Graph: Edge Connectivity
Graph: Edge Connectivity
Edge Connectivity:
The edge connectivity of a connected graph is the cardinality (number) of the smallest set of
edges, removal of which results into a disconnected graph.
In the figure:
The edge connectivity is 1 (one), because by removing the edge (C, D) it will be a disconnected graph.
Similarly in the graph:-
We will have to remove any of two edge set such as { AC , BC} or {CD, CE} to convert the above
connected graph into a disconnected graph. So the edge connectivity of the above graph is 2(two).
Theorem 1:
The Edge connectivity of any graph G is less than or equal to the degree of any
Node (d).
Proof: Let x be any node of the graph having degree (d). To disconnect the graph,
We will have to disconnect at least one node from the graph. Which can happen only if by
removing the entire edges incident to x? So that we will have to remove all edges incident to x, So that
the edge connectivity of the graph will be d .
So that by removing all the edges of x, x will be an isolated vertex & the graph will be a
disconnected graph.
So Edge connectivity ≤ d
Hence Proved.
Note: No. of Vertices (Nodes or Points) can be represented by any one term out of these terms:
|V |, v, n, p
No. of Edges can be represented by any one term out of these terms: |E|, e, m, q
No. of Regions can be represented by any one term out of these: - |R|, r
So doesn’t get confused among any of these.
Theorem 2:
2E
The Edge connectivity of a connected graph is always less than or equal to
N
where E is the no. of edges & N is the no. of vertices.
Proof: Let the minimum degree of the graph is d. So that, the degree of each of N nodes of graph G
is greater than equal to d. So that
The sum of all the degrees of all the nodes is ≥ N d. …………..… (1)
Since the sum of the degree of each of the node of the graph = 2E …………….. (2)
2E
Since d must be an integer. So by taking lower bound of the R.H.S., we get: ---- d ≤ ………. (3)
N
Since as in the above theorem we know that: ------ The Edge connectivity ≤ d …………(4)
2E
So that equation (3) and (4), we get: ------------Edge connectivity ≤
N
Hence Proved.
Theorem 3:
Let a graph G = (V , E ) be a simple graph, show that the maximum degree of any
vertex of graph G is n − 1 where n is the no. of nodes.
Proof: We will prove the above theorem by method of induction on the number of nodes.
Step (i): Let n = 1 So that there is only one node. That is possible only if we are having with
• ← single node
Since loops are not allowed in the simple graph. So that there is not any edge with one node.
So that degree of the vertex =0
∴ Results hold for n = 1
Step (ii) Let the result holds for m no. of Nodes so that with m no. of Nodes
Maximum Degree of each node = m − 1 …………….. (1)
Step (iii) Now we will check the check the above result for m + 1 no. of Nodes.
Since by adding a new vertex in the simple graph, the maximum no. of edges adjacent to every
vertex (node) will be increased by 1 (one).
So now maximum degree of every vertex is = m − 1 + 1 = m
So for m + 1 no. of vertices, the maximum degree of any vertex of graph G is m.
Theorem 4:
n(n − 1)
Show that maximum no. of edges in a simple graph with n vertices is
2
Proof: We know that, the sum of degrees of all vertices of a graph G = (V , E ) is twice the number of
edges of G .
So that Sum of the degree of all the nodes = 2 E ………………….. (1)
Since we know that in a simple graph with n vertices:
Maximum degree of any vertex = n − 1
So the sum of the degree of all the nodes ≤ n(n − 1) ……………. (2)
So from equation (1) and (2)
So 2 E ≤ n(n − 1)
n(n − 1)
E≤
2
n( n − 1)
So Maximum no. of edges in a graph G is Hence Proved.
2
Theorem 5:
Show that the sum of the squares of in degree over all vertices is equal to sum of
the squares of the out degree over all vertices in any Complete Directed Graph.
Where
A = in degree of any node
B = out degree of any node
V = set of vertices
∑ A= ∑B =
v∈V v∈V
So
∑ ( A − B) = 0 …………….. (1)
v∈V
∑ ( A − B)( A + B) = 0
v∈V
∑(A 2
−B )=0
2
v∈V
∑A =∑B
2
2
v∈V v∈V
Hence Proved
Theorem 6:
Let G = (V , E ) be an undirected graph with K components and n no. of vertices &
m no. of edges. Then prove that m + k ≥ n
So 0 + 4 ≥ n ⇒ so m + k ≥ n
Now each time when an edge is added to the graph then either the components remains same
e.g.:
m=2 m=3
n=5 n=5
k =3 k =3
Or will decreased by one
m=2 m=3
n=5 n=5
k =3 k =2
So 3+2 ≥ 5
So m + k ≥ n
So this result holds for each. Of components
Theorem 7:
In a connected graph G = (V , E ) a vertex v is a cut point if ∃ two distinct vertices
v & v such that every path connected v1 & v contains v .
1 2 2
Proof: Let us consider that v is a cut point after deletion the graph will be disconnected. So by
definition of cut point G − {v} is not connected.
Let there is a graph such that, now here in the graph G , v is a cut point.
So that there does not exists any path joining v and v in G which does not contain v .
1 2
Conversely:
Suppose that there exists two vertices v & v such that every path connecting v & v
1 2 1 2
contains v .
Let if possible assume v is not a cut point. Therefore by definition of cut point G − {v} is still
connected.
Since v , v ∈ G − {v} and G − {v} is connected.
1 2
This is contradictory to our supposition that every path connecting v1 & v 2 contain v .
So.
Our assumption is wrong. So every path connecting v & v contains the node v . 1 2
Hence proved.
Theorem 8:
Step 1:
Case (i) if E = 1
Now there can be two sub cases:
(a) Only one vertex and only two regions because the only one edges is the self loop.
V =1 E =1 R=2
L.H .S .
V – E + R = 1 -1 + 2 = 2 = R.H.S.
So Euler’s formula holds for this case.
(b) Two vertices, only one region because there is only one edge between two vertices.
V = 2 E =1 R =1
L.H.S.
V − E + R = 2 − 1 + 1 = 2 = R.H.S.
So,
Euler’s formula holds for this case.
Step II: Let the Euler’s formula is true for E ′ no. of edges, V ′ no. of vertices and R′
no. of regions. So that V ′ − E ′ + R′ = 2 -------------(2)
Step III:
So that let us check for the E ′ + 1 no. of edges.
Case (i) :
If by increasing one edge but not any new node is increased:
Case (ii) if by increasing one edge and one node is also introduced.
Proof: (a) We assure that | E | >1 since each edge contributes in maximum two regions. So
that the each region will have 3 degree or more
Hence Proved.
TREE
Theorem 1:
If G be a connected graph with n nodes then G is a tree if & only if G has
exactly one simple path between any two nodes.
Or
If T = (V , E ) be a tree, then there exists a unique path between every two
vertices ofT .
Or
Prove that there is one and only one path between every pair of vertices in a treeT .
Proof: Since G (T ) a connected graph. So it will contain at least one path between pair of
nodes. E.g. a simple path from a to d .
Which is {a, b, c, d,}
Now let if possible, Let us assume that path is not unique so that exists another path {a, e, d }
between a and d .
Theorem 2 :
In a connected graph G with n nodes: G is a tree if & G has exactly n − 1
edges.
Or
In a tree T (V , E ), | V |> | E | i.e. the no. of vertices are more than the no. of
edges.
Or
Prove that in a tree | E | = |V |-1
Proof: We will prove the theorem by the method of induction on the no. of induction on the
no. of node | V | or n .
Step I: Let |V |=(n=1) so that the tree has only one vertex
| V |=1
| E |=0
Then this is not any edge in tree. So | E |= |V |-1 so that theorem is satisfied for |V | =1
Step II: Let the result is true for |V | ≤ m or n ≤ m there are (m − 1) edges.
Step III: Now we will check the result for (m + 1) nodes. Now deleting exactly
One edge e.g. ( A, E ) from the tree so that we obtain a graph with exactly Two
connected components.
In which, connected component I containing those nodes of G which does not have the
deleted edge, and other connected component II has the deleted edge ( A, E ) from the
right sub tree.
Each component is tree with m & fewer nodes for which the result holds.
Suppose the I component has r nodes & II component has q nodes
So the I component will have (r − 1) edges.
& the Ii component will have (q − 1) edges.
r + q = m +1
And these are total (r − 1 + q − 1) = r + q − 2 no. of edges in the connected components.
So no of nodes = r + q
m +1 = r + q ……………. (1)
Total no of edges in the connected components = (r − 1) + ( q − 1)
= r −1− 2
= (m + 1) − 2 (from (1))
= −m − 1
So that the connected components has (m − 1) edges, after adding one deleted edges ( A, E )
So that the total no. of edges in the main graph = m − 1 + 1 = m
So that for (m+1) nodes there are m edges in the tree.
So this result is also true for (m + 1) nodes.
So By P.M.I, the result is the true for each |V |
Hence Proved.
Theorem :3
A rooted there with in nodes has n − 1 edges.
Proof : Since in the rooted tree the root node has zero in degree & all other ahs one in
degree. So that each of the node except the node is the lower and node of exactly one edge.
So that there are n- 1 edges.
Theorem 4 :
A regular m-ary tree with i interior nodes has n = mi + 1 nodes in all.
Proof : Each of the i interior nodes has m Childs. So that there are
maximum nodes.
So that Maximum no. of Childs (nodes) with i interior = mi
Since there is one more root node is not a child of any node.
So total no. of nodes = mi +1
Hence Proved.
Theorem 5 :
Show that the sum of the degree of vertices of a tree with n nodes is 2 n
Proof: Since in the tree having n nodes there are total ( n -1) no. of edges.
The degree contributed by 1 edge =2
The degree contributed by ( n -1) edge =2 n -2
Hence Proved.
Theorem 6 :
Prove that if sum of the degrees of the graph of n nodes is 2 n -2 the
connected graph will be a tree.
Theorem 7 :
Prove that a regular binary tree has an odd. Of vertices.
Proof: Let us assume that the complement of a spanning tree contains a cut set.
So that three is no. common edge in a spanning tree & cut set.
This means that removal of cut set from the graph, will not separated the
graph into two connected components.
This is contradiction to the definition of cut set.
So our assumption is wrong. So that complement of a spanning tree does
not contain a cut set.
Hence proved.
Theorem 9 :
Prove that any edge of a connected graph G is a branch of some
spanning tree of G .
Proof: Let T will be some spanning tree of a connected graph T then there is
nothing to prove. Let e is not in T
Then add edge e to T to T will from a cycle. Remove any edge other
e
than from this cycle. Then again we get another spanning tree T containing
edge e .
∴ Any edge of a connected graph is a branch of some spanning tree.
Hence Proved