∈X p ( x)

= log ∑ q( x) = log(1) = 0.
x∈X

Thus, − D( p || q ) ≤ 0 or D( p || q ) ≥ 0 .

3
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

p( x)
(ii) D( p || q) = ∑ p( x) log q( x)
Symmetry Property:
D( p || q) = D(q || p )
p( x) q( x)
∑ p( x) log q( x) = ∑ q( x) log p( x)
∑ p( x) log p( x) − ∑ p( x) log q( x) ≠ ∑ q( x) log q( x) − ∑ q( x) log p( x)
Therefore Kullback Leibler distance does not follow symmetry property.

Triangle Inequality:
D( p || q) + D(q || r ) ≥ D( p || r )
p( x) q( x) p ( x)
∑ p( x) log q( x) + ∑ q( x) log r ( x) ≥ ∑ p( x) log r ( x)
On solving this we get
q( x)
∑ (− p( x) + q( x)) log r ( x) ≥ 0
This relation does not hold if p( x) > q( x).
Therefore Kullback Leibler distance does not follow triangle inequality property.

(iii) I ( X ; Y ) = ∑ p ( x, y ) I ( x; y )
log( x, y )
= ∑ p ( x, y )
log p ( x) p( y )
= D( p( x, y ) || p( x) p ( y ))

Q.1.10

4
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q.1.11 The codeword lengths are possible if and only if they satisfy Kraft-McMillan
inequality, which in this case is
1 1 1 d
3
+ 3 + 3 + 8 ≤1
2 2 2 2
d 5
≤
256 8
d ≤ 160

5
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q. 1.12 First note that it is a discrete random variable with a valid probability
∞
1
distribution, since ∑ Pn = ∑ 2
= 1.
n n = 2 An log n

However, H(X) = ∑ − P log P

n
n n = +∞ (the series does not converge!).

Thus the entropy of a discrete random variable can also be infinite.

⎧ a −1 0≤ x≤a
Q.1.13 P ( x) = ⎨
⎩ 0 otherwise

a
Differential entropy = − ∫ p ( x) log p ( x) dx
0
a
1
= − ∫ log a dx = log 2 a
0
a
The plot is given below. Note that the differential entropy can be negative.
H(X)
3

1 log a
2

-1

-2

-3
0 2 4 6 8 10

Q.1.14 DMS with source probabilities {0.35, 0.25, 0.20, 0.15, 0.05}

(i) Huffman code

6
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 7
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

(ii) R = ∑p l
i
i i = 0.35 × 2 + 0.25 × 2 + 0.20 × 2 + 0.15 × 3 + 0.05 × 3 = 2.2

bits.

H (X ) 1
(iii) η=
R
H (X ) = ∑p i log
pi
= 2.121

2.121
η= = 0.964 = 96.4% .
2.2

Q.1.15 (i)

2 1.00
S1 0.35 Codes
3 S1 2
S1 0.25 S2 3
1 S3 11
S1 0.20 S4 12
2 0.40 S5 13
S1 0.15
3
S1 0.05

(ii) R = ∑p l
i
i i = 0.35 + 0.25 + 0.20 × 2 + 0.15 × 2 + 0.05 × 2 = 1.40 ternary

digits/ symbol

Q.1.16 (i) DMS with source probabilities :

Symbols Prob. Efficient fixed length code

S1 0.20 000
S2 0.20 001
S3 0.15 010
S4 0.15 011
S5 0.10 100
S6 0.10 101
S7 0.05 110
S8 0.05 111

7
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 8
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

L
Average code length = R = ∑ n( xk ) P( xk ) = 3
k =1

(ii) Huffman code

L
R = ∑ n( xk ) P( xk ) = 2.9 bits.
k =1
n
(iii) The entropy H(X) = − ∑ P( xi )logP( xi ) = 2.8464 bits
i =1

Huffman code gives shorter code length (η = 98.15%)

Q.1.17 Symbol Probability Self information Code

0
x1 0.5 0.5 0
x2 0.4 0 0.528 10
1
x3 0.1 1 0.332 11

H(X) = 1.36 bits / symbol pair

R1 = 1.5 bits / symbol pair
η = 90.66 %

(ii)

8
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 9
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

H(X) = 2.72 bits / symbol pair

R2 = 2.78 bits / symbol pair
η = 97.84 %

Symbol Self Huffman Probability

Triple Information Codes
x1 x1 x1 0.375 000 0.125
x1 x1 x2 0.3322 110 0.1
x1 x2 x1 0.3322 100 0.1
x2 x1 x1 0.3322 101 0.1
x1 x2 x2 0.2915 0100 0.08
x2 x1 x2 0.2915 0110 0.08
x2 x2 x1 0.2915 0111 0.08
x2 x2 x2 0.2538 0010 0.064
x1 x1 x3 0.1330 11100 0.025
x1 x3 x1 0.1330 11101 0.025
x3 x1 x1 0.1330 010100 0.025
x1 x2 x3 0.11287 010101 0.02
x1 x3 x2 0.11287 010110 0.02
x2 x3 x1 0.11287 010111 0.02
x2 x1 x3 0.11287 001100 0.02
x3 x1 x2 0.11287 001101 0.02
x3 x2 x1 0.11287 001110 0.02
x2 x2 x3 0.09545 111100 0.016
x2 x3 x2 0.09545 111101 0.016
x3 x2 x2 0.09545 111110 0.016
x1 x3 x3 0.0382 1111110 0.005
x3 x1 x3 0.0382 1111111 0.005
x3 x3 x1 0.0382 00111110 0.005
x2 x3 x3 0.03186 00111100 0.004
x3 x2 x3 0.03186 00111101 0.004
x3 x3 x2 0.03186 001111110 0.004
x3 x3 x3 0.00996 001111111 0.001

9
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and10
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

H(X) = 4.0826 bits / triple

R2 = 4.118 bits / symbol triple
H (X )
η = = 99.14 %
R3

Q.1.18 For a B-symbol block x1 x2 …… xB,

n n n
H (x1 x2 …… xB) = − ∑ ∑−∑
j1 =1 j2 =1 j B =1
p ( x j1 x j2 ..... x jB ) log p ( x j1 x j2 ...... x jB )

P(x1 x2 ….. xB) = p(x1) p(x2⎜x1) p(x3⎜x1 x2) …… p(xB|x1 x2 …. xB-1)

Assuming the B r, vs to be statistically independent
∴ H (x1 x2 …… xB) = H(x1) + H(x2) + ….. + H(xB)
= BH(X)

Q 1.19 Hint: Apply equation 1.27.

Q.1.20 Lempel Ziv Code for

01001111100101000001010101100110000
Parsing 0,1,00,11,111,001,01,000,0010,10,101,100,110
gives
Dictionary Dictionary Location Contents Codeword
1 0001 0 00000
2 0010 1 00001
3 0011 00 00010
4 0100 11 00101
5 0101 111 01001
6 0110 001 00111
7 0111 01 00011
8 1000 000 00110
9 1001 0010 01100
10 1010 10 00100
11 1011 101 10101
12 1100 100 10100
13 1101 110 01000

Encoded stream
00000 00001 00010 00101 01001 00111 00011 00110 01100 00100
10101 10100 01000 00110

Q.1.21 (i) Lempel Ziv Code for

133002021113000022122233
Parsing 13,30,02,021,11,300,00,22,12,223
gives
Dictionary Dictionary Location Contents Codeword

10
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and12
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q.1.22 (i) For run length code we encode a run in terms of [how many] and [what].
Therefore to encode a run of n bits we require = ⎡log 2 n ⎤ + 1 bits.
The run length code is therefore

⎡log 2 n1⎤ + 1 ⎡log 2 n2⎤ + 1 ⎡log 2 n3⎤ + 1

Compression it will provide:
Every run requires ⎡log 2 n ⎤ + 1 bits.
1 1 1 1 1
Average length of a run = 1 + 2 2 + 3 3 + − − − − − + (n − 1) ( n −1) + n ( n −1)
2 2 2 2 2
(n + 1)2 − 2 2n 3n
=
[
2 ( n −1) 2 n +1 − 2 n
2
]
Average Compression = ⎡log 2 n⎤ + 1
⎡ (n + 1)2 2 n − 2 3n ⎤
⎢ ( n −1) n +1 2 ⎥
⎢⎣ 2 [
2 − 2 n ⎥⎦ ]
(ii) Huffman Coding

n/2n-1

n-1/2n-1

0
1/24 3/16
1
1/23 15/16
0
1/22 3/4
1
1/2

Q.1.23 For example, let us assume a run of 214 1’s. For first level of run length coding we
require 14+1=15 1’s. For second level of run length coding we need 4+1=5 1’s. This
multi-layer run length coding is useful when we have large runs.
To encode n 1’s maximum possible compression can be calculated as:
For first run required bits = ⎡log 2 n ⎤ + 1
For second run required bits = ⎡log ⎡log 2 n ⎤ + 1⎤ + 1

12
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and13
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

So for a run of n 1’s maximum possible compression =

n
⎡log 2 ⎡log 2 ⎡log 2 n + 1⎤ + 1⎤ + 1⎤ − − − − − +1

13
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 2

Q.2.1 p(y = 0) = p(y = 0 | x = 0) p(x = 0)

+ p(y = 0 ⎜x = 1) p (x = 1)
= (1 - p) p0 + qp1
= (1 - p) p0 + q(1 - p0)

p(y = 1) = p(y = 1 ⎜x = 0) p(x = 0) + p(y = 1 ⎜x = 1) p(x = 1)

= pp0 + (1- q) (1-p0)

p( y = 0 x = 0) p( x = 0) (1 − p) p 0
p(x = 0 ⎜y = 0) = =
p( y = 0) (1 − p ) p0 + q(1 − p0 )

p( y = 1 x = 1) p( x = 1) (1 − q) (1 − p 0 )
p(x = 1 ⎜y = 1) = =
p( y = 1) pp0 + (1 − q) (1 − p0 )

Q 2.2
1–p
p0 0 0
p
e
q
p1 1 1
1–q
p(y = 0) = (1 - p) p0
p(y = e) = p p0 + q(1- p0)
p(y = 1) = (1- q) (1-p0)

2 3 P ( y j | xi )
I (x; y) = ∑∑ P( y j | xi ) P(x j )log
i =1 j =1 P( y j )

1 p 1 q
= (1 − p) p 0 log + pp 0 log + (1 − q )(1 − p 0 ) log + q(1 − p 0 ) log
p0 pp 0 q(1 − p 0 ) (1 − p0 ) pp 0 + q(1 − p 0 )

1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

dI ( X ; Y )
Hint: For capacity find value of p0 by = 0 and then substitute it in above
dp 0
equation.

Q 2.3 (a) CA = max I (x; y)

p
n m P ( xi , y j )
I (x; y) = ∑∑ P( xi , y j )log
i =1 j =1 P( xi ) P( y j )

2 1 1 1 ⎛ 1 ⎞
I ( x, y ) = p log + p log + (1 − p) log ⎜⎜ ⎟⎟
3 p 3 p ⎝ 1 − p ⎠
1 ⎛ 1 ⎞
= p log + (1 − p) log ⎜⎜ ⎟⎟ = − p log p − (1 − p) log(1 − p)
p ⎝ 1 − p ⎠

dI ( x, y )
= − log p − 1 + log(1 − p) + 1 = 0
dp
⇒ - log p = log (1 - p) ⇒ p = 1/2
∴ CA = ½ log 2 + ½ log 2 = log 2 = 1 bit/use.

1 ⎛ 1 ⎞
(b) I(x; y) = p 2 log + p1 log ⎜⎜ ⎟⎟
p2 ⎝ 1− p2 ⎠
⎛ 1 ⎞
(1 − p1 − p 2 ) log ⎜⎜ ⎟
⎟
⎝ 1 − p2 ⎠
1 1
= p 2 log + (1 − (1 − p 2 ) log
p2 1− p2

I(x; y) is maximized for p2 = ½.

∴CB = log 2 = 1 bit/use.

(c)

2
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

p p ⎛ p⎞ ⎛ p⎞
H (Z ) = − log − ⎜1 − ⎟ log⎜1 − ⎟
3 3 ⎝ 3⎠ ⎝ 3⎠
p 1 2p 2
H ( Z | X ) = − log − log
3 3 3 3
I(X; Z) = H(Z) – H(Z|X)
1 1 2 ⎛ 2 ⎞ ⎛ 3 ⎞
I(X; Z) = p log + p log ⎜⎜ ⎟⎟ + (1 − p ) log ⎜⎜ ⎟⎟
3 p 3 ⎝ 3− p ⎠ ⎝3− p⎠

−p 2p 2p
= log p + log 2 − log p (3 − p) + (1 − p) log 3 − (1 − p ) log(3 − p )
3 3 3
−p 2p 2p ⎛ p⎞
= log p + log 2 − + (1 − p) log 3 − ⎜1 − ⎟ log (3 − p)
3 3 3 ⎝ 3⎠

dI ( x; z ) − 1 1 2 1 1
= log p − + − log 3 + log (3 − p ) + = 0
dp 3 3 3 3 3
2 1 ⎛3− p⎞
⇒ − log 3 + log ⎜⎜ ⎟⎟ = 0
3 3 ⎝ p ⎠
⎛3− p⎞
⇒ 2 − 3 log 3 + log ⎜⎜ ⎟⎟ = 0
⎝ p ⎠
⎛3− p⎞
⇒ log⎜⎜ ⎟⎟ = 3 log 3 − 2 = 2.755
⎝ p ⎠
3− p
⇒ = 6.75 ⇒ p = 0.387.
p

CAB = 0.199 bit/use

(d) CAB is less than both CA and CB. This is because we don’t have the flexibility to
choose the input symbol probabilities for the second half of the composite channel. The
plot of I(X; Z) versus probability p is given below.

3
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

0.2

0.15
I(X;Z)
0.1

0.05

p
0
0 0.2 0.4 0.6 0.8 1

Q.2.4

Probabilities of different output symbols

C = max {I ( x ; y )}
pi

⎡ p( y i x j ) ⎤
I ( x ; y) = ∑∑ p( x ) p( y i x j ) log ⎢
j ⎥
j i ⎣⎢ p( y i ) ⎦⎥
0.5 0.5
= 0.5 p1 log + 0.5 (1 − p1 − p 2 − p3 ) log
0.5 (1 − p 2 − p3 ) 0.5 (1 − p 2 − p3 )
0.5 0.5
+ 0.5 p1 log + 0.5 p 2 log
0.5 ( p1 + p 2 ) 0.5 ( p1 + p 2 )

0.5 0.5
+ 0.5 p 2 log + 0.5 p3 log
0.5 ( p 2 + p3 ) 0.5 ( p 2 + p3 )

0.5 0.5
+ 0.5 p3 log + 0.5 (1 − p1 − p 2 − p3 ) log
0.5 (1 − p1 − p 2 ) 0.5 (1 − p1 − p 2 )
= - 0.5 [p1 log (1 - p2 - p3) + (1 - p1 – p2 - p3) log (1 - p2 - p3) + p1 log (p1 + p2)
+ p2 log (p1 + p2) + p2 log (p2 + p3) + p3 log (p2 + p3) + p3 log (1 - p1 - p2)
+ (1 - p1 - p2 - p3) log (1 - p1 - p2) ]
= - 0.5 [ (1 - p2 - p3) log (1 - p2 – p3) + (p1 + p2) log (p1 + p2) + (p2 + p3)
log (p2 + p3) + (1 - p1 - p2) log (1 - p1 - p2) ]

4
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

∂I ( x; y ) 1
= [log( p1 + p 2 ) + 1 − log (1 − p1 − p 2 ) − 1] = 0
∂p1 2
1
⇒ log( p1 + p 2 ) = log (1 − p1 − p 2 ) ⇒ p1 + p 2 = (1)
2
∂I ( x; y )
= − 0.5 [− log(1 − p 2 − p3 ) − 1 + log ( p 2 + p3 ) + 1] = 0
∂p3
⇒ log( p 2 + p3 ) = log (1 − p 2 − p3 )
1
⇒ p 2 + p3 = 1 − p 2 − p3 ⇒ p 2 + p3 = (2)
2
Substituting (1) & (2) in I(x; y) we get
C = -0.5 [0.5 log 0.5 + 0.5 log 0.5 + 0.5 log 0.5 + 0.5 log 0.5]
= - 4/2 × ½ log ½ = log 2 = 1 bit/use.

Q.2.5 (a) C = Blog2 (1 + SNR) SNR = 20 dB = 100

= 3000 log2 (1 + 100)
= 19,974.6 bits/sec ≈ 20 kb/s.

(b) SNR = 25 dB = 316.2

C = 3000 log2 (1 + 316.2) = 24,927.7 bits/sec ≈ 25 kb/s.
It is interesting to note that for this telephone channel under consideration, the
SNR (in dB) directly translates to the capacity (in kb/s). See the graph below!
40

35

30
Capacity

25

20

15
15 20 25 30 35 40
S NR (dB )

Q.2.6

5
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Capacity = max {I ( x ; y )}
{ pi }

⎧ 1− p ⎫ ⎧ p ⎫
I ( x ; y) = p1 (1 − p) log ⎨ ⎬ + p(1 − p1 − p2 ) log⎨ ⎬
⎩ (1 − p) p1 + p(1 − p1 − p2 ) ⎭ ⎩ (1 − p) p1 + p(1 − p1 − p2 ) ⎭

⎧ 1− p ⎫ ⎧ p ⎫
+ p2 (1 − p) log⎨ ⎬ + pp1 log⎨ ⎬
⎩ pp1 + p(1 − p − p2 ) ⎭ ⎩ pp1 + p(1 − p) p2 ) ⎭

⎧ 1− p ⎫ ⎧ p ⎫
+ (1 − p1 − p 2 ) (1 − p ) log ⎨ ⎬ + pp 2 log ⎨ ⎬
⎩ pp 2 + (1 − p )(1 − p1 − p 2 ) ⎭ ⎩ pp 2 + (1 − p )(1 − p1 − p 2 ) ⎭

I(x; y) = - [p1(1-p) + p(1–p1–p2)] log [(1-p)p1 + p(1–p1–p2)] + p1(1-p) log (1-p)

+ p(1–p1–p2)] logp - [p2(1-p) + pp1] log[pp1 + (1-p)p2] + p2[(1 - p) log (1-p)
+ pp1 log p – [1-p1-p2)(1-p) + pp2] log[pp2 + (1-p)(1-p1-p2) (1-p) log(1-p)
+ pp2 logp

∂I ( x; y )
=0 ⇒ (2p-1) log [(1-p)p1 + p(1-p1-p2)] – plog [pp1 + p2(1-p)]
∂p1
+ (1-p) log[pp2 + (1-p) (1-p1-p2)] = 0

∂I ( x; y )
=0 ⇒ p log [(1-p)p1 + p(1-p1-p2)] – (1-p)log [pp1 + (1-p)p2]
∂p 2
- (2p-1) log [pp2 + (1-p) (1-p1-p2)] = 0

Q.2.7 A TV displays 30 frames /second

2 × 105 pixels /frame
16 bits / pixel
∴Transmission rate = 30 × 2 × 105 × 16 bits /second
= 96 × 106 bits / second.
C = B log2 (1 + SNR) SNR = 25dB = 316.22

96 ×10 6
Bandwidth required = = 11.55 ×10 6 Hz
log 2 (1 + 316.22)

Q.2.8

6
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 7
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

1 p
(a) I ( x; y ) = p1 log + p(1 − p1 ) log
p1 + p(1 − p1 ) p1 + p(1 − p1 )

p
+ (1 − p) (1 − p1 ) log
(1 − p) (1 − p1 )

= - [p1 + p(1-p1)] log[p1 + p(1-p1)] – (1-p) – p(1-p1) logp –(1-p)(1-p1) log(1-p1)

dI ( x; y )
= - (1-p) log [p1 + p(1-p1)] – (1-p) – p log p + (1 - p) [1 + log (1 - p1)]
dp1
= - (1 - p) log (1 - p1) - (1-p) log [p1 + p (1 - p1) – p log p

⎡ 1 − p1 ⎤ p
⇒ log ⎢ ⎥= log p
⎣ p1 + p (1 − p1 ) ⎦ 1 − p
p
1 − p1
⇒ = ( P) 1− p = α
p1 + p (1 − p1 )

1−α p
p1 = → Input probability that results in capacity
1 − α p + pα

(b)

p1 = 1, 1 …. N times = 1
p2 = p + (1 - p)p + (1-p)2.p + ….. + (1-p)N-1 p
⎡1 − (1 − p ) N ⎤
=⎢ ⎥ = 1 − (1 − p )
N

⎣ 1 − (1 − p ) ⎦
p3 = (1 - p)N

Equivalent channel for combination of N Z channels

7
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 8
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

(c) As N → ∞, the equivalent channel tends to (provided p ≠ 0)

I(x; y) = p log 1/1 + (1-p) log 1/1 = 0

Thus, the capacity → 0 as N → ∞, provided p ≠ 0.
Of course, the capacity → 1 as N → ∞, for p = 0.

Q.2.9 Given: SNR=20dB

(
(i) Cutoff rate R0 = 1 − log 2 1 + e − E / N 0 )
= 0.99

(ii) Pe < 2 n ( R0 − Rc )
Assuming binary antipodal signaling-
10-6 < 2 –(0.99-Rc)

Rc < 18.94

(iii) n is based on the cutoff rate. Cutoff rate is a design parameter.

∞
p( x)
Q.2.10 Consider D ( p q) ∆ ∫ p( x) log q( x) dx
−∞
Where p(x) and q(x) are two probability distributions.
∞ ∞
q( x) ⎛ q( x) ⎞
− D ( p q) ∆ ∫ p( x) log dx ≤ ∫ p( x) ⎜⎜ − 1⎟⎟dx
−∞
p ( x) −∞ ⎝ p ( x) ⎠
∞ ∞
= ∫ q( x) dx − ∫ p( x) dx = 0
−∞ −∞

⇒ D ( p q) ≥ 0
x2
1 −
2σ
Now, let q(x) be the Gaussian distribution : e
2πσ 2
Let p(x) → arbitrary distribution with variance σ2.

8
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 9
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

∞
p( x)
D ( p q) ∆ ∫ p( x) log q( x) dx
−∞

∞ ∞
⎛ 1 ⎞
= ∫
−∞
p ( x ) log ⎜
2 ⎜ ⎟
⎟
⎝ q( x) ⎠
dx + ∫ p( x) log 2 p( x) dx
−∞
∞ ⎛ x2 ⎞

= ∫ p ( x) log 2 ⎜ 2πσ 2 e 2σ ⎟ dx − H ( X )
1 2

2 ⎜ ⎟
−∞ ⎝ ⎠
Differential entropy for differential entropy
Gaussian distribution
1 1
= log (2πσ 2 ) + log 3 − H ( X )
2 2
1
= log 2 (2π eσ 2 ) − H ( X )
2
1
D ( p q) ⇒ H ( X ) ≤ log 2 (2π eσ 2 )
2
Differential entropy for Gaussian distribution.

9
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 3

Q3.1 C = {0000, 1100, 0011, 1111}

Since:
(i) All zero codeword {0000} is a valid code word.
Sum of any two code is also a valid code word. Hence ‘C’ is a linear code.
(ii) Minimum distance = Minimum weight = 2

1100 1100 0011

+ 0011 + 1111 + 1111
1111 0011 1100

Q3.2 (i) (6, 1, 6) ⇔ (n, k, d*)

⎡ d − 1⎤ ⎡ 6 − 1⎤
t=⎢ ⎥ = ⎢ ⎥ =2
⎣ 2 ⎦ ⎣ 2 ⎦

Hamming bound for binary codes

⎡ ⎛n⎞ ⎛n⎞ ⎛ n ⎞⎤
⎢1 + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ + ........ + ⎜⎜ ⎟⎟ ⎥ ≤ 2 n − k
⎣ ⎝1⎠ ⎝2⎠ ⎝ t ⎠⎦

⎛6⎞ ⎛6⎞
⇒ 1 + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ ≤ 2 5
⎝1⎠ ⎝2⎠

⇒ 22 < 32.

Hence (6, 1, 6) is a valid code word and it is simply the repetition code
with n = 6.

(ii) (3, 3, 1) ⇒ d* = 1 ⇒ t = 0.

By Hamming bound [1 + 0] ≤ 2º ≤ 1 = 1.

Hence (3, 3, 1) is also valid code construction.

This code is equivalent to sending information bits as such with no

change, hence does not have any error detecting or correction capability.

(iii) (4, 3, 2) ⇒ d* = 2 ⇒ t = 0.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

By Hamming bound 1 < 2.

Hence it is a valid code construct (Parity code)

⎡1 0 1 0 0⎤
Q3.3 G = ⎢⎢ 1 0 0 1 1 ⎥⎥
⎢⎣ 0 1 0 1 0 ⎥⎦ 3 x 5

⇒ k = 3, n = 5.

(i) All possible code words

C=I.G

⎡0 0 0 ⎤
⎢0 0 1 ⎥
⎢ ⎥
⎢0 1 0 ⎥
⎢ ⎥ ⎡1 0 1 0 0⎤
0 1 1 ⎢1 0 0 1 1⎥
C= ⎢ ⎥
⎢ ⎥
⎢1 0 0 ⎥
⎢ ⎥ ⎢⎣ 0 1 0 1 0 ⎥⎦
⎢1 0 1 ⎥
⎢1 1 0 ⎥
⎢ ⎥
⎢⎣ 1 1 1 ⎥⎦

⎡ 0 0 0 0 0 ⎤
⎢ 0 1 0 1 0 ⎥
⎢ ⎥
⎢ 1 0 0 1 1 ⎥
⎢ ⎥
1 1 0 0 1
C ⎢
= ⎥
⎢ 1 0 1 0 0 ⎥
⎢ ⎥
⎢ 1 1 1 1 0 ⎥
⎢ 0 0 1 1 1 ⎥
⎢ ⎥
⎢⎣ 0 1 1 0 1 ⎥⎦

(ii) G in systematic form:

⎡1 0 0 1 0⎤
G = ⎢⎢ 0 1 0 1 1 ⎥⎥
⎢⎣ 0 0 1 0 1 ⎥⎦
Jk P
Parity check matrix H = [- PT, In-k]
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡1 1 0 1 0 ⎤
H= ⎢ ⎥
⎣0 1 1 0 1 ⎦

(iii) Generator matrix of an equivalent systematic code

⎡1 0 1 0 0⎤
G = ⎢⎢ 1 1 0 1 0 ⎥⎥
⎢⎣ 0 1 0 0 1 ⎥⎦
(iv) d* = 2; Hence it can detect 1 error but can’t correct that error

⎡1 0 ⎤
⎢1 1 ⎥
⎢ ⎥
HT = ⎢0 1 ⎥
⎢ ⎥
⎢1 0 ⎥
⎢⎣ 0 1 ⎥⎦
⎡0 0 0 0 0 ⎤ ⎡0 0 ⎤
⎢0 ⎥ ⎡1 0 ⎤ ⎢0 1 ⎥
⎢ 0 0 0 1 ⎥ ⎢1 ⎥ ⎢ ⎥
⎢ 1 ⎥
⎢0 0 0 1 0 ⎥ ⎢1 0 ⎥
Syndrome = eH T = ⎢ ⎥ ⎢0 1 ⎥ = ⎢ ⎥
⎢0 0 1 0 0 ⎥ ⎢ ⎥ ⎢0 1 ⎥
⎢0 1 0 0 0 ⎥ ⎢1 0 ⎥ ⎢1 1 ⎥
⎢ ⎥ ⎢⎣ 0 1 ⎥⎦ ⎢ ⎥
⎣⎢1 0 0 0 0 ⎦⎥ ⎣⎢1 0 ⎦⎥

We can choose 4 error patterns with distinct syndrome which can be

corrected using this syndrome table.

00000 → 00
00100 → 01
01000 → 11
10000 → 10

Standard Array

00000 01010 10011 11001 10100 11110 00111 01101

00100 01110 10111 11101 10000 11010 00017 01001
01000 00010 11011 10001 11100 10110 01111 00101
10000 11010 00011 01001 00100 01110 10111 11101

(v) Minimum distance, d* = 2.

(vi) Since d* = 2, number of errors it can detect = d* - 1 = 1.

(vii) The error patterns that can be detected:

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

00001
00010
00100
01000
d*−1
10000it can correct =
(viii) Number of errors =0
2

(ix) Since it cannot correct any error. The symbol error is same as uncoded
probability of error. However, since 1 bit error can be detected, a request
for repeat transmission can be made on this basis. This scheme will also
fail if 2 or more bits are in error. Thus, if automatic repeat request (ARQ)
is used, the probability of error will be

⎛5⎞ ⎛5⎞ ⎛5⎞ ⎛5⎞

p(e) = ⎜⎜ ⎟⎟ p 2 (1 − p) 5 + ⎜⎜ ⎟⎟ p 3 (1 − p ) 2 + ⎜⎜ ⎟⎟ p 4 (1 − p) + ⎜⎜ ⎟⎟ p 5
⎝2⎠ ⎝3⎠ ⎝4⎠ ⎝5⎠

(x) Since all zero is a valid codeword and sum of any two code words is also a
valid codeword, it is a linear code.

Q3.4 C = {00000, 10100, 01010, 11111}

For this, (n, k) = (5, 2)

By observation if we use the mapping as follow

Information Word Codeword

00 → 00000
01 → 10101
10 → 01010
11 → 11111

⎡1 0 1 0 1⎤
G= ⎢ ⎥
⎣0 1 0 1 0⎦

Q3.5 Given: (n, k, d*) is a binary code with d* even. We need to show that there exists
a (n, k, d*) code in which all codewords have even weight.

Construct a code with all codewords of even weight by adding even a parity bit to
all the codewords of the given binary (n, k, d*) code with d* even. By adding an
even parity, all code words will become even weight and also since d* is even.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

This process of adding a parity bit will not alter the minimum weight. Hence, the
d* will be even for this new code.

Q3.6 Let C be a linear code (binary).

If the codewords C1, C2 ∈ C, then, C1 ⊕ C2 = C3 ∈ C.
Let us now add a parity check bit to the codewords of C.
Let Pi be the parity bit of Ci and let this new augmented codeword be denoted by
C i . By definition, the parity bit is the binary sum of all the elements of the
codeword vector, i.e.:
P1 = C11 ⊕ C12 ⊕ …. ⊕ C1n
P2 = C21 ⊕ C22 ⊕ …. ⊕ C2n
P3 = C31 ⊕ C32 ⊕ …. ⊕ C3n
Now, C3 = C1 ⊕ C2
= (C11 ⊕ C12 ⊕ …. ⊕ C1n) ⊕ (C21 ⊕ C22 ⊕ …. ⊕ C2n)
P3 = P1 ⊕ P2.

Hence even after adding an overall parity bit,

C1 ⊕ C 2 = C 3
Thus the modified code is also binary linear code.

Q3.7 (i) New codeword C3 consists of U1 U2 U3----- Un V1 V2 V3 ------- Vn

So in new code there are total 2n codewords.
As U1 U2 U3----- Un encodes 2k1 words and V1 V2 V3 ------- Vn encodes 2k2
words, the new code encodes 2k1.2k2 = 2k1+k2 words.
Therfore C3 is (2n, k1+k2) code.

(ii) U1 U2 U3----- Un => minimum distance d1

V1 V2 V3 ------- Vn => minimum distance d2
For C3 (U|U+V) if U is all zero codeword then minimum distance is d2.
If V is all zero codeword then minimum distance is 2d1.
Therefore minimum distance of C3 is min{2d1, d2}

Q3.8 (a) S = {0101, 1010, 1100}

<S> = {0000, 0101, 1010, 1100, 1111, 1001, 0110, 0011}

(b) S = {1000, 0100, 0010, 0001}

<S> = {0000, 1000, 0100, 0010, 0001, 1100, 1010, 1001, 0110, 0101,
0011, 1110, 1101, 1011, 0111, 1111}

(c) S = {11000, 01111, 11110, 01010}

<S> = {00000, 11000, 01111, 01010, 10111, 00110, 10010, 10001, 00101,
10100, 01001, 11101, 01100, 11011, 11110, 00011}
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q3.9 Binary code (23, 12, 7)

d*−1
t= = 3 , errors that can be corrected.
2
p = 0.01.
⎛ 23 ⎞ 4 ⎛ 23 ⎞ 5 ⎛ 23 ⎞ 23
P(e) = ⎜⎜ ⎟⎟ p (1 − p)19 + ⎜⎜ ⎟⎟ p (1 − p)18 + .......... + ⎜⎜ ⎟⎟ p
⎝ 4 ⎠ ⎝ 5 ⎠ ⎝ 23 ⎠

⎛ 23 ⎞ ⎛ 23 ⎞ ⎛ 23 ⎞ 2 ⎛ 23 ⎞
= 1 − ⎜⎜ ⎟⎟ (1 − p ) 23 − ⎜⎜ ⎟⎟ p(1 − p) 22 − ⎜⎜ ⎟⎟ p (1 − p ) 21 − ⎜⎜ ⎟⎟ p 3 (1 − p) 20
⎝ 0 ⎠ ⎝ 1 ⎠ ⎝ 2 ⎠ ⎝3⎠

= 1 – 079361 – 0.18437 – 0.02048 – 0.00144

= 1 – 0.99992 ≈ 0.00008.

Q3.10 C is a binary code with parity check matrix H.

C1 → H1 ⇒ CHT = 0 ⇒ GHT = 0
C1 = [C : P]
G1 = [G : P]
G1 H 1T = [G : P] H 1T
⎡ | 1⎤
⎢ 1 ⎥
⎢ HT | ⎥
= [G : P] ⎢ ⎥
⎢ | 1⎥
⎢_ _ _ _ _ _⎥
⎢ ⎥
⎣0 0 .... 0 | 1⎦

= [GHT G ⊕ P]
= 0.
H : 0
Hence H1 = : 0
..... : :
1 1 1 1...1

is the parity check matrix of the extended code C, obtained from C by adding an
overall parity bit.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 7
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q3.11 (5, 3) code over GF (4)

+ 0 1 2 3 . 0 1 2 3
0 0 1 2 3 0 0 0 0 0
1 1 0 3 2 1 0 1 2 3
2 2 3 0 1 2 0 2 3 1
3 3 2 1 0 3 0 3 1 2

⎡1 0 0 1 1 ⎤
G = ⎢⎢0 1 0 1 2⎥⎥
⎢⎣0 0 1 1 3⎥⎦
I P

(i) Parity check matrix

⎡− 1 − 1 − 1 1 0⎤
H = ⎢
⎣− 1 − 2 − 3 0 1⎥⎦
⎡1 1 1 1 0⎤
⇒H = ⎢ since -1 = 1, -2 = 2, -3 = 3 over GF (4).
⎣1 2 3 0 1⎥⎦
(ii) Minimum distance, d* = 3.

Hence this code can detect 2 errors.

(iii) This code can correct 1 error.

(iv) With erasures d * ≥ 2t + 1 + r

As t = 1
r=0

(v) Check for the condition for perfect code:

⎧⎛ n ⎞ ⎛ n ⎞ ⎛n⎞ ⎛n⎞ ⎫
M ⎨⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟(q − 1) + ⎜⎜ ⎟⎟(q − 1) 2 + ... + ⎜⎜ ⎟⎟(q − 1) t ⎬ = q n
⎩⎝ 0 ⎠ ⎝ 1 ⎠ ⎝ 2⎠ ⎝t ⎠ ⎭

Here, n = 5, k = 3, t = 1, q = 4, M = 43. Yes, it’s a perfect code!

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 8
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

3.13 Hamming bound for non-binary.

⎡ ⎛n⎞ ⎛n⎞ ⎛ n ⎞⎤
⎢1 + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟⎥ = 2 n − k
⎣ ⎝1⎠ ⎝2⎠ ⎝ 3 ⎠⎦

if n = 7 [1 + 7 + 21 + 35] = 27-k
64 = 27-k
which can be satisfied for k = 1.
∴binary perfect code with d* = 7 & n = 7 is possible.
If n = 23
1 + 23 + 253 + 1771 = 223-k
2048 = 223-k
⇒ k = 12.
∴binary perfect code with d* = 7 & n = 23 is possible.

k 2m − 1
Q3.14 rH = = m
n 2 −1− m
2m − 1
lt rH = lt m
k →∞ m →∞ 2 − 1 − m

Since 2m terms goes toward ∞, exponentially. Hence we can neglect the effect of
m & 1 both
2m
⇒ lt rH = lt m = 1 .
k →∞ m →∞ 2

Q3.15 Hint: Check for the optimality of codes with a larger value of n, but same k and t,
or with a smaller value of k with the same n and t. Then draw a conclusion.

Q.3.16 For orthogonality every column must be orthogonal to other column i.e. inner
product of 2 column vectors must be 0.

Column(1,2) inner product

(x1, x2, x3).(-x2*, x1*, 0) = 0
Column(1,3) inner product
(x1, x2, x3).(-x3*, 0, x1*) = 0
Column(1,4) inner product
(x1, x2, x3).(0, -x3*, x2*) = 0
Column(2,3) inner product
(-x2*, x1*, 0).(-x3*, 0, x1*) ≠ 0
Therefore the given STBC code is not orthogonal.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 4

Q4.1 (a) → Not a linear cyclic block code since {1111} missing.
(b) → Not cyclic.
(c) → Not cyclic.
(d) → Cyclic equivalent.
(e) → Cyclic.

F [ x]
Q4.2 (a) over GF(2) contains {0, 1, x, 1 + x}.
x2 + 1
The addition and multiplication tables are given below:

+ 0 1 x 1+x . 0 1 x 1+x
0 0 1 x 1+x 0 0 0 0 0
1 1 0 x+1 1 1 0 1 x x+1
x x x+1 0 1 x 0 x 1 x+1
1+x 1+x x 1 0 1+x 0 x+1 x+1 0

F [ x]
over GF(2) is not a field because of the absence of a multiplicative
x2 + 1
inverse for 1+x.

F [ x]
Also, over GF(2) is not a field because (x2 +1) in not prime polynomial
x2 + 1
over GF(2).

F [ x]
(b) over GF(3) is a field because (x2 + 1) is prime over GF(3).
x +1
2

F [ x]
over GF(3) contains {0, 1, 2, x, x +1, x + 2, 2x, 2x+1, 2x+2}.
x2 + 1

+ 0 1 2 x x+1 x+2 2x 2x+1 2x+2

0 0 1 2 x x+1 x+2 2x 2x+1 2x+2
1 1 2 0 x+1 x+2 x 2x+1 2x+2 2x
2 2 0 1 x+2 x x+1 2x+2 2x 2x+1
x x x+1 x+2 2x 2x+1 2x+2 0 1 2
x+1 x+1 x+2 x 2x+1 2x+2 2x 1 2 0
x+2 x+2 x x+1 2x+2 2x 2x+1 2 0 1
2x 2x 2x+1 2x+2 0 1 2 x x+1 x+2
2x+1 2x+1 2x+2 2x 1 2 0 x+1 x+2 x
2x+2 2x+2 2x 2x+1 2 0 1 x+2 x x+1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

. 0 1 2 x x+1 x+2 2x 2x+1 2x+2

0 0 0 0 0 0 0 0 0 0
1 0 1 2 x x+1 x+2 2x 2x+1 2x+2
2 0 2 1 2x 2x+2 2x+1 x x+2 x+1
x 0 x 2x 2 x+2 2x+2 1 x+1 2x+1
x+1 0 x+1 2x+2 x+2 2x 1 2x+1 2 x
x+2 0 x+2 2x+1 2x+2 1 x x+1 2x 2
2x 0 2x x 1 2x+1 x+1 2 2x+2 x+2
2x+1 0 2x+1 x+2 x+1 2 2x 2x+2 x 1
2x+2 0 2x+2 x+1 2x+1 x 2 x+2 1 2x

This is a field because it satisfies all the eight conditions stated in Definition 3.9.

Q4.3 (a) Irreducible polynomials over GF(2) of degree 1 to 5.

x, x+1, x2+x+1, x3 + x + 1, x3 + x2 + 1, x4 + x3 + 1, x4 + x + 1, x4 + x3 + x2 + x
+ 1, x5 + x4 + x3 + x + 1, x5 + x4 + x2 + x + 1, x5 + x3 + x2 + x + 1, x5 + x3 + 1,
x5 + x2 + 1.

(b) Irreducible polynomials over GF(3) of degree 1 to 3.

x, 2x, x+1, x+2, 2x+1, x2 + 2x + 2, x2 + 1, + x2 + x + 2, 2x2 + x + 1, 2x2 + 2x + 1,
x3 + x2 + 2, 2x3 + x +1 , 2x3 + x + 2, x3 +2x + 1, x3 + 2x + 2, x3 + x2 + x + 2, x3 +
x2 + 2x + 1, x3 + 2x2 + 1, x3 + 2x2 + 2x + 2, 2x3 + x2 + 2, 2x3 + x2 + x + 1, 2x3 + x2
+ 2x + 2, 2x3 + 2x2 + 1, 2x3 + 2x2 + x + 2, 2x3 + 2x2 + 2x + 1.

Q4.4 x5 + 1 = (x + 1) (x4 + x3 + x2 + x + 1), hence two cyclic nontrivial codes of length

5 are possible.
(i) With g(x) = x4 + x3 + x2 + x + 1
n = 5, n - w = 4 k = 1
G = [1 1 1 1 1 ]1x5
Min distance d* = 5.

(ii) With g(x) = x+1.

⎡1 1 0 0 0 ⎤
⎢0 1 1 0 0 ⎥
G= ⎢ ⎥
⎢0 0 1 1 0 ⎥
⎢ ⎥
⎣0 0 0 1 1 ⎦ 4 x5

Systematic form of G.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡1 0 0 0 1 ⎤
⎢0 1 0 0 1 ⎥
G= ⎢ ⎥ H = [1 1 1 1 1]
⎢0 0 1 0 1 ⎥
⎢ ⎥
⎣0 0 0 1 1 ⎦

d* = 2.

Q4.5 xn -1 is product of r distinct irreducible polynomials over GF(q).

Hence 2r – 2, different nontrivial cyclic codes of length n exits (excluding trivial

case of g(x) = 1 & g(x) = xn – 1.

If g(x) is given, then

Minimum distance of codes ≤ Weight of generator matrix.

Q4.6 (a) To factorize x8 – 1 over GF(3)

(x8 – 1) = (x4 – 1) (x4 + 1)

= (x2 – 1) (x2 + 1) (x2 + x + 2)(x2 + 2x + 2)
= (x – 1) (x + 1) (x2 + x + 2) (x2 + 2x + 2) (x2 + 1)
= (x + 1) (x + 2) (x2 + 1) (x2 + x + 2) (x2 + 2x + 2)

(b) Hence 25 -2 = 30 non-trivial ternary code exists.

(c) For GF(4):

(x2 - y2) = (x2 + y2)
x8 – 1 = x8 + 1 = (x+ 1)8
Hence 7 non-trivial quaternary cyclic codes exist. The generator polynomials
are: gr(x) = (x+ 1)r for r = 1, 2,…,7.

Q4.7 g(x) = x10 + x8 + x5 + x4 + x2 + x + 1 n = 15 ⇒ k = 5, since degree of

g(x) is (n – k).

⎡1 0 1 0 0 1 1 0 1 1 1 0 0 0 0 ⎤
⎢0 1 0 1 0 0 1 1 0 1 1 1 0 0 0 ⎥⎥
⎢
(a) G = ⎢ 0 0 1 0 1 0 0 1 1 0 1 1 1 0 0 ⎥
⎢ ⎥
⎢0 0 0 1 0 1 0 0 1 1 0 1 1 1 0 ⎥
⎢⎣ 0 0 0 0 1 0 1 0 0 1 1 0 1 1 1 ⎥⎦

Minimum distance, d* = 7.
(b) See at the end of the solution for this question.

(c) Errors that can be detected = 7 – 1 = 6

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

(d) Errors that can be corrected (7-1)/2 = 3.

(e) Systematic form of G

⎡1 0 0 0 0 1 0 1 0 0 1 1 0 1 1⎤
⎢0 1 0 0 0 1 1 1 1 0 1 0 1 1 0 ⎥⎥
⎢
G = ⎢0 0 1 0 0 0 1 1 1 1 0 1 0 1 1⎥
⎢ ⎥
⎢0 0 0 1 0 1 0 0 1 1 0 1 1 1 0⎥
⎢⎣ 0 0 0 0 1 0 1 0 0 1 1 0 1 1 1 ⎥⎦
I P

(b) H = [- PT In-k]

⎡ 1 1 0 1 0 1 0 0 0 0 0 0 0 0 0 ⎤
⎢ 0 1 1 0 1 0 1 0 0 0 0 0 0 0 0 ⎥⎥
⎢
⎢ 1 1 1 0 0 0 0 1 0 0 0 0 0 0 0 ⎥
⎢ ⎥
⎢ 0 1 1 1 0 0 0 0 1 0 0 0 0 0 0 ⎥
⎢ 0 0 1 1 1 0 0 0 0 1 0 0 0 0 0 ⎥
= ⎢ ⎥
⎢ 1 1 0 0 1 0 0 0 0 0 1 0 0 0 0 ⎥
⎢ 1 0 1 1 0 0 0 0 0 0 0 1 0 0 0 ⎥
⎢ ⎥
⎢ 0 1 0 1 1 0 0 0 0 0 0 0 1 0 0 ⎥
⎢ 1 1 1 1 1 0 0 0 0 0 0 0 0 1 0 ⎥
⎢ ⎥
⎣⎢ 1 0 1 0 1 0 0 0 0 0 0 0 0 0 1 ⎦⎥

Q4.8 g(x) = x6 + 3x5 + x4 + x3 + 2x2 + 2x + 1 over GF(4)

n – k = 6, n = 15, ⇒ k = 9

(a) Check the validity of g(x)

x6 + 3x5 + x4 + x3 + 2x2 + 2x + 1) x15 + 1 (x9+3x8+3x7+2x5+x4+2x2 + 2x + 1
x15 + 3x14 + x13 + x12 + 2x11 + 2x10 + x9
3x14 + x13 + x12 + 2x11 + 2x10 + x9 + 1
3x14 + 2x13 + 3x12 + 3x11 + x10 + x9 + 3x8
3x13 + 2x12 + x11 + 3x10 + 3x8 + 1
3x13 + 2x12 + 3x11 + 3x10 + x9 + x8 + 3x7
2x11 + x9 + 2x8 + 3x7 + 1
2x11 + x10 + 2x9 + 2x8 + 3x7 + 3x6 + 2x6
x10+ 3x9 + 3x6 + 2x6 + 1
x10 + 3x9 + x8 + x7 + 2x6 + 2x5 + x4
x8 + x7 + x6 + x4 + 1
x8 + 3x7 + x6 + x5 + 2x4 + 2x3 + x2
2x7 + x5 + 3x4 + 2x3 + x2 + 1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

2x7 + x6 + 2x5 + 2x4 + 3x3 + 3x2 + 2x

x6 + 3x5 + x4 + x3 + 2x2 + 2x + 1
x6 + 3x5 + x4 + x3 + 2x2 + 2x + 1

No remainder.

Hence g(x) is a valid generator polynomial.

(b) x15 + 1 = (x9 + 3x8 + 3x7 + 2x5 + x4 + 2x2 + 2x + 1).g(x) = h(x).g(x)

⎡ 1 3 3 0 2 1 0 1 2 1 0 0 0 0 0 ⎤
⎢ 0 1 3 3 0 2 1 0 1 2 1 0 0 0 0 ⎥⎥
⎢
⎢ 0 0 1 3 3 0 2 1 0 1 2 1 0 0 0 ⎥
H = ⎢ ⎥
⎢ 0 0 0 1 3 3 0 2 1 0 1 2 1 0 0 ⎥
⎢ 0 0 0 0 1 3 3 0 2 1 0 1 2 1 0 ⎥
⎢ ⎥
⎣⎢ 0 0 0 0 0 1 3 3 0 2 1 0 1 2 1 ⎦⎥

(c) d* = 5 (five columns of H add up to the zero vector).

(d) Code rate = k/n = 9/15 = 3/5.

(e) v(x) = x8 + x5 + 3x4 + x3 + 3x + 1.

V=[0 0 0 0 0 01 0 0 1 3 1 0 3 1]

⎡1 0 0 0 0 0 ⎤
⎢3 1 0 0 0 0 ⎥⎥
⎢
⎢3 3 1 0 0 0 ⎥
⎢ ⎥
⎢0 3 3 1 0 0 ⎥
⎢2 0 3 3 1 0 ⎥
⎢ ⎥
⎢1 2 0 3 3 1 ⎥
⎢0 1 2 0 3 3 ⎥
⎢ ⎥
HT = ⎢1 0 1 2 0 3 ⎥
⎢2 1 0 1 2 0 ⎥
⎢ ⎥
⎢1 2 1 0 1 2 ⎥
⎢ ⎥
⎢0 1 2 1 0 1 ⎥
⎢0 0 1 2 1 0 ⎥
⎢ ⎥
⎢0 0 0 1 2 1 ⎥
⎢0 0 0 0 1 2 ⎥
⎢ ⎥
⎢⎣ 0 0 0 0 0 1 ⎥⎦
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

VHT = [ 1 _ _ _ _ _ ] non zero.

Hence v(x) = x8 + x6 + 3x4 + x5 + 3x + 1 is not a valid code word.
Also we can check this by dividing x8 + x6 + 3x4 + x3 + 3x + 1 by x6 + 3x6
+ x4 + x3 + 2x2 + 2x + 1 and since reminder is not zero, so it is not valid code
word.

Q4.9 Since we are working over GF(2), -1 = +1.

For Hamming code, gH(x) = (x3 + x + 1)
g1(x) = (x+1) (x3 + x + 1)
= x4 + x3 + x2 + 1.
n = 7, n – k = 4, k = 3

⎡1 1 1 0 1 0 0 ⎤
G = ⎢⎢ 0 1 1 1 0 1 0 ⎥⎥
⎣⎢ 0 0 1 1 1 0 1 ⎥⎦

⎡1 0 0 1 1 0 1 ⎤
= ⎢⎢ 0 1 0 1 1 1 0 ⎥⎥
⎢⎣ 0 0 1 0 1 1 1 ⎥⎦

⎡1 1 0 1 0 0 0 ⎤
⎢1 1 1 0 1 0 0 ⎥⎥
H1 = ⎢ d min = 4
⎢0 1 1 0 0 1 0 ⎥
⎢ ⎥
⎣1 0 1 0 0 0 1 ⎦

4 −1
t= =1
2

So this code can correct one bit error.

S = eHT
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 7
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡ 0 0 0 0 0 0 0 ⎤ ⎡ 0 0 0 0 ⎤
⎢ 0 0 0 0 0 0 1 ⎥⎥ ⎢ 0 0 0 1 ⎥⎥
⎢ ⎢
⎢ 0 0 0 0 0 1 0 ⎥ ⎢ 0 0 1 0 ⎥
⎢ ⎥ ⎢ ⎥
⎢ 0 0 0 0 1 0 0 ⎥ 0 1 0 0 ⎥
⎡ 1 1 0 1 ⎤ ⎢
⎢ 0 0 0 1 0 0 0 ⎥ ⎢ ⎢ 1 0 0 0 ⎥
⎢ ⎥ ⎢ 1 1 1 0 ⎥⎥ ⎢ ⎥
⎢ 0 0 1 0 0 0 0 ⎥
⎢ ⎢ 0 1 1 1 ⎥
0 1 1 1 ⎥ ⎢
⎢ 0 1 0 0 0 0 0 ⎥ ⎢ ⎥ 1 1 1 0 ⎥
= ⎢ ⎥ .⎢ 1 0 0 0 ⎥=⎢ ⎥
⎢ 1 0 0 0 0 0 0 ⎥
⎢ ⎢ 1 1 0 1 ⎥
0 1 0 0 ⎥ ⎢
⎢ 1 1 0 0 0 0 0 ⎥ ⎢ ⎥ 0 0 1 1 ⎥
⎢ ⎥ ⎢ 0 0 1 0 ⎥ ⎢ ⎥
⎢ 0 1 1 0 0 0 0 ⎥ ⎢ ⎢ 1 0 0 1 ⎥
⎢ ⎥ ⎣ 0 0 0 1 ⎥⎦ ⎢ ⎥
⎢ 0 0 1 1 0 0 0 ⎥ ⎢ 1 1 1 1 ⎥
⎢ 0 0 0 1 1 0 0 ⎥ ⎢ 1 1 0 0 ⎥
⎢ ⎥ ⎢ ⎥
⎢ 0 0 0 0 1 1 0 ⎥ ⎢ 0 1 1 0 ⎥
⎢⎣ 0 0 0 0 0 1 1 ⎥⎦ ⎢⎣ 0 0 1 1 ⎥⎦

The first matrix contains ALL single errors and adjacent double errors. The
second matrix is the HT. The matrix on the RHS is the syndrome matrix.

Since we can map all the two adjacent possible errors to district syndrome
vectors, we will be able to correct all the two adjacent double error.

Q4.10 Hint: Work along similar lines as Example 4.28.

Q4.11 Hint: Start from the basic definition of the Fire code.

Q4.12 (i) Assuming this Fire code over GF(2). For a valid generator polynomial of a Fire
code g(x) = (x2t-1-1) p(x) where p(x) does not divide (x2t-1-1).

x6 + x +1) x11 + 1 ( x5 + 1
x11 + x6 + x5
x6 + x5 + 1
x6 + x +1
x5 + x

Therefore g(x) is a valid generator polynomial.

(ii) A Fire code can correct all burst errors of length t or less. So this code can
correct all burst of length 6 or less.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 8
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q.4.13 g(x) is a codeword that is a burst of length (n-k)+1. Probability of burst error of
1
length (n-k) + 1 can not be detected is ( n =k ) +1 .
2
Therefore g(x) can detect a fraction 1-2-(n-k+1) of all burst pattern of length (n-
k+1).

Q.4.14 Shift Register Encoder:

g(x) = x6 + 3x5 + x4 + x3 + 2x2 + 2x +1

x x2 x3 x4 x5 x6

1
2 2 3 3

Meggitt Decoder:

Shift register circuit for dividing by g(x).

x x2 x3 x4 x5 x6

2 2

For Meggitt decoder refer flow chart of Fig. 4.9.

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 9
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q4.15 g(x) = (x23 + 1) (x17 + x3 + 1)

g(x) = x40 + x26 + x23 + x17 + x3 + 1

n – k = 40. From Singleton bound

d * ≤ n − k + 1 = 41 .

⎡ 41 − 1⎤
(i) Errors that can be corrected ≤ ⎢
⎣ 2 ⎥⎦
≤ 20.
(ii) Burst errors that can be corrected
⎡1 ⎤
≤ ⎢ (n − k )⎥ ≤ 20.
⎣2 ⎦

However, the actual d* = 6. Thus the Singleton bound is quite loose here. This
g(x), however, is excellent for detecting and correcting burst errors
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 5

Q5.1 The addition and multiplication tables for of GF(3) are:
+ 0 1 2 . 0 1 2
0 1 2 0 0 0 0
1 2 0 1 0 1 2
2 0 1 2 0 2 1

Representation of GF(32):
Choose the primitive polynomial p(x) = x2 + x + 2
Table 1
Exponential Polynomial Ternary Decimal Minimal
Notation Notation Notation Notation Notation
0 0 00 0 1
0
α 1 01 1 x+2
α 1
z 10 3 X2 + x + 2
α2 2z + 1 21 7 x2 + 1
α 3
2z + 2 22 8 x2 + x + 2
4
α 2 02 2 x +1
5 2
α 2z 20 6 x + 2x + 2
α 6
Z + 2 12 5 x2 + 1
α 7
Z+1 11 4 x2 + 2x + 2

GF(9) Addition and Multiplication Tables:

+ 0 1 z 2z+1 2z+2 2 2z z+2 z+1
0 0 1 z 2z+1 2z+2 2 2z z+2 z+1
1 1 2 z+1 2z+2 2z 0 2z+1 2 z+2
z z z+1 2z 1 z z+2 0 2z+2 2z+1
2z+1 2z+1 2z+2 1 2z+2 z 2z 2z+1 z+2 6
2z+2 2z+2 2z 2 2 Z+1 2z+1 1 2z+2 Z
2 2 0 z+2 2z 2z+1 1 2z+2 Z+1 Z
2z 2z 2z+1 0 z+1 Z+2 1 2z+2 z+1 2
Z+2 z+2 2 2z+2 0 1 z+1 2 2z+1 2z
Z+1 z+1 z+2 2z+1 2 0 2 1 2z 2z+1

. 0 1 z 2z+1 2z+2 2 2z z+2 z+1

0 0 0 0 0 0 0 0 0 0
1 0 1 z 2z+1 2z+2 2 2z z+2 z+1
z 0 z 2z+1 2z+2 2 2z z+2 z+1 1
2z+1 0 2z+1 2z+2 2 2z z+2 z+1 z+1 Z
2z+2 0 2z+2 2 2z z+2 z+1 1 z 2z+1
2 0 2 2z z+2 z+1 1 z 2z+1 2z+2
2z 0 2z z+2 z+1 1 z 2z+1 2z+2 2
Z+2 0 z+2 Z+1 1 z 2z+1 2z+2 z 2z
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Z+1 0 z+1 1 z 2z+1 2z+2 2 2z z+2

Q5.2 (i) Using the minimal polynomials as calculated previously in Table-1 in the
previous solution:

GF(3)
x8 – 1 = (x + 1) (x + 2) (x2 + 1) (x2 + 2x + 2) (x2 + x + 2)

Ternary BCH code with t = 2 (double error correcting code).

g(x) = LCM [f1(x), f2(x), f3(x) f4(x) ]
= LCM [x2 + x + 2, x2 + 1, x2 + x + 2, x + 1]
= (x2 + 2x + 2) (x2+1) (x+2)
g(x) = x5 + 2x4 + x3 + x2 + 2

Minimum distance d* = 5
n = 8, n – k = 5 ⇒ k = 3.
Code rate = k/n = 3/8.

(ii) The primitive blocklength = 26 = qm – 1 = 33 – 1.

p(x) = x3 + 2x + 1
Representation of GF(32)

Exponential Polynomial Ternary Decimal Minimal

Notation Notation Notation Notation Notation
0 0 000 0
α0 1 001 1 x+1
α1 2 010 3 x3 + 2x +1
α2 22 100 9 x3 + 2x2 + 2x +1
α3 2+z 012 5 x3 + 2x + 2
α4 22 + 2z 120 15 x3 + 2x2 +1
α5 2z2 + z + 2 212 23 x3 + x2 + x + 2
α6 22 + 1 + 2 111 13 x3 + 2x2 + x + 1
α7 22 + 2z + 2 122 17
α8 2z2 + 2 202 29
α9 z+1 011 4 x3 + 2x + 2
α10 22 + z 110 12 x3 + 2x + 1
α11 22 + z + 2 112 14
α12 22 + z 102 11 x3 + 2x2 + 1
α13 2 002 2 x+2
α14 2z 020 6
α15 2z2 200 18 x3 + x2 + x + 2
α16 2z + 1 021 7
α17 2z2 + 2 210 21
α18 22 + 2z +1 121 16 x3 + 2x2 + x + 1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

α19 2z2 + 2 + 2z 202 26 x3 + x2 + x + 2

α20 2z2 + z + 1 211 22
α21 22 + 1 101 10
α22 2z + 2 022 8
α23 2z2 + 2z 220 24
α24 2z2 + 2z + 1 221 25
α25 2z2 + 1 201 19

t = 3 (triple error correcting code)

g(x) = LCM [f1(x), f2(x), f3(x), f4(x) f5(x) f6(x)]
= x12 + x11 + 2x9 + 2x8 + 2x7 + 2x3 +2x2 + x + 1
d* = 9 (thus the code is over designed).

Q5.3 Representation of GF(25)

Primitive polynomial p(x) = x5 + x2 + 1

Table II
Exponential Polynomial Notation Binary Decimal Minimal
Notation Notation Notation Notation
0 0 00000 0
0
α 1 00001 1 x+1
α1 z 00100 2 f1 (x)
α 2
z2 01000 4 f1 (x)
α3 z3 10000 8 f3 (x)
α4 z4 00101 16 f1 (x)
α5 z2 + 1 01010 5 f5 (x)
α6 z2 + z 10100 10 f5 (x)
α7 z4 + z2 01101 20 f7 (x)
α8 z3 + z2 + 1 11010 13 f1 (x)
α9 z4 + z3 + z 10001 26 f5 (x)
α10 z4 + 1 10001 17 f5 (x)
α11 z2 + z + 1 00111 7 f11 (x)
α12 z3 + z2 + z 01110 14 f3 (x)
α13 z4 + z 3 + z2 + 1 11100 28 f11 (x)
α14 z4 + z 3 + z2 + 1 11101 29 f7 (x)
α15 z4 + z 3 + z2 + z + 1 11111 31 f15 (x)
α16 z4 + z3 + 1 + z 11011 27 f1 (x)
α17 z4 + z + 1 10001 19 f3 (x)
α18 z +1 00011 3 f5 (x)
α 19
z2 + z 00110 6 f7 (x)
α20 z3 + z 01100 12 f3 (x)
α21 z4 + z2 11000 24 f11 (x)
α22 z4 + z2 + 1 10101 21 f11 (x)
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

α23 z3 + z2 + z + 1 01111 15 f15 (x)

α24 z4 + z 3 + z2 + z 11110 30 f3 (x)
α25 z4 + z3 +1 11001 25 f7 (x)
α26 z4 + z2 + z + 1 10111 23 f11 (x)
α27 z3 + z + 1 010111 11 f15 (x)
α28 24 + z2 + 2 101110 22 f7 (x)
α29 23 + 1 01001 9 f15 (x)
α30 24 + 2 10010 18 f15 (x)

The generator polynomial for correcting at least t errors is given by

g(x) = LCM [f1(x) f2(x), ..., f2t(x)].
g(x) = x5 +(z+1) x4 + (z2 + z + 1) x3 + (z + 1) x2 + zx +z4

Q5.4 Over GF(2) for RS code N=21 – 1=1

Generator polynomial (x – α0) = (x – 1)

Q5.5 Taking values of 2’s from GF(2n) construction

(i) RS (15, 11) code
n = 15, k = 11 d* = n – k + 1 = 5
t = (n – k)/2 = 2.
g(x) = (x – 2) (x – 22) (x – 23) (x – 24)
= x4 + (23 + 22 + 1)x3 + (23 + 22)x2 + z3(x) + (z2 + 2 + 1)
= x4 + α13 x3 + x6 x2 + α3 x + α10

(ii) RS (15, 7) n = 15, k = 7

d* = 15 – 7 + 1 = 9
t = (n – k)/2 = 4

g(x) = (x - α) (x - α2) (x - α3) (x - α4) (x - α5) (x - α6) (x - α7) (x - α2) (x - α8)

(iii) Taking values of α’s from Table II RS (31, 21)

n = 31, k = 21, d* = 11, t = (n – k)/2 = 5
g(x) = (x - α) (x - α2) (x - α3) …………… (x - α9) (x - α10)

Q5.6 Hint: Similar to example 5.10.

Q5.7 Hint: Start with a RS code over GF(qm) for a designed distance d. Then show that
BCH code is a subfield subcode of a Reed Solomon code of the same designed distance.

Q5.8 The theorem is true if µ = 1. If it is true for all µ - 1 by µ - 1 Vendermonde

matrices, then it is also true for all µ by µ Vendermonde matrices. Replace X1 by the
indeterminate x and transpose the matrix. Then the determinant is a function of x, given
by
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡1 x x2 L x µ −1 ⎤
⎢ ⎥
1 X2 X 22 L X 2µ −1 ⎥
D( x) = det ⎢
⎢M M M ⎥
⎢ 2 µ −1 ⎥
⎢⎣ 1 X µ X µ L X µ ⎥⎦
The determinant can be expanded in terms of elements of the first row multiplied by the
cofactors of these elements of the first row. This gives a polynomial in x of degree µ - 1,
which can be written
D ( x) = d µ −1 x µ −1 + L + d 1 x + d 0
The polynomial D(x) has at most µ - 1 zeros. The coefficient d µ −1 is itself the
determinant of a Vandermonde matrix, and by the induction hypothesis, is nonzero. If for
any i, 2 ≤ i ≤ µ, we set x = Xi, then two rows of the matrix are equal, and D(Xi) = 0. Thus
for each i ≠ 1, Xi is a zero of D(Xi), and because they are all distinct and there are µ - 1 of
them, the polynomial can easily factored:

⎡ µ ⎤
D( x) = d µ −1 ⎢∏ ( x − X i )⎥.
⎣ i =2 ⎦
Therefore the determinant of the original Vandermonde matrix is
⎡ µ ⎤
D( X 1 ) = d µ −1 ⎢∏ ( X 1 − X i )⎥.
⎣ i=2 ⎦
This is non zero because d µ −1 is nonzero and X 1 is deferent from each of the remaining
X i . Hence the determinant of the µ by µ Vendermonde matrix is nonzero, and by
induction, the theorem is true for all µ.

⎡1 1 1 1 ⎤
⎢1 2 3 10 ⎥⎥
⎢
⎢1 2 2 32 10 2 ⎥
Q5.9 (i) Given: H = ⎢ 3 ⎥.
⎢1 2 33 10 3 ⎥
⎢ ⎥
⎢ 8 8⎥
⎣⎢1 2 38 10 ⎦⎥
⎡ 10 ⎤
We observe that sr = ⎢∑ i r ⎥ = 0 for r = 0, 1, 2, 3, ..,8.
⎣ i =1 ⎦ mod11
Thus all the 10 columns of H add up to the zero vector. Hence d* = 10 ⇒ t = 4.

(ii) The size of H is (n-k) x n. Thus n – k = 9, n = 10 ⇒ k = 1.

In this case, n – k + 1 = 9 + 1 = 10 = d*.
Thus, it is optimal in the sense of the Singleton bound.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡1 1 1 1 ... 1 ⎤
⎢1 2 3 4 ... 10 ⎥
⎢ ⎥
⎢1 2 2 32 4 2 ... 10 2 ⎥
Q5.10 (i) Given H = ⎢ 3 3 3 3⎥
⎢1 2 3 4 ... 10 ⎥
⎢1 2 4 34 4 4 ... 10 4 ⎥
⎢ 5 5 5 5⎥
⎣⎢1 2 3 4 ... 10 ⎦⎥
Use the same approach as the previous question.
(ii) First convert the matrix into H = [ – PT | I ] form. Since the math is over GF(11), and
q = 11 = odd, its simply modulo 11 arithmetic. The generator matrix is given by
G = [ I | P].

Q5.11 (i) x8 + x4 +x3 +x2 +1

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 6

Q6.1 R = 1/2 , v = 4, d* = 6

(i) There are 24 = 16 states for this encoder. The state transitions and the output is given
in the table below.

Input Current State Next State Output

0 0 0 0 0 0 0 0 0 0 0
1 0 0 0 0 1 0 0 0 1 1
0 0 0 0 1 0 0 0 0 0 0
1 0 0 0 1 1 0 0 0 1 1
0 0 0 1 0 0 0 0 1 0 1
1 0 0 1 0 1 0 0 1 1 0
0 0 0 1 1 0 0 0 1 0 1
1 0 0 1 1 1 0 0 1 1 0
0 0 1 0 0 0 0 1 0 0 1
1 0 1 0 0 1 0 1 0 1 0
0 0 1 0 1 0 0 1 0 0 1
1 0 1 0 1 1 0 1 0 1 0
0 0 1 1 0 0 0 1 1 0 0
1 0 1 1 0 1 0 1 1 1 1
0 0 1 1 1 0 0 1 1 0 0
1 0 1 1 1 1 0 1 1 1 1
0 1 0 0 0 0 1 0 0 0 1
1 1 0 0 0 1 1 0 0 1 0
0 1 0 0 1 0 1 0 0 0 1
1 1 0 0 1 1 1 0 0 1 0
0 1 0 1 0 0 1 0 1 0 0
1 1 0 1 0 1 1 0 1 1 1
0 1 0 1 1 0 1 0 1 0 0
1 1 0 1 1 1 1 0 1 1 1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

0 1 1 0 0 0 1 1 0 0 0
1 1 1 0 0 1 1 1 0 1 1
0 1 1 0 1 0 1 1 0 0 0
1 1 1 0 1 1 1 1 0 1 1
0 1 1 1 0 0 1 1 1 0 1
1 1 1 1 0 1 1 1 1 1 0
0 1 1 1 1 0 1 1 1 0 1
1 1 1 1 1 1 1 1 1 1 0

(iii) dfree = 6.

(iv) G(D) = [ 1 D4+D3+D2+D+1].

(v) Non catastrophic since GCD [1 D4+D3+D2+D+1] = 1.

Q 6.2 (i)
+

000 000 000 000

111
110
110

010

(ii) dfree = 8 (see the trellis diagram).

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q6.3 (i)
+
+

+ +

Trellis diagram will have 16 states. Generate mechanically, or write a small

computer program. The state transitions in a tabular for is given below.

Present state Next state Output

i/p = 0 i/p = 1 i/p = 0 i/p = 1
0000 0000 1000 000 001
0001 0000 1000 000 001
0010 0001 1001 110 111
0011 0001 1001 110 111
0100 0010 1010 010 011
0101 0010 1010 010 011
0110 0011 1011 100 101
0111 0011 1011 100 101
1000 0100 1100 001 000
1001 0100 1100 001 000
1010 0101 1101 111 110
1011 0101 1101 111 110
1100 0110 1110 011 010
1101 0110 1110 011 010
1110 0111 1111 101 100
1111 0111 1111 101 100

(ii) k0 = 1, n0 = 3, Rate, R = 1/3

ν = 4, m = 1.

(iii) d* = 5, dfree = 5
(iv) G(D) = [1 + D D2 + D3 D2 + D3 + D4]

Q6.4 (i) k = 3, n = 4, v = 1 m = 4 + 2 + 2 = 8, R = ¾.

[ ]
k0
v = ∑ max deg g ij ( D) = sum of highest power of gij(D) in each row of the
i =1 j

matrix G(D) = 2 + 2 + 4 = 8. The G(D) is given below.

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡ D + D2 D2 D + D2 0 ⎤
⎢ ⎥
(ii) G ( D) = ⎢ D 2 D D D2 ⎥
⎢ 0 0 D4 D 2 ⎥⎦
⎣ 3x4

⎡0 0 0 0⎤
G 0 = ⎢⎢ 0 0 0 0 ⎥⎥
⎢⎣ 0 0 0 0 ⎥⎦

⎡1 0 1 0⎤ ⎡1 1 1 0⎤
G1 = ⎢⎢ 0 1 1 0 ⎥⎥ G2 = ⎢⎢ 1 0 0 1 ⎥⎥
⎢⎣ 0 0 0 0 ⎥⎦ ⎢⎣ 0 0 0 1 ⎥⎦

⎡0 0 0 0⎤ ⎡0 0 0 0⎤
G3 = ⎢⎢ 0 0 0 0 ⎥⎥ G4 = ⎢⎢ 0 0 0 0 ⎥⎥
⎢⎣ 0 0 0 0 ⎥⎦ ⎢⎣ 0 0 1 0 ⎥⎦

⎡ D 0 1 D2 D + D2 ⎤
⎢ ⎥
Q6.5 G ( D) = ⎢ D 2 0 0 1+ D 0 ⎥
⎢ 1 0 D2 0 D 2 ⎥⎦
⎣

k0 = 3, n0 = 5, R = 3/5

(ii) GCD [G(D)] = 1

⇒ It’s a non-catastrophic convolutional code.

Q6.6 Hint:

Use:
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

⎡ P0T −I ...⎤
⎢ T ⎥
⎢ P1 0 P 0
T
−I ⎥
⎢ P2T 0 P T
0 P0T −I ⎥
⎢ ⎥
1

H =⎢ : : : : ⎥ , and
⎢PT 0 T
P 0 PmT− 2 0 ... P0T −I ...⎥
⎢ m m −1
⎥
⎢ ⎥
T T
Pm 0 P m −1 0 ...
⎢ T
...⎥⎦
⎣ P m 0 ...

⎡I P0 0 P1 0 P2 ... 0 Pm 0 0 0 0 ...⎤
⎢0 0 I P0 0 P1 ... 0 Pm −1 0 Pm 0 0 ...⎥⎥
⎢
⎢0 0 0 0 I P0 ... 0 Pm − 2 0 Pm −1 0 Pm ...⎥
G=⎢ ⎥,
⎢: : : : : : : : 0 Pm − 2 0 Pm −1 ...⎥
⎢ : : 0 Pm − 2 ...⎥
⎢ ⎥
⎢⎣ : : ⎥⎦

where Gl = [gijl] and g ij ( D) = ∑ g ijl D l .

l

Q6.7 (i) g1(D) = 2D3 + 3D2 + 1 over GF(4)

g2(D) = D3 + D + 1

This is a rate ½ encoder.

Hint: You can break up the problem by representing each symbol ∈{0, 1, 2, 3} by two
bits. Thus, this encoder takes in one quaternary symbol (2 bits) and converts it to 2
quaternary symbols (4 bits). The memory unit of this encoder consists of 3 delay
elements. Thus the number of states = 44 = 256. In reality, such an encoder will be
implemented using a 8-bit shift register, with two bit-shifts at a time. The logic circuit
will be implemented using GF(4) arithmetic. It is advisable to write a small computer
program to calculate the dmin for this code.

(ii) GCD [g1, g2 ] = 1 ⇒ non catastrophic

Q6.8 g1 = D3 + D2 + 1
g2 = D3 + D
g3 = D2 + 1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Input Present state Next state Output

0 000 000 000
1 000 100 101
1 100 110 111
0 110 011 110
0 011 001 011
0 001 000 111
1 000 100 101
0 100 010 111
1 010 101 011
1 001 110 100
0 110 101 110
1 011 101 110
0 101 010 101
1 010 101 001

(i) Encoded Stream for input = 01100011110101 :

000 101 111 110 011 111 101 111
011 100 001 110 101 001

(ii) Encoded Stream for input = 1010101010 .....

101 010 001 101 001 101 001 101

(iii) 0011011

Q6.9 Hint: Solve the problem similar to problem 6.7. You can break up the problem by
representing each symbol ∈{0, 1, 2} by two bits. Thus, this encoder takes in one ternary
symbol (2 bits) and converts it to 2 ternary symbols (4 bits). The memory unit of this
encoder consists of 4 delay elements. Thus the number of states = 34 = 81. Write a small
computer program to calculate the dmin for this code.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 7

Q7.1 Given 2/3 convolutional encoder with

1 D D + D2
G(D)=
D2 1+ D 1+ D + D2

(a) the encoder has 16 states in the Trellis diagram

(b) the free Euclidean distance:

QPSK 8-PSK
s1 s2
s3 s1
δ0 δ2 δ1
δ0
s2 δ1 s0 s4 δ3 s0
Es Es

s7

δ 02 = 2Es δ 02 = 0.586Es
δ12 = 4Es δ12 = 2Es
δ 22 = 3.414Es
δ 32 = 4Es

Gray coded using 8 psk signal constellation

d 2free = 2Es + 2Es +.586 Es = 4.586 Es.

dfree = 2.1415 √E

(c) The asymptotic coding gain w.r.t. QPSK

( )
d 2free / Es coded
g ∞ = g SNR →∞ = 10 log 2
( )
d free / Es uncoded
,

= 10 log (4.586/2)
= 3.6040 dB.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q7.2

(a) The free Euclidean distance:

d 2free = 2Es + 2Es + 2Es
dfree = 2.4495 √E.

(d 2
free )
/ Es coded
(b) g ∞ = g SNR →∞ = 10 log
(d 2
free /E )
s uncoded
,

= 10 log (6/2)
= 4.7712 dB.

Q7.3
(i)
Present State Input Next State Output
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 1 0 0 0 0 1 0
0 0 0 1 0 1 1 0 0 0 0
0 0 0 1 1 1 1 0 0 1 0
0 0 1 0 0 0 1 0 1 0 1
0 0 1 0 1 0 1 0 1 1 1
0 0 1 1 0 1 0 0 1 0 1
0 0 1 1 1 1 0 0 1 1 1
0 1 0 0 0 0 0 1 0 0 0
0 1 0 0 1 0 0 1 0 1 0
0 1 0 1 0 1 1 1 0 0 0
0 1 0 1 1 1 1 1 0 1 0
0 1 1 0 0 0 1 1 1 0 1
0 1 1 0 1 0 1 1 1 1 1
0 1 1 1 0 1 0 1 1 0 1
0 1 1 1 1 1 0 1 1 1 1
1 0 0 0 0 0 0 1 1 0 0
1 0 0 0 1 0 0 1 1 1 0
1 0 0 1 0 1 1 1 1 0 0
1 0 0 1 1 1 1 1 1 1 0
1 0 1 0 0 0 1 1 0 0 1
1 0 1 0 1 0 1 1 0 1 1
1 0 1 1 0 1 0 1 0 0 1
1 0 1 1 1 1 0 1 0 1 1
1 1 0 0 0 0 0 0 1 0 0
1 1 0 0 1 0 0 0 1 1 0
1 1 0 1 0 1 1 0 1 0 0
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 3
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

1 1 0 1 1 1 1 0 1 1 0
1 1 1 0 0 0 1 0 0 0 1
1 1 1 0 1 0 1 0 0 1 1
1 1 1 1 0 1 0 0 0 0 1
1 1 1 1 1 1 0 0 0 1 1

(ii) There are 8 states in the trellis diagram. Construct it using the table above. The
trellis diagram has parallel paths. In succinct form, the trellis diagram can be written as

S0,S2,S0,S2

S4,S6,S4,S6

S7,S5,S7,S5

S0,S2,S0,S2

S1,S3,S1,S3

S6,S4,S6,S4

S5,S7,S5,S7

S3,S1,S3,S1

(iii) d 2free = 2Es

N( d 2free ) = 2 x 8 = 16.

(iv) After set partitioning: d 2free = 4Es

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 4
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

S0,S4,S0,S4

S2,S6,S2,S6

S1,S5,S1,S5

S0,S4,S0,S4

S7,S3,S7,S3

S6,S2,S6,S2

S3,S7,S3,S7

S3,S1,S3,S1

(v) Encoded symbols for

Natural Mapping: S0, S4, S0, S2, S2

Mapping using set partitioning: S2, S0, S0, S4, S4 .

(vi) Asymptotic coding gain for

Natural Mapping: 10 log (2 / 0.586) = 5.33 dB

Mapping using set partitioning: 10 log (4 / 0.586) = 8.34 dB.

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 5
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q7.4(a) Set partitioning of asymmetric constellation diagram:

∆0 = δ0

A0 A1
∆1 = δ1

A00 A01 A10 A11

∆2 =

(b) If the minimum Euclidean distance between parallel transitions is given by ∆m~ +1 and
the minimum Euclidean distance between non-parallel paths of the trellis is given by
~ ), the free Euclidean distance of the TCM encoder is
dfree( m

~ )].
dfree = min[ ∆m~ +1 , dfree( m

If all the symbols are used with equal frequency, the asymmetric constellation diagram
will result in a smaller value of dfree in general (i.e., poorer performance).

Q7.5 Hint: Perform set partitioning as follows;

"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 6
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q7.6 Hint: For the right hand side of the inequality, divide both the numerator and
denominator by K and take lim K → ∞.

Q7.7 Hint: Use equation 7.55.

Q7.8
a) The number of states depends on the desired BER and the computational
complexity permissible.
b) The internal memory of the encoder will be decided by the number of states in the
trellis.
c) The decision regarding parallel paths can be taken upon calculation of the dfree.
Parallel paths cannot be ruled out right in the beginning. The number of branches
emanating from each node will depend on the constellation size of the modulation
scheme.
d) The decision is between MPSK and MQAM. Beyond M = 16 it is better to use
MQAM (see their BERs in AWGN!).
e) Since AWGN is under consideration, we will use the Ungerboek’s design rules.

Q7.9
Viterbi decoding metric is m(rl, sl) = ln p(rl/sl). The metric of the above form is chosen
because it behaves like a distance measure between the received signal and the signal
associated with the corresponding branch in the trellis. It has an additive property,
namely, that the total metric for a sequence of symbols is the sum of the metric for each
channel input and output pair.

Q7.10.
(a) The direct application of equation 7.55 gives the new d p2 ( L) equal to ten times
the previous d p2 ( L) . However, to obtain such a large value of d p2 ( L) , we need to
construct trellis with a larger number of states.
(b) The new d 2free will be larger resulting in an improvement in the performance of
the new TCM scheme over AWGN channels as well.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 7
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

Q7.11
(a) The trellis diagram of the encoder:
00 00 00
11 11 11
01 01 01
10 10 10

00 00 00

10 11 10 11 10
11
10 10 10

S3

(b) T ( D) =
(
D3 1 − D + D3 )
(
1 − 2D + D 2 + D 4
.
) D D2

D 1 D2

S0 S2 S1 S0
(c)
D
DLI

S3

DLI D2L

DLI L D2L

S0 S2 S1 S0

DLI
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 8
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

L3 ID 3 (1 − LID ) + L4 I 2 D 6
T ( D , L, I ) =
( )
1 − LID − L2 ID + L3 I 2 D 2 + L3 I 2 D 4
.

(d) d Hfree = 3.

(e) Only one path (see the trellis diagram).

Q7.12: Hint: Carry out the mathematical excercise along the lines in Section 7.6.

Q7.13 (i) TCM2

(iii) For AWGN Channel: TCM2

For Fading Channel: TCM1
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 1
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

SOLUTIONS FOR CHAPTER 8

Q8.1 (a) Let M be the size of alphabet. Then to crack this code we need ‘M-1
attempts’ by using brute force attack. In this case we require 26 – 1 = 25
attempts.

(b) 25 ms

Q8.2 THE CAT IS OUT OF THE BOX

Q8.3. (a) Assuming plain text is staggered between k rows, we apply brute force
attack for each ‘k’, 1<k<n , where ‘n’ is the length of the cipher text.
For each ‘k’, divide cipher text into ⎣n / k ⎦ sized blocks (total k blocks).
Number of attacks = (n - 2).
(b) Given k, apply the attack corresponding to k as given in part a) to decrypt.

Q8.4. First stage:

Size of the key < 36 (36 is worst case).
Number of possible rectangular matrices = 36! (Factorial 36).
Second stage:
Key size k < 36 again
We need to attack for key sides from 1 to 36.
⎡ 36 ⎤
So total number of attacks = 36! ⎢∑ k!⎥ .
⎣ k =1 ⎦

(b) Given the cipher text, use the second key to form the second matrix (write
horizontally, read vertically). After reading horizontally, we get intermediate
cipher text. Next, use the first key to form the rectangular matrix. For finding
plain text corresponding to XY location, look at X-row and Y-column.

Q8.5. (a) Number of attacks needed < number of vectors = 2n

Q8.6 27!

Q8.7
P = 29, Q = 61
N = PQ = 29 * 61 = 1769
Q(N) = (P-1) (Q-1) = 28 * 60 =1680

Let E = 11 , as 11 * 611 – 4 * 1680 =1

Take D = 611
11 is the public key
611 is the private key.
"Copyrighted Material" -"Additional resource material supplied with the book Information, Theory, Coding and 2
Cryptography, Second Edition written by Ranjan Bose & published by McGraw-Hill Education (India) Pvt. Ltd. This
resource material is for Instructor's use only."

(b) R = 82, S = 83 , A = 65

822 = 1417 mod 1769

824 = 74 mod 1769
828 = 169 mod 1769
823 = 1209 mod 1769
8211 = 886 mod 1769

832 = 1582 mod 1769

834 = 1358 mod 1769
838 = 166 mod 1769
833 = 1118 mod 1769
8311 = 545 mod 1769

652 = 687 mod 1769

654 = 1415 mod 1769
658 = 1480 mod 1769
6511 = 371 mod 1769

Therefore RSA is encoded as {886, 545, 371}.

(c) P= 37, Q = 67
N=PQ =37 *67 = 2479
Q(N) = (P-1) (Q-1) = 36 * 66 = 2376
E=5
D = 1901
Public key = 5
Private key = 1901

Q8.8 Hint: Find prime numbers below 1090 and 10100 and subtract from the
later.

Q8.9
(i) Each pair should have different keys so total number of distinct keys
⎛ N ⎞ N ( N − 1)
= N C2 = ⎜⎜ ⎟⎟ = .
⎝2⎠ 2
(ii) There is 1 public key and N private keys.

Q8.10 1/3

Q8.11 (i) Total number of subblocks = n = p2 / q2

It requires n! brute force attacks.

(ii) Key size = log2n!