J. KSIAM Vol.11, No.

4, 19-38, 2007

APPLICATIONS OF GRAPH THEORY

S. Pirzada and Ashay Dharwadker

Abstract

Graph theory is becoming increasingly significant as it is applied to other areas

of mathematics,
science and technology. It is being actively used in fields as varied as
biochemistry (genomics),
electrical engineering (communication networks and coding theory), computer science
(algorithms
and computation) and operations research (scheduling). The powerful combinatorial
methods
found in graph theory have also been used to prove fundamental results in other
areas of pure
mathematics. This paper, besides giving a general outlook of these facts, includes
new graph
theoretical proofs of Fermat’s Little Theorem and the Nielson-Schreier Theorem. New
applications to DNA sequencing (the SNP assembly problem) and computer network
security
(worm propagation) using minimum vertex covers in graphs are discussed. We also
show how to
apply edge coloring and matching in graphs for scheduling (the timetabling problem)
and vertex
coloring in graphs for map coloring and the assignment of frequencies in GSM mobile
phone
networks. Finally, we revisit the classical problem of finding re-entrant knight’s
tours on a

chessboard using Hamiltonian circuits in graphs.

INTRODUCTION

Graph theory is rapidly moving into the mainstream of mathematics mainly

because of its applications in diverse fields which include biochemistry,
electrical
engineering (communications networks and coding theory), computer science
(algorithms and computations) and operations research (scheduling). The wide
scope of these and other applications has been well-documented cf. [5] [19]. The
powerful combinatorial methods found in graph theory have also been used to
prove significant and well-known results in a variety of areas in mathematics
itself.

The best known of these methods are related to a part of graph theory called

19
8. Pirzada and Ashay Dharwadker

matchings, and the results from this area are used to prove Dilworth’s chain
decomposition theorem for finite partially ordered sets. An application of
matching in graph theory shows that there is a common set of left and right coset
representatives of a subgroup in a finite group. This result played an important
role in Dharwadker’s 2000 proof of the four-color theorem [8] [18]. The existence
of matchings in certain infinite bipartite graphs played an important role in
Laczkovich’s affirmative answer to Tarski’s 1925 problem of whether a circle is
piecewise congruent to a square. The proof of the existence of a subset of the real
numbers R that is non-measurable in the Lebesgue sense is due to Thomas [21].
Surprisingly, this theorem can be proved using only discrete mathematics
(bipartite graphs). There are many such examples of applications of graph theory
to other parts of mathematics, but they remain scattered in the literature [3]
[16].
In this paper, we present a few selected applications of graph theory to other
parts

of mathematics and to various other fields in general.

1. THE CANTOR-SCHRODER-BERNSTEIN THEOREM

Here we discuss the graph theoretical proof of the classical result of Schréder and
Bernstein. This theorem was presumed to be an obvious fact by Cantor
(cf. remark 1.2) and later proved independently by Schréder (1896) and Bernstein
(1905). The proof given here can be found in [14] and is attributed to Kénig.

1.1. Theorem (Cantor-Schroder-Bernstein). For the sets A and B, if there is an

injective mapping f; AB and an injective mapping g: B—A, then there is a

bijection from A onto B, that is, A and B have the same cardinality.

Proof. Without loss of generality, assume A and B to be disjoint. Define a

bipartite
graph G = (A, B, E), where xy € E if and only if either f(x) = y or g(v) =x, x € A,
y € B. By the hypothesis, 1 < a(v) < 2 for each v of G. Therefore, each component

20
21

Applications of Graph Theory

of G is either a one-way infinite path (that is, a path of the form xo, X1,....
Xns--+)s
or a two-way infinite path (of the form ..., X), Xot,---5 X-1 > X0s X1y--+5
Xny---), OF A
cycle of even length with more than two vertices, or an edge. Note that a finite
path of length greater or equal to two cannot be a compound of G. Thus, in each
component there is a set of edges such that each vertex in the component is
incident with precisely one of these edges. Hence, in each component, the subset

of vertices from A is of the same cardinality as the subset of vertices from B. =

1.2. Remark. Cantor inferred the result as a corollary of the well-ordering

principle. The above argument shows that the result can be proved without using

the axiom of choice.

2. FERMAT’S (LITTLE) THEOREM

There are many proofs of Fermat’s Little Theorem. The first known proof was
communicated by Euler in his letter of March 6, 1742 to Goldbach. The idea of
the graph theoretic proof given below can be found in [12] where this method,
together with some number theoretic results, was used to prove Euler’s

generalization to non-prime modulus.

2.1, Theorem (Fermat). Let a be an integer and let p be a prime such that a is not

divisible by p. Then, a” — a is divisible by p.

Proof. Consider the graph G = (V, £), where the vertex set V is the set of all
sequences (a), a,..., Gp») of natural numbers between | and a (inclusive), with
a; # a; for some i#/. Clearly, Vhas a’ — a elements. Let u = (i, wu, ..., Up),

v = (up, W1,..., Mp1) € V. Then, we say wv © E. With this assumption, each vertex
8. Pirzada and Ashay Dharwadker

of G is of degree 2. So, each component of G is a cycle of length p. Therefore, the

a’ -a ;
. That is, p/a’-a. =

number of components is

3, THE NIELSON-SCHREIER THEOREM

Let H be a group and § be a set of generators of H. The product of generators and

their inverses which equals identity (1) is called a trivial relation among the
generators in S if | can be obtained from that product by repeatedly replacing xx"
or x x by 1. Otherwise such a product is called a non-trivial relation. A group H
is free if H has a set of generators such that all relations among the generators
are
trivial.

Babai [2] proved the Nielson-Schreier Theorem for subgroups of free

groups, as well as other results in diverse areas, from his Contraction Lemma. The
particular case of this lemma when G is a tree, and its use in proving the Nielson-

Schreier Theorem, was also observed by Serre [20].

3.1. Contraction Lemma. Let H be a semi-regular subgroup of the automorphism

group of a connected graph G. Then, G is contractible onto some Cayley graph of
H. The proof of this lemma is technical, although it only uses ideas from group
theory and graph theory.

Let H be a group and h © H. Let he be a permutation of H obtained by

multiplying all the elements of AH on the right by A”. The collection
Ar = the: h © A} is a regular group of permutations (under composition) and is
called the (right) regular permutation representation of H.

It can be seen [2] that G is a Cayley graph of the group H if and only if G

is connected and Hg is a subgroup of the automorphism group of G.

22
23

Applications of Graph Theory

The automorphism group of a graph G is the group of all permutations p of the

vertices of G with the property that p(x)p(y) is an edge of G if and only if xy is
an
edge of G.

A group H of permutations acting on a set V is called semi-regular if for

each x € V, the stabilizer H, = {h € H: x= x} consists of the identity only, where
x" denotes the image of x under A. If H is transitive and semi-regular, then it is
regular,

Let (H, o) be a group and S be a set of generators of H, not necessarily

minimal.

The Cayley graph G(H, S) of (H, 0) with respect to S, has vertices

x, y,...€ H, and xy is an edge if and only if either x = yoa or y = xoa, for some
aeéS.

If G is any graph and e = xy an edge of G, then by contraction along e, we

mean the graph G’ obtained by identifying the vertices x and y.

We say that a graph G; is contractible onto a graph G» if there is a

sequence of contractions along edges which transforms G; to G2.

3.2. Corollary. If J is a subgroup of a group H, then any G(H, S) is contractible

onto G(J, T) for some set T of generators of J.

3.3. Theorem (Nielson-Schreier). Any subgroup of a free group is free.

Proof. We first show that in any group H and for any set S of generators of H, the
Cayley graph G(H, S) contains a cycle of length > 2 if and only if there is a
nontrivial relation among the generators in S. To show this, suppose

Xo, Mines Xn = Xo is a_ cycle of G(H, S). Then, there are

a, € S, 1 < i < nn, such that x,,a" =x,, where e, € {l,-l}. Hence,

Xx, =X, 40" =x, a5" a s..sxyaj'ay.a”, that is, the —_ identity

n
8. Pirzada and Ashay Dharwadker

l=ay'a;...a*. If this were a trivial relation, then there would exist an integer i,
| <i<u, such that a;=a;-; and ¢, = —¢,,,. However, this implies that x;) = xj+1, a
contradiction. Similarly, if aj'a;...a°"=1 is a nontrivial relation, then
X0, Xijeee) Xl, Yn, Where x, =x,,4;',1<isn, and xo = x,, is a closed trial in

G(H, S), which must contain a cycle.

Suppose now that H is a free group, S a minimal set of generators of H,

and J a subgroup of H. Since there is no nontrivial relation on the elements of 5,
G(H, S) does not contain a cycle. Also, from the Corollary above, G(H, 5) is
contractible onto G(/, T) for some set T of generators of J. Because any
contraction of a cycle-free graph is again cycle free, G(/, T) must be cycle free,
and, thus, there is no nontrivial relation on the elements of 7. Hence, / must be a

free group, freely generated by 7. =

4. THE SNP ASSEMBLY PROBLEM

In computational biochemistry there are many situations where we wish to resolve

conflicts between sequences in a sample by excluding some of the sequences. Of
course, exactly what constitutes a conflict must be precisely defined in the
biochemical context. We define a conflict graph where the vertices represent the
sequences in the sample and there is an edge between two vertices if and only if
there is a conflict between the corresponding sequences. The aim is to remove the
fewest possible sequences that will eliminate all conflicts. Recall that given a
simple graph G, a vertex cover C is a subset of the vertices such that every edge
has at least one end in C. Thus, the aim is to find a minimum vertex cover in the
conflict graph G. (in general, this is known to be a NP-complete problem [13]).
We look at a specific example of the SNP assembly problem given in [15] and
show how to solve this problem using the vertex cover algorithm [6].

A Single Nucleotide Polymorphism (SNP, pronounced “snip”) [15] is a single base

mutation in DNA, It is known that SNPs are the most common source of genetic

24
25

Applications of Graph Theory

Figure 4.1. The DNA double helix and SNP assembly problem

polymorphism in the human genome (about 90% of all human DNA

polymorphisms). The SNP Assembly Problem [15] is defined as follows. A SNP
assembly is a triple (S, F, R) where S = {s), ..., Sn} is a set of n SNPs,
F= {fi, ..., fm} is a set of m fragments and R is a relation R: S x F — {0, A, B}
indicating whether a SNP s; € S does not occur on a fragment fj € F (marked by 0)
or if occurring, the non-zero value of s; (A or B). Two SNPs s; and s; are defined
to
be in conflict when there exist two fragments f, and fj such that exactly three of
R(si, fi), R(si, 1), R(s;, fe), R(S;, 7) have the same non-zero value and exactly
one
has the opposing non-zero value. The problem is to remove the fewest possible
SNPs that will eliminate all conflicts. The following example from [15] is shown
in the table below. Note that the relation R is only defined for a subset of S x F

obtained from experimental values.

26
8. Pirzada and Ashay Dharwadker

R {ft fr fs fs f
st A B B
52 B A A A 0
83 0 0 B B A
Ss A 0 A 0 B
85 A B B B A
S6 B A A 0

Note, for instance, that s; and ss are in conflict because R(s), 2) = B, R(s1, fs)
= B,
R(ss, #2) = B, R(ss, fs) = A. Again, sq and s¢ are in conflict because R(s4, fi) =
A,
R(s4, 3) = A, R(s6, fi) = B, R(s6, f3) = A. Similarly, all pairs of conflicting
SNPs are
easily determined from the table. The conflict graph G corresponding to this SVP

assembly problem is shown below in figure 4.2.

(5) )

Figure 4.2. The conflict graph G

We now use the vertex cover algorithm [6] to find minimal vertex covers in the
conflict graph G. The input is the number of vertices 6, followed by the adjacency
matrix of G shown below in figure 4.3. The entry in row 7 and column / of the
adjacency matrix is | if the vertices s; and s; have an edge in the conflict graph
and

0 otherwise.
27

Applications of Graph Theory

oOroococo
oorcvcco
ooooco;9o
ProoUr oO
ooro9or
oOorcooqo

Figure 4.3. The input for the vertex cover algorithm

The vertex cover program [6] finds two distinct minimum vertex covers, shown in

figure 4.4.

Minimum Vertex Cover: Si, S4

Minimum Vertex Cover: Sy, S5

Figure 4.4. The output of the vertex cover algorithm

Thus, either removing s;, s4 or removing s4, ss; solves the given SNP assembly

problem.
28
8. Pirzada and Ashay Dharwadker

5. COMPUTER NETWORK SECURITY

A team of computer scientists led by Eric Filiol [11] at the Virology and
Cryptology Lab, ESAT, and the French Navy, ESCANSIC, have recently used the
vertex cover algorithm [6] to simulate the propagation of stealth worms on large
computer networks and design optimal strategies for protecting the network

against such virus attacks in real-time.

Figure 5.1. The set {2, 4,5} is a minimum vertex cover in this computer network

The simulation was carried out on a large internet-like virtual network and
showed that that the combinatorial topology of routing may have a huge impact

on the worm propagation and thus some servers play a more essential and
29

Applications of Graph Theory

significant role than others. The real-time capability to identify them is

essential
to greatly hinder worm propagation. The idea is to find a minimum vertex cover
in the graph whose vertices are the routing servers and whose edges are the
(possibly dynamic) connections between routing servers. This is an optimal
solution for worm propagation and an optimal solution for designing the network
defense strategy. Figure 5.1 above shows a simple computer network and a

corresponding minimum vertex cover {2, 4, 5}.

6. THE TIMETABLING PROBLEM

In a college there are m professors x), x2, ..., X» and m subjects v1, 2, ..., ¥n
to be
taught. Given that professor x; is required (and able) to teach subject y; for pj
periods (p = [pj] is called the teaching requirement matrix), the college
administration wishes to make a timetable using the minimum possible number of
periods. This is known as the ¢imetabling problem [4] and can be solved using the
following strategy. Construct a bipartite multigraph G with vertices
X15 X25 1:5 Xm» Vis Vas +++5 Yn Such that vertices x; and y; are connected by p;
edges.
We presume that in any one period each professor can teach at most one subject
and that each subject can be taught by at most one professor. Consider, first, a
single period. The timetable for this single period corresponds to a matching in
the
graph and, conversely, each matching corresponds to a possible assignment of
professors to subjects taught during this period. Thus, the solution to the
timetabling problem consists of partitioning the edges of G into the minimum
number of matchings. Equivalently, we must properly color the edges of G with
the minimum number of colors. We shall show yet another way of solving the
problem using the vertex coloring algorithm [7]. Recall that the ine graph L(G) of
G has as vertices the edges of G and two vertices in L(G) are connected by an
edge if and only if the corresponding edges in G have a vertex in common. The
line graph L(G) is a simple graph and a proper vertex coloring of L(G) yields a

proper edge coloring of G using the same number of colors. Thus, to solve the
8. Pirzada and Ashay Dharwadker

timetabling problem, it suffices to find a minimum proper vertex coloring of Z(G)

using [7]. We demonstrate the solution with a small example.
Suppose there are four professors x1, x2, x3, x4 and five subjects v1, 2, V3,

¥4, ¥s to be taught [4]. The teaching requirement matrix p = [p,] is given below in

figure 6.1.
P| Mi | ¥2) V3 | Ya | Ys
x} 2/0}]1)140
x2} O0}1)0)1)0
x3/0]/1)/1)140
xy} O}O]0)1) 1

Figure 6.2. The bipartite multigraph G

30
31

Applications of Graph Theory

We first construct the bipartite multigraph G shown above in figure 6.2. Next,

we construct the line graph 1(G). The adjacency matrix of L(G) is given below.

01110000000
10110000000
1101200031000
11100100110
00000110000
00011000110
000010031100
00100010100
000102111010
000101200101
00000000010

Now, we use the vertex coloring algorithm [7] to find a minimum proper

4-coloring of the vertices of L(G).

Figure 6.3. A minimum proper 4-coloring of the vertices of L(G)

8. Pirzada and Ashay Dharwadker

Vertex Coloring: ( 1 , green )( 2, red )(3, blue )(4, yellow) (5, yellow )
(6, green )(7, green) (8, yellow )(9, red )( 10, blue ) (11, yellow ). This,

in turn, yields a minimum proper edge 4-coloring of the bipartite multigraph G:

Figure 6.4. A minimum proper 4-coloring of the edges of G

Edge Coloring: ( {xi,yi}, green ) ( {x:,yi} , red ) ( {xi,y3} , blue )

( {x1,ya} , yellow ) ( {X2,y2} , yellow ) ( {x2,y4} , green ) ( {x3,y2} , green )
( {x3,y3} , yellow ) ( {x3,ya} , red ) ( {x4,y4} , blue ) ( {x4,ys} , yellow ).
Interpret
the colors green, red, blue, yellow as periods 1, 2, 3, 4 respectively. Then, from
the edge coloring of G, we obtain a solution of the given timetabling problem as

shown below in figure 6.5.

32
33

Applications of Graph Theory

1/2\)3\4
1 Mi | Vi | V3 | M4
x2 | Wa J2
X3 | V2 | Va ¥3
x4 va | Vs

Figure 6.5. The timetable

7. MAP COLORING AND GSM MOBILE PHONE NETWORKS

Given a map drawn on the plane or the surface of a sphere, the famous four color
theorem asserts that it is always possible to properly color the regions of the map
such that no two adjacent regions are assigned the same color, using at most four
distinct colors [8] [18] [1]. For any given map, we can construct its dual graph as
follows. Put a vertex inside each region of the map and connect two distinct
vertices by an edge if and only if their respective regions share a whole segment
of their boundaries in common. Then, a proper vertex coloring of the dual graph

yields a proper coloring of the regions of the original map.

Figure 7.1. The map of India

8. Pirzada and Ashay Dharwadker

Figure 7.2. The dual graph of the map of India

We use the vertex coloring algorithm [7] to find a proper coloring of the map of
India with four colors, see figures 7.1 and 7.2 above.

The Groupe Spécial Mobile (GSM) was created in 1982 to provide a

standard for a mobile telephone system. The first GSM network was launched in
1991 by Radiolinja in Finland with joint technical infrastructure maintenance
from Ericsson. Today, GSM is the most popular standard for mobile phones in the
world, used by over 2 billion people across more than 212 countries. GSM is a
cellular network with its entire geographical range divided into hexagonal cells.
Each cell has a communication tower which connects with mobile phones within
the cell. All mobile phones connect to the GSM network by searching for cells in
the immediate vicinity. GSM networks operate in only four different frequency

ranges. The reason why only four different frequencies suffice is clear: the map of

the cellular regions can be properly colored by using only four different colors!
So,

the vertex coloring algorithm may be used for assigning at most four different

frequencies for any GSM mobile phone network, see figure 7.2 below.

34
35

Applications of Graph Theory

Figure 7.3. The cells ofa GSM mobile phone network

8. KNIGHT’S TOURS

In 840 A.D., al-Adli [17], a renowned shatranj (chess) player of Baghdad is said
to have discovered the first re-entrant knight's tour, a sequence of moves that
takes the knight to each square on an 8x8 chessboard exactly once, returning to
the original square. Many other re-entrant knight's tours were subsequently
discovered but Euler [10] was the first mathematician to do a systematic analysis
in 1766, not only for the 8x8 chessboard, but for re-entrant knight's tours on the
general nxn chessboard. Given an nxn chessboard, define a knight's graph with a
vertex corresponding to each square of the chessboard and an edge connecting
vertex i with vertex / if and only if there is a legal knight's move from the
square
corresponding to vertex i to the square corresponding to vertex j. Thus, a re-
entrant knight's tour on the chessboard corresponds to a Hamiltonian circuit in the
knight's graph. The Hamiltonian circuit algorithm [9] [13] has been used to find

re-entrant knights tours on chessboards of various dimensions.

(1]

[2]

[3]

[4]

8. Pirzada and Ashay Dharwadker

Figure 8.1. A re-entrant knight’s tour on the 8x8 chessboard

REFERENCES

K. Appel and W. Haken, Every Planar Map is Four Colorable, Bull.

Amer. Math. Soc. 82 (1976) 711-712.

L. Babai, Some applications of graph contractions, J. Graph Theory, Vol.

1 (1977) 125-130.

E. Bertram and P. Horak, Some applications of graph theory to other

Parts of mathematics, The Mathematical Intelligencer (Springer-Verlag,
New York) (1999) 6-11.

J.A. Bondy and U.S.R. Murty, Graph Theory with Applications, 1976,

36
37

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Applications of Graph Theory

Elsevier Science Publishing Company Inc.

L. Caccetta and K. Vijayan, Applications of graph theory, Fourteenth

Australasian Conference on Combinatorial Mathematics and Computing
(Dunedin, 1986), Ars. Combin., Vol. 23 (1987) 21-77.

Ashay Dharwadker, The Vertex Cover Algorithm, 2006,

http://www.geocities.com/dharwadker/vertex_ cover

Ashay Dharwadker, The Vertex Coloring Algorithm, 2006,

http://www.geocities.com/dharwadker/vertex_coloring

Ashay Dharwadker, 4 New Proof of The Four Colour Theorem, 2000,

http://www. geocities.com/dharwadker

Ashay Dharwadker, A New Algorithm for finding Hamiltonian Circuits,

2004, http://www.geocities.com/dharwadker/hamilton

L. Euler, Solution d'une question curieuse qui ne paroit soumise a aucune

analyse, Mémoires de I'Académie Royale des Sciences et Belles Lettres de
Berlin, Année 1759 15, 310-337, 1766.

Eric Filiol, Edouard Franc, Alessandro Gubbioli, Benoit Moquet and

Guillaume Roblot, Combinatorial Optimisation of Worm Propagation on
an Unknown Network, Proc. World Acad. Science, Engineering and
Technology, Vol 23, August 2007.

http://www. waset.org/pwaset/v23/v23-68.pdf

K. Heinrich and P. Horak, Euler's theorem, Am. Math. Monthly, Vol. 101
(1994) 260.

R.M. Karp, Reducibility among combinatorial problems, Complexity of

Computer Computations, Plenum Press, 1972.

D. Konig, Theorie der endlichen und unendlichen graphen, Akademische

Verlagsgesllschaft, Leipzing (1936), reprinted by Chelsea, New York
(1950).

G. Lancia, V. Afna, S. Istrail, L. Lippert, and R. Schwartz, SNPs

Problems, Complexity and Algorithms, ESA 2002, LNCS 2161, pp. 182-
193, 2001. Springer-Verlag 2001.
[16]

[17]

[18]

[19]

[20]

[21]

8. Pirzada and Ashay Dharwadker

L. Lovasz, L. Pyber, D. J. A. Welsh and G. M. Ziegler, Combinatorics in

pure mathematics, in Handbook of Combinatorics, Elsevier Sciences

B. V., Amsterdam (1996).

H. J. R. Murray, A History of Chess, Oxford University Press, 1913.

Shariefuddin Pirzada and Ashay Dharwadker, Graph Theory, Orient
Longman and Universities Press of India, 2007.

F. S. Roberts, Graph theory and its applications to the problems of society,

CBMS-NSF Monograph 29, SIAM Publications, Philadelphia (1978).
J.P. Serre, Groupes Discretes, Extrait de I’ Annuaire du College de
France, Paris (1970).

R. Thomas, A combinatorial construction of a non-measurable set, Am.

Math, Monthly 92 (1985) 421-422.

S. Pirzada

Department of Mathematics
University of Kashmir
Srinagar 190006

India

[email protected]

http://www. geocities.com/dharwadker/pirzada

Ashay Dharwadker

H-501 Palam Vihar

District Gurgaon
Haryana 122017
India

[email protected]

http:/Avww. geocities.com/dharwadker

38