Lecture 3

Introduction to Vector Space Theory

Matrices

Linear Block Codes

c = mG

where, c code word (vector)

m message word (vector)

G Generator matrix

Vector Space-Introduction
An n-dimensional vector has a form x = ( x1, x2, x3, , xn ) . The set Rn of n-dimensional vectors is a vector space. Any set V is called a vector space if it contains objects that behave like vectors: ie, they add & multiply by scalars according to certain rules. In particular, they must be closed under vector addition and scalar multiplication. But addition & scalar multiplication need not be defined conventionally!
3

Contd
Let V denote the vector space.The addition on V is vector addition.The scalar multiplication combines a scalar from a Field F and a vector from V. Hence V is defined over a field F. V must form a commutative group under addition For any element a in F and any element v in V, a.V is an element in V. Distributive law- a.(u+v)=a.u+a.v Associative law- (a.b).v=a.(b.v)
4

Contd.
Important vector spaces: R, R2, R3, Rn with usual + and scalar multn. Mmn ; the set of all m x n matrices Pn; all polynomials of degree n
Consider a vector space over binary field F2.Consider the sequence u=u0un-1 where the ui s are from {0,1}.We can construct such 2n ntuples over F2.Let Vn denote this set. Vn is a Vector space over F2
5

Subspaces
A set W of vectors is a subspace of vector space V if and only if W is a subset of V and W is itself a vector space under the same addition and scalar multiplication. For any two vectors u,v W, (u+v) W. For any element a in F and any u in W, a.u must be in W.
6

Contd
To test if W is a Subspace We should, but need not, check all the properties of a vector space in W: most hold because Ws vectors are also in the bigger vector space V. But we must check closure in W: linear combinations of vectors in W must also lie in W.

This means the zero & additive inverses must be in W too.

7

Examples
Let u1,.,uk be a set of k vectors in V over a field F. The set of all linear combinations of u1,.,uk forms a subspace of V. The set of polys of degree 2 or less is a subspace of the set of polynomials of degree 3 or less. The set of integers is not a subspace of R, because the set of scalars includes fractions, eg 1/2.
8

Spanning Sets &Linear Independence

A set S = {u1,u2,.......,un} of vectors is said to span a vector space V if every vector in V can be expressed as a linear combination of the vectors in S. Ex:( x, y, z ) = x i + y j + z k, so every vector in R3 is a linear combination of i, j & k. If any vector in a set can be expressed as a linear combination of the others, we call the set linearly dependent. If not, the set is linearly independent.
9

Basis set
A set of linearly independent vectors is a basis for a Vector space V if each vector in V can be expressed in one and only one way as a linear combination of the set. In any Vector space or subspace there exists at least one set B of linearly independent vectors which span the space. The no. of vectors in the Basis of a Vector space is the dimension of the Vector space. One example of a basis are the vectors (1,0,,0), (0,1,,0),, (0,0, , 1).
10

Orthogonality
Let u= (u0 , u1 ,.....u n 1 ) and v= (v0 , v1 ,....vn 1 )
be two n-tuples in Vn. We define the inner product(dot product) as u.v=

u0 v0 + u1v1 + ........ + un 1vn 1 where the multiplication and addition are

carried out in mod-2.. The inner product is a scalar. If u.v=0, then u and v are said to be orthogonal to each other The inner product has the following properties (1) u.v=v.u (2) u.(v+w)=u.v+u.W (3)(au).v=a(u.v)
.
11

Matrices
A k x n matrix over F2 is a rectangular array with k rows and n columns.

g01 g02 ..... g0,n1 g00 g g g ..... g 11 12 1,n1 10 . . . . G= . . . . . . gk 1,0 gk 1,1 gk 1,2 ..... gk 1,n1
where each

gij

with

0 i k and 0 j n
12

is an element from the binary field F2.

 G is also represented by its k rows g 0 , g 0 ,.....g k 1 as

g0 g 1 G= . . g k 1
Each row of G is an n-tuple and each column is a k-tuple over F2.
13

 If k (with k n ) rows of G are linearly independent, then the 2k linear combinations of of these rows form a k dimensional subspace of the vector space Vn of all the ntuples over F2. This subspace is called the

row space of G

Elementary row operations will not change the row space of G

14

 Let S be the row space of a k x n matrix G over F2 whose rows are linearly independent. Let Sd be the null space of S. Then the dimension of Sd is n-k. Consider (n-k) linearly independent vectors in Sd. These vectors span Sd. We can form an (n-k) x n matrix H as
h0 h00 h h 1 10 H = . = . . . g n k 1 hn k 1,0 h01 h11 . . hn k 1,1 h02 h12 . . hn k 1,2 ..... hn k 1,n 1 ..... ..... . . h0,n 1 h1,n 1 . .
15

The row space of H is Sd

 Since each row gi is a vector in S and each row hj of H is a vector in Sd, the inner product of gi and hj must be zero. As the row space S of G is the null space of the row space Sd of H, S is called the null space

or dual space of H.

16

Theorem
For any k x n matrix G over F2, with k linearly independent rows, there exists an (n-k) x n matrix over the same field with (n-k) linearly independent rows such that for any row gi in G and any hj in H, gi.hj = 0. The row space of G is the null space of H and vice versa.

17