Professional Documents
Culture Documents
Outline of Presentation: Modern Algebra C Research Problems Conclusion References
Outline of Presentation: Modern Algebra C Research Problems Conclusion References
Introduction
Modern Algebra
Code Construction
Research Problems
Conclusion
References
1 3/8/2021
Introduction…
General Block Diagram of Communication System
Information Source Channel Modulator
Source Encoder Encoder (writing unit)
Channel
Noise (Storage
Medium)
Shannon’s central theme was that , if the signaling rate of the system is
less than the channel capacity, reliable communication can be achieved
if one chooses proper encoding and decoding technique.
There has been increasing demand for efficient and reliable digital data
transmission and storage system.
3 3/8/2021
Introduction…
The design of good codes and efficient decoding method,
initiated by Hamming, Slepian and other in the early 1950’s,
4 3/8/2021
Introduction…
Linear Block code (n , k), where
n- codeword length ,
Hamming Codes
Cyclic code
BCH codes,
R-S codes,
Turbo codes
7 3/8/2021
Introduction…
Forward Error Correcting (FEC) codes includes
Fire codes
Concatenated codes
Interleaved codes
Product Codes
Burton Code
R. M. Code
8 3/8/2021
Introduction…
Applications of codes
Practical application of Error Correcting Codes are for
High data rate applications
Satellite-based DVB
Optical Communication
Fields
Vector Spaces
10 3/8/2021
Modern Algebra
Groups
Let G be a set of elements.
A binary operation * on G is rule that assigns to each pair of
elements a and b a uniquely defined third element c = a* b in G.
When such a binary operation * is defined on G, we say that G is
closed under *.
For example, let G be the set of all integers & let the binary
operations on G be real addition +.
We all know that for any two integers i & j is a uniquely defined
integer in G.
Hence, the set of integer is closed under real addition.
11 3/8/2021
Modern Algebra
Groups:
A binary operation * on G is said to be associative if, for any a ,b, & c in
G.
𝑎 ∗ 𝑏 ∗ 𝑐 = 𝑎 ∗ 𝑏 ∗ 𝑐.
G contains an element e such that for any a in G, a ∗ e = e ∗ a = a This
element e is called an identity element.
For any element a in G there exist another element a’ in G such that
𝑎 ∗ 𝑎′ = 𝑎′ ∗ 𝑎 = 𝑒. The element a’ is called an inverse of a(a is also an
inverse of a’).
A group G is said to be commutative if its binary operation * also satisfies
the following conditions :
For any a & b in G,
𝑎 ∗ 𝑏 = 𝑏 ∗ 𝑎.
12 3/8/2021
Modern Algebra…
Groups :
Example
The set of all integer is a commutative group under real addition.
In this case, the integer 0 is the identity element, & the integer –i is the
inverse of integer i.
The set of all rational numbers excluding zero is a commutative group under
real multiplication.
The integer 1 is the identity element with respect to real multiplication, & the
rational number 𝑏 𝑎 is the multiplicative inverse of 𝑎 𝑏.
Groups with finite numbers of elements do exist, as we shall in the next
example.
The number of elements in a group is called as order of the group.
A group of finite order is called a finite group.
13 3/8/2021
Modern Algebra…
Group:
Example :
Consider the set of two integers 𝐺 = 0,1 . Let us define a binary
operation, denoted by ⊕, on G as follows
Solution
0 ⊕ 0 = 0. 0⊕1 = 1 1⊕0 = 1 1⊕1 = 0
15 3/8/2021
Modern Algebra…
Fields:
It follows from the definition that a field consist of at least two
elements, the additive and multiplicative identity.
For Example:
16 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Let V be a set of elements on which a binary operation called
addition, +, is defined. Let F be a field.
The set V is called a vector space over the field F if it satisfies the
following conditions:
V is Commutative under addition.
17 3/8/2021
Modern Algebra…
Vector Spaces Vn :
(Distributive Law) For any element u and v in V and any element a and b in
F,
𝑎 . 𝒖 + 𝒗 = 𝑎 . 𝒖 + 𝑎 . 𝒗,
𝑎 + 𝑏 . 𝒗 = 𝑎. 𝒗 + 𝑏 . 𝒗.
(Associative Law) For any v in V and any a and b in F.
𝑎 . 𝑏 . 𝒗 = 𝑎 . (𝑏 . 𝒗) .
Let 1 be the unit element of F. Then, for any v in V, 1 . 𝒗 = 𝒗.
The elements of V are called as vectors & elements of field F are called as
scalars.
The additions on V is called a vector addition.
The multiplication that combines a scalar in F and vector in V is referred to
as scalar multiplication (or product ).
The additive identity of V is denoted by 0.
18 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Let n =5. the vector space V5 of all 5-tuples over GF(2) consist
of the following set of 32 vectors which are distinct :
00000 00001 00010 00011
00100 00101 00110 00111
01000 01001 01010 01011
01100 01101 01110 01111
10000 10001 10010 10011
10100 10101 10110 10111
11000 11001 11010 11011
11100 11101 11110 11111
Consider the vector space of all 5 tuple over GF(2) the set
00000 , 00111 , 11010 , 11101
21 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Linear Combination of vectors
Let 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑘 be k vectors in vector space V over a field F, let 𝑎1 , 𝑎2 , ⋯ , 𝑎𝑘 be k
scalars from F. The sum of product of scalar and vector that is
𝑎1 𝑣1 + 𝑎2 𝑣2 + ⋯ + 𝑎𝑘 𝑣𝑘
Clearly, the sum of two linear combinations of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑘 .
𝑎1 𝑣1 + 𝑎2 𝑣2 + ⋯ + 𝑎𝑘 𝑣𝑘 + 𝑏1 𝑣1 + 𝑏2 𝑣2 + ⋯ + 𝑏𝑘 𝑣𝑘
= 𝑎1 + 𝑏1 𝑣1 + 𝑎2 + 𝑏2 𝑣2 + ⋯ + 𝑎𝑘 + 𝑏𝑘 𝑣𝑘
Scalar Product :
The product of scalar c in F & a linear combination of 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑘 .
𝑐 . 𝑎1 𝑣1 + 𝑎2 𝑣2 + ⋯ + 𝑎𝑘 𝑣𝑘 = 𝑐. 𝑎1 𝑣1 + 𝑐. 𝑎2 𝑣2 + ⋯ + 𝑐. 𝑎𝑘 𝑣𝑘
22 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Statement:
Proof:
24 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Example:
The vectors (10110) , (01001), & (11111) are linearly dependent. Since
1 . 10110 + 1. 01001 + 1. 11111 = (00000);
A set of vectors is said to span a vector space V if every vector in V is a linear combination of
vectors in set .
25 3/8/2021
Modern Algebra…
Vector Spaces Vn :
Basis Vector
In any vector space or subspace there exist at least one set B of
linearly independent vectors that span the space.
29 3/8/2021
Modern Algebra…
Irreducible polynomial P(X):
Among the four polynomials of degree 2, 𝑋 2 , 𝑋 2 +1, & 𝑋 2 + 𝑋
are not irreducible, since they are divisible by X or X + 1;
31 3/8/2021
Modern Algebra
Extension field GF(2m): (Where m is the degree of
irreducible polynomial)
Let m = 4, the polynomial 𝑝 𝑋 = 1 + 𝑋 + 𝑋 4 is a primitive
polynomial over GF(2), set 𝑝 𝛼 = 1 + 𝛼 + 𝛼 4 = 0.
α is the primitive element, exist in the extension field GF(2m) of
GF(2).
Then ∝4 = 1+ ∝, using this relation, we can construct 𝐺𝐹 24 .
The identity ∝4 = 1 +∝ is used repeatedly to form the
polynomial representation for the element of 𝐺𝐹 24 .
𝛼 5 = 𝛼. 𝛼 4 = 𝛼 1 + 𝛼 = 𝛼 + 𝛼 2
𝛼 6 = 𝛼. 𝛼 5 = 𝛼 𝛼 + 𝛼 2 = 𝛼 2 + 𝛼 3
𝛼 6 = 𝛼. 𝛼 6 = 𝛼 𝛼 2 + 𝛼 3 = 𝛼 3 + 𝛼 4 = 𝛼 3 + 1 + 𝛼 = 1 + 𝛼 + 𝛼 3
32 3/8/2021
Modern Algebra
𝛼 5 + 𝛼 7 = 𝛼 + 𝛼 2 + 1 + 𝛼 + 𝛼 3 = 1 + 𝛼 2 + 𝛼 3 = 𝛼13 ,
1 + 𝛼 5 + 𝛼10 = 1 + 𝛼 + 𝛼 2 + 1 + 𝛼 + 𝛼 2 = 0.
33 3/8/2021
Modern Algebra
Extension field GF(2m): (Where m is the degree of irreducible
polynomial)
Table 1 : All 16 elements of GF(16) i.e. GF(24) are given in power of α, polynomial form
of α & 4 tuple form (binary form) is given
34 3/8/2021
Modern Algebra …
Factorization of Xn + 1 over GF(2), where n =𝟐𝒎 -1
Let 𝑓 𝑥 be a polynomial with coefficients from GF(2). Let 𝛽 be an
element in extension field GF 2𝑚 .
36 3/8/2021
Modern Algebra …
Factorization of Xn + 1 over GF(2), where n =𝟐𝒎 -1
22
Consider the element 𝛼 5 𝑖𝑛 𝐺𝐹 16 . Since the 𝛼 5 = 𝛼 20 = 𝛼 5 ,
The only conjugate of 𝛼 5 is 𝛼 10 , Both 𝛼 5 & 𝛼 10 both have order n =3 . The
minimal polynomial of 𝛼 5 & 𝛼 10 is 𝑋 2 + 𝑋 + 1 . Whose degree is factor of
m = 4.
The conjugates for α7 are α11 , α13 ,α14 , The minimal polynomial for α7 is:
𝑋 + 𝛼 7 𝑋 + 𝛼 11 𝑋 + 𝛼 13 𝑋 + 𝛼 14
= 𝑋 2 + 𝛼 7 + 𝛼 11 𝑋 + 𝛼 18 𝑋 2 + 𝛼 13 + 𝛼 14 𝑋 + 𝛼 27
= 𝑋 2 + 𝛼 8 𝑋 + 𝛼 3 𝑋 2 + 𝛼 2 𝑋 + 𝛼 12
= 𝑋 4 + 𝛼 8 + 𝛼 2 𝑋 3 + 𝛼 12 + 𝛼 10 + 𝛼 3 𝑋 2 + 𝛼 20 + 𝛼 5 X + 𝛼 15
= 𝑋4 + 𝑋3 + 1
37 3/8/2021
Modern Algebra…
Factorization of Xn + 1 over GF(2), where n =𝟐𝒎 -1
The conjugecy class of each element from GF(16) and corresponding
polynomial are given in the table below.
There fore the polynomial corresponding to elements of GF(16) except
zero is as follows.
𝑋 15 + 1 = 𝑋 + 1 𝑋 4 + 𝑋 + 1 𝑋 4 + 𝑋 3 + 𝑋 2 + 𝑋 + 1 𝑋 2 + 𝑋 + 1 𝑋 4 + 𝑋 3 + 1
k- message length.
𝑋𝑛 + 1 = 𝑔 𝑋 . 𝑋 .
Theorem 1:
The generator polynomial g(X) of an (n , k) cyclic code is a
factor of X n + 1, where n =2m – 1.
Theorem 2:
If g(X) is a polynomial of degree (n-k) and is a factor of X n + 1,
then g(X) generates an (n , k) cyclic code where n =2𝑚 -1
39 3/8/2021
Construction of Codes
Linear Block Code (n , k):
Information in most current digital data communication and storage
systems is coded in the binary digits 0 & 1, we discuss only the linear
block codes with system form the binary field GF(2).
40 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
These binary n-tuples are called as codeword (Code Vector) of the
message u.
41 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
A block code of length n and 2k code words is called a linear (n, k)
code if and only if its 2k code words from a k-dimensional subspace
of the vector space of all the n-tuples over the field GF(2).
For a block code with 2k code words and length n, unless it has a
certain special structure, the encoding apparatus will be
prohibitively complex for large k & n.
42 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
Matrix representation of codes
In fact, a binary code is linear if & only if the modulo-2 sum of two
codeword is also a codeword.
Because an (n, k) linear code C is a k-dimensional subspace of the vector
space Vn of all the binary n-tuples, it is possible the find k linearly
independent codeword, 𝑔0 , 𝑔1 , ⋯ , 𝑔𝑘;1 every codeword v in C is a linear
combination of these k codeword that is,
𝒗 = 𝑢0 𝑔0 + 𝑢1 𝑔1 + ⋯ + 𝑢𝑘;1 𝑔𝑘;1
Where, 𝑢𝑖 = 0 or 1 for 0 ≤ 𝑖 < 𝑘 .
We arrange linearly independent codeword as the rows of a k × 𝑛 matrix
as follows:
43 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
𝒗 = 𝑢0 𝑔0 + 𝑢1 𝑔1 + ⋯ + 𝑢𝑘;1 𝑔𝑘;1
44 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
𝒗 = 𝑢0 𝑔0 + 𝑢1 𝑔1 + ⋯ + 𝑢𝑘;1 𝑔𝑘;1
Clearly, the rows of G generate (or spam) the linear code C.
Therefore the encoder has only to store the k rows of G and to form a
linear combination of these k rows based on the input messages
𝒖 = 𝑢0 , 𝑢1 , ⋯ , 𝑢𝑘;1 .
45 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
Where 𝑝𝑖,𝑗 = 0 or 1.
Then 𝑮 = 𝑷 | 𝑰𝑘 .
46 3/8/2021
Construction of Codes …
Linear Block Codes (n , k):
A desirable property for a linear block code to possess the systematic
structure of the codeword.
A codeword is divided into two parts, the message part and the
redundant checking part ((n-k)Parity bits).
47 3/8/2021
Construction of Codes …
Linear Block Code (n , k):
Let 𝒖 = 𝑢0 , 𝑢1 , ⋯ , 𝑢𝑘;1 be the message to be encoded.
The corresponding codeword is
𝒗 = 𝑣0 , 𝑣1 , ⋯ , 𝑣𝑛;1
𝒗 = 𝑢0 , 𝑢1 , ⋯ , 𝑢𝑘;1 . 𝑮
The components of 𝒗 are:
𝑣𝑛;𝑘:𝑖 = 𝑢𝑖 𝑓𝑜𝑟 0 ≤ 𝑖 < 𝑘 & 𝑣𝑗 = 𝑢0 𝑝0𝑗 + 𝑢1 𝑝1𝑗 + ⋯ +𝑢𝑘;1 𝑝𝑘;1,𝑗 .
For 0 ≤ 𝑗 < 𝑛 − 𝑘 the rightmost digits of a codeword 𝒗 are identical to
the information digits 𝑢0 , 𝑢1 , ⋯ , 𝑢𝑘;1 to be encoded.
The leftmost n-k redundant digits are linear sums of the information
digits.
The n - k equations are called parity check equations of the code.
48 3/8/2021
Construction of Codes …
Encoding:
Encoding i.e. constructing encoded sequence is as follows :
The (7, 4) linear code given in table has the following matrix as a
generator matrix.
𝒈0 1 1 0 1 0 0 0
𝒈1 0 1 1 1 0 1 0
𝐺 = ⋮ =
1 1 1 0 0 1 0
𝒈𝑘;1 1 0 1 0 0 0 1
If 𝒖 = (1101) is the message to be encoded its corresponding codeword
𝑣 = 1. 𝑔0 + 1. 𝑔1 + 0. 𝑔2 + 1. 𝑔3
𝑣 = 1101000 + 0110100 + 1010001
𝑣 = 0001101
49 3/8/2021
Construction of Codes …
Linear Block Code (n , k):
The right most 4 digits of each codeword are identical to the
corresponding information digits.
A linear systematic (n, k) code is completely specified by a kxn
matrix G of the following form:
Table 2 for (7, 4) systematic block code is given below :
Message Codeword Message Codeword
50 3/8/2021
Construction of Codes …
Encoding:
The matrix G given is in systematic form. Let 𝒖 = 𝑢0 , 𝑢1 , ⋯ , 𝑢𝑘;1 be
the message to be encoded & 𝒗 = 𝑣0 , 𝑣1 , 𝑣2 , 𝑣3 , 𝑣4 , 𝑣5 , 𝑣6 be the
corresponding codeword, Then
1 1 0 1 0 0 0
𝑣 = 𝑢0 , 𝑢1 , 𝑢2 , 𝑢3 . 0 1 1 0 1 0 0
1 1 1 0 0 1 0
1 0 1 0 0 0 1
By matrix multiplication, we get the following digits of the codeword
v
𝑣6 = 𝑢3 𝑣5 = 𝑢2 𝑣4 = 𝑢1
𝑣3 = 𝑢0 𝑣2 = 𝑢1 + 𝑢2 + 𝑢3 𝑣1 = 𝑢0 + 𝑢1 + 𝑢2
𝑣0 = 𝑢0 + 𝑢2 + 𝑢3
The codeword corresponding to the message
u = 1011 is v = 1001011
51 3/8/2021
Construction of Codes …
Calculations: C=u. G
1 1 0 1 0 0 0
0 1 1 0 1 0 0
𝐶= 1 0 1 1 .
1 1 1 0 0 1 0
1 0 1 0 0 0 1
Compute r.HT
57 3/8/2021
Construction of Codes …
Cyclic Codes (n, k)
The systematic (n, k) cyclic code is a sub class of linear block code
which are constructed using generator polynomial
Theorem:
The generator polynomial g(X) of an (n , k) cyclic code is a factor of
Xn +1
Theorem:
If g(X) is a polynomial of degree (n-k) and is a factor of X n + 1, then
g(X) generates an (n , k) cyclic code.
58 3/8/2021
Construction of codes …
Generalized Block Diagram of Encoding of (7,4)
Code
𝑢0
uo 𝑢1 𝑢2
Codeword
63 3/8/2021
Fig 6 Diagram for Error Correction
Construction of Codes …
Analytical method of Cyclic Code (n , k)
Encoding in systematic form consists of three steps:
Obtain the remainder b(X) from dividing Xn−k u(X) by the generator
polynomial g(X)
Combine b(X) and Xn−k u(X) to obtain the code polynomial b(X) +Xn−k u(X).
64 3/8/2021
Construction of Codes …
Example of Analytical Method:
g(x)
b(x)=1 0 0
65 3/8/2021
Codeword = 1 0 0 1 0 1 1
Construction of Codes …
Shift Register Method:
Consider the (7,4) cyclic code Generated by 𝑔 𝑋 = 1 + 𝑋 + 𝑋 3 .
Suppose that the message 𝑢 = 1011 is to be encoded.
66 3/8/2021
Construction of Codes …
Example of Cyclic Encoding:
u=1011
𝑆1 = 𝑆0 + 𝑚 + 𝑆2
𝑆0 = 𝑆2 + 𝑚
𝑆2 = 𝑆1
68 3/8/2021
Construction of Codes …
Consider r=(10010110) is received
r=(10010110)
71 3/8/2021
Construction of Codes …
Decoding Using Shift Register Method
Error Pattern and their syndromes:
The received polynomial r(X) shifted into the syndrome register
from the left end.
The seven single-error patterns and their corresponding
syndromes are listed in Table 4
Table 4: Error Pattern and their syndromes in polynomial
and binary form
72 3/8/2021
Construction of Codes …
Cyclic Decoding (Meggit Decoder ):
We see that e6(X) = X6 is the only error pattern with an error at
location X6.
73 3/8/2021
Construction of Codes …
We see that after four more shifts the contents in the syndrome register are (101),
and the erroneous digit r2 is the next digit to come out from the buffer register.
75 3/8/2021
Construction of Codes …
76 3/8/2021
LDPC codes are called regular if its parity-check matrix ' H ' satisfies first two
conditions, otherwise called as an irregular LDPC code
‘H’ has small density of 1’s, hence the name Low Density parity-check codes
78 3/8/2021
LDPC Codes ….
The digital communication system with an Error Correcting
Coding technique mainly consists of
79 3/8/2021
Encoder
Encoding is the process in which k-bit message
(information) is augmented with ‘n-k’ parity- check
bits to form an n-bit LDPC codeword
Let
‘m’ be message;
(m =[ 1 1 0]=[ m1, m2, m3], k =03 bit )
Then
LDPC code for the message shall be
c=mG
where,
‘G’ is Generator matrix
80 3/8/2021
Encoder….
Generator matrix ‘G’ is
Gk x n = [Ik P k x (n-k)]
where,
G is a generator matrix of size k×n
I is a identity matrix of size k×k
P is a parity bit matrix of size k×(n-k)
The ‘G’ matrix is derived from Parity Check matrix ‘H’, where ,
H(n-k)x n = [PT In-k ]
Relation between matrix G and H is such that
H.GT = 0
81 3/8/2021
Encoder….
Let us consider an example (6, 3)
Consider the 3 bit message
m =[ 1 1 0]=[ m1, m2, m3]
Code formation required ‘G’ matrix which can be
derived from ‘H’ matrix,
Let ‘H’ matrix be of size 3×6
1 1 0 1 0 0
H 0 1 1 0 1 0
1 1 1 0 0 1
H PT
I n k
82 3/8/2021
Encoder….
Generator Matrix ‘G’
G Ik P
1 0 0 1 0 1
G 0 1 0 1 1 1
LDPC Code Formation
0 0 1 0 1 1
C mG
1 0 0 1 0 1
Thus the codeC
contains 1 00bit &13 parity
1 3message 1 1 1
0 bits
0 0 1 0 1 1
C 1 1 0 0 1 0
83 3/8/2021
Decoder
Code word is sent through Additive White Gaussian Noise
(AWGN) channel with BPSK modulation.
Received code word is checked for error using ‘H’ matrix as;
H .C T 0
1
1
1 1 0 1 0 0 1.1 1.1 0.1 1.0 0.1 0.0
0
0
0
1 1 0 1
0 0.1 1.1 1.0 0.0 1.1 0.0
1 1.1 1.1 1.0 0.0 0.1 1.0
1 1 1 0 0 1
0
0 0 0
86 3/8/2021
Decoder….
The Tanner Graph for the ‘H’ matrix under consideration is
Tanner Graph
as under. 8
7 V0 V1 V2 V3 V4 V5
1 1 0 1 0 0
6
H 0 1 1 0 1 0 4
1 1 1 0 0 1 2
1 C0 C1 C2
0
0 1 2 3 4 5 6 7
The edges starting & ending at the same vertex is called short
cycle
The number of edges in a short cycle is called as length of
short cycle
Here in a given Tanner graph ( v0 - c0 - v1 - c2 - v0 ) formed
short cycle of length 4
Smallest short cycle in Tanner graph is called Girth of the Tanner
graph 87 3/8/2021
Decoder….
Steps in Message Passing Algorithm(MPA):
Bits in the received codeword's are assigned to variable
nodes of Tanner graph (Intrinsic Bits)
Variable node gets extrinsic bits from their neighboring
check nodes (Interaction amongst vi & cj in H matrix)
Correctness of the intrinsic bit is decided on the basis of
majority of bits received at variable node.
Correction is made accordingly at variable node and again
condition H.CT = 0 is verified.
The process is executed till correction of complete code
word.
88 3/8/2021
Decoder….
Example for Code correction
Correct Code was C 1 1 0 0 1 0
Erroneous code received at Receiver C 1 1 0 0 1 1
At receiver the bits are assigned to variable nodes vi
v0 1, v1 1, v2 0, v3 0, v4 1, v5 1
Degree of variable nodes decide the edges between check node and
variable node
1 C0 C1 C2
check nodes
0
0 1 2 3 4 5 6 7
5 1 0
4
1
3
1 C0 C1 C2
check nodes
0 91 3/8/2021
0 1 2 3 4 5 6 7
Decoder….
Variable node v2
Degree of node is 2
Three bits are received at v2
one intrinsic bit = 0 from channel
2 extrinsic bits from check node c1 ,c2 (0, 1)
Majority selection amongst bits (0, 0,1) is 0 and same is updated at v2
Tanner Graph
8
Trans.codeword 1 1 0 0 1 0
Recd.codeword 1 1 0 0 1 1
7 V0 V1 V2 V3 V4 V5
variable
nodes
6
1
5
0
4
1 C0 C1 C2
check nodes
0
0 1 2 3 4 5 6 7
92 3/8/2021
Decoder….
Variable node v3
Degree of node is 1
Two bits are received at v3
one intrinsic bit = 0 from channel
one extrinsic bits from check node c0 (0)
Majority selection amongst bits (0, 0) is 0 and same is updated at v3
Tanner Graph
8
Trans.codeword 1 1 0 0 1 0
Recd.codeword 1 1 0 0 1 1
7 V0 V1 V2 V3 V4 V5
variable
nodes
6 0
5
1 C0 C1 C2
check nodes
0 93 3/8/2021
0 1 2 3 4 5 6 7
Decoder….
Variable node v4
Degree of node is 1
Two bits are received at v4
one intrinsic bit = 1 from channel
one extrinsic bits from check node c1 (1)
Majority selection amongst bits (1, 1) is 1 and same is updated at v4
Tanner Graph
8
Trans.codeword 1 1 0 0 1 0
Recd.codeword 1 1 0 0 1 1
7 V0 V1 V2 V3 V4 V5
variable
nodes
6
5
1
4
1 C0 C1 C2
check nodes
0 94 3/8/2021
0 1 2 3 4 5 6 7
Decoder….
Variable node v5
Degree of node is 1
Two bits are received at v5
one intrinsic bit = 1 from channel
one extrinsic bits from check node c2 (0)
Majority selection amongst bits (1, 0) in this case authenticity of
extrinsic bit shall be consider and is updated at v5
Tanner Graph
8
Trans.codeword 1 1 0 0 1 0
Recd.codeword 1 1 0 0 1 1
7 V0 V1 V2 V3 V4 V5
variable
nodes
6
4
0
3
1 C0 C1 C2
check nodes
0
0 1 2 3 4 5 6 7
Updated code word is verified with the = 0, If found true the H.CT
correction is completed . Otherwise the process is continued till the
satisfaction of the condition
95 3/8/2021
Sum Product Algorithm
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
Calculate H.zT
𝑎𝑙𝑙 𝑒𝑙𝑒𝑚𝑒𝑛𝑡 = 0, 𝑇𝑒𝑟𝑚𝑖𝑛𝑎𝑡𝑒
𝐻𝑧 𝑇 =
𝑎𝑙𝑙 𝑒𝑙𝑒𝑚𝑒𝑛𝑡 ≠ 0, 𝐶𝑜𝑛𝑡𝑖𝑛𝑢𝑒
1
1 1 0 1 0 0 0
𝐻 = 0 1 1 0 1 0 &z= 1
0
1 1 1 0 0 1 1
1
𝐻𝑧 𝑇 = 1 0 1 0 𝑇
102 3/8/2021
Calculation of LLR Values
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
103 3/8/2021
Calculation of LLR Values
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
104 3/8/2021
Calculation of LLR Values
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
105 3/8/2021
LLR Calculation to Check Validity
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
106 3/8/2021
LLR Calculation to Check Validity
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
107 3/8/2021
LLR Calculation to Check Validity
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
109 3/8/2021
Iteration III
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
Aposteriori Probabilities
0.6779 3.9907 0 2.0695 0 0
𝐿𝑖,𝑗 = 0 0.4940 −1.2123 0 −4.3388 0
2.1426 0 0 0 −4.6041 −1.8963
0 0 −58721 2.3674 0 0.5984
110 3/8/2021
E11, E12, E14 Calculation Iteration III
C1 C2 C3 C4
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
Calculate H.zT
𝑎𝑙𝑙 𝑒𝑙𝑒𝑚𝑒𝑛𝑡 = 0, 𝑇𝑒𝑟𝑚𝑖𝑛𝑎𝑡𝑒
𝐻𝑧 𝑇 =
𝑎𝑙𝑙 𝑒𝑙𝑒𝑚𝑒𝑛𝑡 ≠ 0, 𝐶𝑜𝑛𝑡𝑖𝑛𝑢𝑒
0
1 0 0 0
1 1 0
𝐻= 0 1 1 0 1 0 &z = 1
0
1 1 1 0 0 1 1
1
𝐻𝑧 𝑇 = 0 0 0 0 𝑇
117 3/8/2021
References