Professional Documents
Culture Documents
Textbook Mathematical and Statistical Applications in Life Sciences and Engineering 1St Edition Avishek Adhikari Ebook All Chapter PDF
Textbook Mathematical and Statistical Applications in Life Sciences and Engineering 1St Edition Avishek Adhikari Ebook All Chapter PDF
https://textbookfull.com/product/mathematical-and-statistical-
applications-in-food-engineering-1st-edition-surajbhan-sevda-
editor/
https://textbookfull.com/product/mathematical-methods-in-
engineering-and-applied-sciences-mathematics-and-its-
applications-1st-edition-hemen-dutta-editor/
https://textbookfull.com/product/an-introduction-to-statistical-
analysis-in-research-with-applications-in-the-biological-and-
life-sciences-1st-edition-wiley/
https://textbookfull.com/product/mathematical-and-statistical-
methods-for-actuarial-sciences-and-finance-maf-2016-1st-edition-
marco-corazza/
Advanced Mathematical Techniques in Engineering
Sciences First Edition Davim
https://textbookfull.com/product/advanced-mathematical-
techniques-in-engineering-sciences-first-edition-davim/
https://textbookfull.com/product/recent-advances-in-mathematics-
for-engineering-mathematical-engineering-manufacturing-and-
management-sciences-1st-edition-mangey-ram-editor/
https://textbookfull.com/product/physical-sciences-and-
engineering-advances-in-life-sciences-and-oncology-a-wtec-global-
assessment-1st-edition-paul-janmey/
Mathematical
and Statistical
Applications in
Life Sciences and
Engineering
Mathematical and Statistical Applications in Life
Sciences and Engineering
Avishek Adhikari Mahima Ranjan Adhikari
•
123
Editors
Avishek Adhikari Yogendra Prasad Chaubey
Department of Pure Mathematics Department of Mathematics and Statistics
University of Calcutta Concordia University
Kolkata, West Bengal Montreal, QC
India Canada
v
vi Preface
of the present volume. We are thankful to all the contributors to this volume and
especially to the corresponding authors for their cooperation and timely submission
of the articles and revisions. We have derived substantial help in reviewing these
articles from the following reviewers to whom we owe our immense gratitude:
Jorge A. Achcar (Brazil), Koby Asubonteng (USA), Saria Awadalla (USA), M.
L. Lakhal-Chaieb (Canada), Ashish K. Chattopadhyay (India), Propser Donovon
(Canada), Sujit K. Ghosh (USA), David Hanagal (India), Tamanna Howlader
(Bangladesh), Abdulkadir Hussein (Canada), Nachimuthu Manickam (USA),
Manoel Santos-Neto (Brazil), Partha Sarathi Roy (Japan), Satyajit Roy (India),
Helton Saulo (Brazil), and Jingjing Wu (Canada).
We apologize for any omissions. We are also grateful to Springer for publishing
this volume and to all individuals who have extended their support and cooperation
in order to bring the project of IMBIC for publishing this commemorative volume
to fruition.
Part I Mathematics
1 Hole: An Emerging Character in the Story of Radio k-Coloring
Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Ushnish Sarkar and Avishek Adhikari
2 Robust Control of Stochastic Structures Using Minimum Norm
Quadratic Partial Eigenvalue Assignment Technique . . . . . . . . . . . 43
Kundan Goswami, Sonjoy Das and Biswa Nath Datta
3 Single-Time and Multi-Time Hamilton–Jacobi Theory Based on
Higher Order Lagrangians . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Savin Treanţă and Constantin Udrişte
Part II Statistics
4 On Wavelet-Based Methods for Noise Reduction of cDNA
Microarray Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Tamanna Howlader, S. M. Mahbubur Rahman
and Yogendra Prasad Chaubey
5 A Transformation for the Analysis of Unimodal Hazard Rate
Lifetimes Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Kobby Asubonteng, Govind S. Mudholkar and Alan Hutson
6 The Power M-Gaussian Distribution: An R-Symmetric Analog
of the Exponential-Power Distribution . . . . . . . . . . . . . . . . . . . . . . 141
Saria Salah Awadalla, Govind S. Mudholkar and Ziji Yu
7 Stochastic Volatility Models (SVM) in the Analysis of Drought
Periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Jorge Alberto Achcar, Roberto Molina de Souza
and Emílio Augusto Coelho-Barros
ix
x Contents
Mahima Ranjan Adhikari, Ph.D. is the Founder President of the Institute for Mathematics,
Bioinformatics, Information Technology and Computer Science (IMBIC), Kolkata, and former
Professor of Pure Mathematics at the University of Calcutta. He has published a number of papers
in several Indian and foreign journals including Proceedings of American Mathematical Society
and eight textbooks including Basic Modern Algebra with Applications (Springer) and Basic
Algebraic Topology and its Applications (Springer). Twelve students have already been awarded
Ph.D. degree under his guidance on various topics such as algebra, algebraic topology, category
theory, geometry, analysis, graph theory, knot theory, and history of mathematics. He is a member
of American Mathematical Society and on the editorial board of several Indian and foreign
journals and research monographs. He was elected as the president of the Mathematical Science
Section (including Statistics) of the 95th Indian Science Congress, 2008. He visited several
institutions in India, USA, UK, China, Japan, France, Greece, Sweden, Switzerland, Italy, and
many other countries on invitation.
xi
xii Editors and Contributors
Contributors
Jorge Alberto Achcar Medical School, University of São Paulo, Ribeirão Preto,
SP, Brazil; Department of Social Medicine, FMRP University of São Paulo, Monte
Alegre Ribeirão Preto, SP, Brazil
Jean-Francois Angers University of Montreal, Montreal, QC, Canada
Kobby Asubonteng AstraZeneca Pharmaceuticals, Gaithersburg, MD, USA
Saria Salah Awadalla Division of Epidemiology and Biostatistics, UIC School of
Public Health (SPH-PI), Chicago, IL, USA
N. Balakrishnan Department of Mathematics and Statistics, McMaster University
Hamilton, Hamilton, ON, Canada
Atanu Biswas Applied Statistics Unit, Indian Statistical Institute, Kolkata, India
Emílio Augusto Coelho-Barros Federal Technological University of Paraná,
Cornélio Procópio, PR, Brazil
Richard J. Cook Department of Statistics and Actuarial Science, University of
Waterloo, Waterloo, ON, Canada
Sonjoy Das Department of Mechanical and Aerospace Engineering, University at
Buffalo, Buffalo, NY, USA
Biswa Nath Datta Department of Mathematical Sciences, Northern Illinois
University, De Kalb, IL, USA
Isha Dewan Theoretical Statistics and Mathematics Unit, Indian Statistical
Institute, New Delhi, India
Jie Fan Division of Biostatistics, Department of Public Health Sciences, Miller
School of Medicine, University of Miami, Miami, FL, USA
Editors and Contributors xiii
1.1 Introduction
1.1.1 Motivation
the users in these cells are made through their corresponding cell area base stations.
Now the base stations transmit or receive any communication in specified frequencies
which are obtained by dividing the radio spectrum uniformly into disjoint frequency
bands. Suppose a radio receiver is tuned to a signal on channel c0 , broadcast by its
local transmitter (i.e., the closest one). The quality of reception will be degraded if
there is excessive interference from other nearby transmitters. Also, due to the reuse of
the same channel c0 in the vicinity, there will be “co-channel” interference. Moreover,
since in reality, as the signal energy from one channel spills over into its nearby
channel, neither transmitters nor receivers operate exclusively within the frequencies
of their assigned channel. Hence, stations using channels close to c0 are also potential
source of interference. Imposing constraints to channel separations between pair of
transmitters, potentially capable of interference, is a way to maintain signal quality
of the broadcast. Let d(x, y) be the distance between transmitter sites x and y in
some suitable metric and δc (x, y) be the minimum allowed spectral separation of the
channels assigned to sites x and y. Then for any two pairs of transmitter sites (x1 , y1 )
and (x2 , y2 ), if d(x1 , y1 ) ≥ d(x2 , y2 ), then δc (x1 , y1 ) ≤ δc (x2 , y2 ). Thus there is a
monotonic trade-off between distance and spectral separation. Moreover, the physical
assumptions made above imply that the spectral constraints are determined only by
distances, in the sense that if d(x1 , y1 ) = d(x2 , y2 ), then δc (x1 , y1 ) = δc (x2 , y2 ).
In other words, if the distance between two stations is small, the difference in their
assigned channels must be relatively large, whereas two stations at a larger distance
may be assigned channels with a smaller difference. Thus, if the channels assigned
to the stations x and y are f (x) and f (y), respectively, then
where l x y depends inversely on the distance between x and y. Also, the range of
frequencies, from which the assignment is to done, should be kept as narrow (and
hence economic) as possible keeping in mind the scarcity and dearness of the radio
spectrum.
Now the problem of channel assignment was first interpreted in a graph-theoretic
approach by Hale [19]. Here, we consider the base stations, i.e., the transceivers
in the network as vertices. Two vertices are adjacent if the corresponding stations
can broadcast or receive the transmission of each other, i.e., if their transmission
ranges intersect. Also, as the channels are uniformly spaced in the radio spectrum,
we can consider the channel assignment of networks as an integer coloring or integer
labeling problem of graphs.
Chartrand et al. [8, 9] have introduced the radio k-coloring problem of a simple
finite graph by taking l x y = k + 1 − d(x, y) in Eq. 1.1. Formally, for any positive
integer k ≥ 2, a radio k-coloring L of a finite simple graph G is a mapping L : V →
N ∪ {0} such that for any two vertices u, v in G,
First, we have to be acquainted with some important notions. Throughout this chapter,
graphs have been taken as simple and finite. The number of vertices of a graph is
its order. Let G be any graph. Then V (G) and E(G) denote the vertex set and
the edge set of G, respectively. Also, let L be any radio k-coloring of G. Then
(maxv∈V L(v) − min v∈V L(v)) is referred as the span of L, denoted by span(L),
and min L {span(L) : L is a radio k − coloring o f G} is referred as the radio k-
chromatic number of G, denoted by r ck (G). Interestingly, if L be a radio k-coloring
of G, then L + a is the same. Therefore, without loss of generality, we shall assume
min v∈V L(v) = 0, for any radio k-coloring L on G. Any radio k-coloring L on G
with span r ck (G) is referred as r ck (G)-coloring or simply r ck -coloring (when the
underlying graph is fixed).
So far, the radio k-coloring problem has been mostly studied for k = 2, k = 3,
k = diam(G) − 2, k = diam(G) − 1, k = diam(G). The radio k-coloring is
referred as radio coloring, antipodal coloring, and near-antipodal coloring, and the
corresponding radio k-chromatic numbers are known as radio number, antipodal
number, and near-antipodal number of G for k = diam(G), k = diam(G) − 1 and
k = diam(G) − 2, respectively, where diam(G) denotes the diameter of G, i.e., the
maximum distance between any two vertices of G. For k = 2, the problem becomes
the L(2, 1)-coloring problem introduced by Griggs and Yeh [18] and for k = 3,
the problem is often referred as L(3, 2, 1)-coloring. Note that the radio 2-chromatic
number r c2 (G) is also often denoted as λ2,1 (G) and termed as λ2,1 (G)-number or
simply λ2,1 -number. Also, r c2 -coloring of G is referred as λ2,1 -coloring of G. In
fact, in our subsequent discussion, we will mostly use λ2,1 (G) instead of r c2 (G).
For an r ck -coloring L on a graph G = (V, E), let L ik (G) = {v ∈ V |L(v) = i} and
li (G) = |L ik (G)|. We replace L ik (G) by L i and lik (G) by li if there is no confusion
k
2 8 4 14
4 6 8 11
1 2
Fig. 1.1 Example of graph with positive 1-hole index and 2-hole index
2 4
0 1
No hole rc 2−coloring
5
3
2
0
2
5
0 Set of holes={4}
3
1
1
L(2,1)−coloring
6
4
2 9
0 9
1 3
2 4 0
6
3 7
G1 G2 G3
as simply μ(G). Note that r ck (G) ≤ μk (G) and in particular, λ2,1 (G) ≤ μ(G). If
λ2,1 (G) = μ(G), then the graph G admits a λ2,1 -coloring with no hole. Such λ2,1 -
coloring of G is said to be no hole λ2,1 -coloring or consecutive λ2,1 -coloring of G
and the graph G is said to be a full colorable graph.
It has been shown in [13] that the following graphs are full colorable:
(i) any connected graph G with Δ(G) = 2 except P3 , C3 , C4 and C6 ;
(ii) any tree T with Δ(T ) ≥ 3 except stars;
(iii) any connected non-tree graph G of order at least six with Δ(G) = 3 and
λ2,1 (G) = 4 except the graphs in the Fig. 1.4;
(iv) any connected non-tree graph G of order either 7 or 8 with Δ(G) = 4 and
λ2,1 (G) = 5 except the graph in the Fig. 1.5.
Several interesting combinatorial properties of holes in a λ2,1 -coloring as well as
existence of full colorable graphs have been studied in [1, 13, 16, 30, 31]. Recently,
holes in a r ck -coloring has been studied in [40], and several interesting properties
have been explored. In fact, the study of holes in an r ck -coloring on a graph is an
emerging topic in literature. However, this problem is comparatively more explored
for r c2 -coloring, i.e., a λ2,1 -coloring of a graph than for general k.
We now define k-islands and using this notion, we will define an equivalence relation
on Λρk (G). These notions will be very helpful in our subsequent discussion. Based
on the fact that an r ck -coloring of a graph G may have at most (k − 1) consecutive
holes [39], an Ω-set and a k-island have been defined in [40]. Let L ∈ Λρk (G). An
Ω-set of L is a non-empty set A of non-negative integers, assigned by L as colors,
such that s ∈ A only if 0 ≤ |s − s | < k, for some s ∈ A. A maximal Ω-set of
L ∈ Λρk (G) is a k-island or simply an island of L, if there is no confusion regarding
1 Hole: An Emerging Character in the Story of Radio k-Coloring Problem 9
k. The minimum and maximum element of an island I are said to be the left and
right coast of I , respectively, denoted by lc(I ) and r c(I ) accordingly. The left and
right coasts of I are together called the coastal colors or coastal labels of I . If the
set of coastal colors of some island is a singleton set, then the island is said to be an
atoll.
An Equivalence Relation on Λρk (G):
First, we discuss a recoloring of an r ck -coloring in Λρk (G), referred as an
α-recoloring (see [40]), which eventually will induce an equivalence relation on
Λρk (G). Let L ∈ Λρk (G) and I0 , I1 , . . . , Iρk be the islands of L such that lc(I j+1 ) =
r c(I j ) + k, 0 ≤ j ≤ ρk − 1. For any i, 0 ≤ i ≤ ρk , and m, 0 ≤ m ≤ (ρk − i), let us
define a new radio k-coloring L̂ on G as follows:
L(u), if L(u) ∈/ i+m
s=i Is ,
L̂(u) = i+m
r c(Ii+m ) − t, if L(u) ∈ s=i Is and L(u) = lc(Ii ) + t, t ≥ 0.
It is easy to verify that L̂ ∈ Λρk (G) and the islands I0 , I1 , . . . , Iρk of L̂ are such that
I j = I j , for j ∈
/ {i, i +1, . . . , i +m} and r c(Ii+m−t ) = r c(Ii+m )−(lc(Ii+t )−lc(Ii )),
lc(Ii+m−t ) = r c(Ii+m ) − (r c(Ii+t ) − lc(Ii )), for 0 ≤ t ≤ m. This recoloring L̂ is an
α-recoloring of L.
Let η be a relation on Λρk (G) defined by L 1 ηL 2 if and only if L 2 is obtained from
L 1 using of finite α-recolorings. It is easy to verify that η is reflexive, symmetric,
and transitive. Thus, η is an equivalence relation on Λρk (G). Therefore, any r ck -
coloring in Λρk (G) is η-related to some L ∈ Λρk (G) with islands I0 , I1 , . . . , Iρk
such that lc(I j+1 ) = r c(I j ) + k, 0 ≤ j ≤ ρk − 1 and |I0 | ≤ |I1 | ≤ · · · ≤ |Iρk |.
Thus, we may assume, without loss of generality, that for any L ∈ Λρk (G), if the
islands of L are I0 , I1 , . . . , Iρk with lc(I j+1 ) = r c(I j ) + k, 0 ≤ j ≤ ρk − 1, then
|I0 | ≤ |I1 | ≤ · · · ≤ |Iρk |. We refer this finite sequence (|I0 |, |I1 |, . . . , |Iρk |) as the
k-island sequence of L or simply the island sequence of L, if there is no confusion
regarding k. For the graph G in Fig. 1.1, the λ2,1 -coloring in Λρ2 (G), i.e., Λρ (G), has
2-islands I0 = {0, 1, 2, 3, 4}, I1 = {6}, I2 = {8}, and the r c3 -coloring in Λρ3 (G) has
3-islands I0 = {0, 2, 4, 6, 8}, I1 = {11}, I2 = {14}. Hence, (1, 1, 5) is a 2-island
sequence as well as a 3-island sequence of G.
In this section, an upper bound of ρk (G) has been given. Moreover, various interesting
features regarding order, regularity (i.e., each vertex having the same degree), radio
k-chromatic number, domination number as well as perfect domination number of
G (defined later) can be found when ρk(G) attains its upper bound. We will start with
the following results which will give a structural insight of a graph G in presence of
a (k − 1)-hole of an r ck -coloring in Λρk (G). But first, we need this definition. Let G
10 U. Sarkar and A. Adhikari
and H are two graphs. Then G + H is the disjoint union of G and H having vertex
set V (G) ∪ V (H ) and edge set E(G) ∪ E(H ). Thus, mG is the disjoint union of m
pairwise disjoint copies of G.
Lemma 1.1 [40] Let G be a graph with ρk (G) ≥ 1 and L ∈ Λρk (G). If {i + 1, i +
2, . . . , i + k − 1} is a (k − 1)-hole in L, then li = li+k and the subgraph of G induced
by L i ∪ L i+k is li K 2 .
Proof Case I: Let li+k ≥ 2. If possible, let x ∈ L i+k be such that x v, where v is
an arbitrary vertex in L i . We define L̂ by
L(u), if u = x,
L̂(u) =
i + k − 1, if u = x.
The following result is going to play a key role in our subsequent discussion.
Corollary 1.1 [40] Let G be a graph with ρk (G) ≥ 1 and I, J are two distinct
islands of L ∈ Λρk (G) with x and y as two coastal colors of I and J , respectively.
Then l x = l y and the subgraph of G induced by L x ∪ L y is l x K 2 .
Theorem 1.1 [40] Let G be a graph with ρk (G) ≥ 1. Then ρk (G) ≤ Δ(G), where
Δ(G) is the maximum degree of any vertex in G.
Proof Let I0 , I1 , . . . , Iρk be the islands of any L ∈ Λρk (G). Let x be any coastal
color of I0 and u ∈ L x . Then by Corollary 1.1, u ∼ v y , where v y ∈ L y in G,
for each coastal color y of I j , 1 ≤ j ≤ ρk (G). Hence, d(u) ≥ ρk (G). Therefore,
ρk (G) ≤ Δ(G).
When ρk (G) attains its upper bound, G shows some interesting structural proper-
ties. Before we discuss the situation, let us take a look at few definitions. In a graph
1 Hole: An Emerging Character in the Story of Radio k-Coloring Problem 11
Theorem 1.3 [40] Let G be a graph of order n with ρk (G) ≥ 1 and ρk (G) =
Δ(G) = Δ. Then
Proof Let I0 , I1 , . . . , Iρk be the islands of any L ∈ Λρk (G) with lc(I j+1 ) = r c(I j ) +
k, 0 ≤ j ≤ ρk − 1 such that |I0 | ≤ |I1 | ≤ · · · ≤ |Iρk |. Then the number of islands of
L is ρk + 1.
We claim that every island of L is an atoll. If not, then there exists an island I
of L such that lc(I ) = r c(I ). Then by Corollary 1.1, for any vertex v with L(v)
as a coastal color of any other island J of L, d(v) ≥ ρk + 1 = Δ + 1, which is a
contradiction. Hence, every island is a singleton set, i.e., an atoll. Thus, for every
vertex v of G, L(v) is a coastal color of some island of L and therefore every vertex
has same degree, by Corollary 1.1. Hence, G is Δ-regular.
Since each island of L is atoll, I j = {k j}, 0 ≤ j ≤ Δ. Hence, r ck (G) = kΔ.
Moreover, by Corollary 1.1, lki = lk j = l (say), 0 ≤ i < j ≤ Δ. Therefore,
n = l(Δ + 1), i.e., n ≡ 0 (mod Δ + 1).
Again, by Corollary 1.1, the subgraph induced by L ki ∪ L k j is l K 2 , for any i, j
with 0 ≤ i < j ≤ Δ. Also, Ii = {ki}, 0 ≤ i ≤ Δ. Therefore, L ki is a dominating set
as well as a perfect dominating set of G, for every i, 0 ≤ i ≤ Δ. Hence, γ(G) ≤ Δ+1 n
.
Using Theorem 1.2, we get γ(G) = Δ+1 . Hence, γ p (G) = Δ+1 .
n n
Things become very interesting when ρ2 (G) = Δ(G), i.e., ρ(G) = Δ(G). By
Theorem 1.3, the order of G is t (Δ + 1), for some integer t ≥ 1. Let GΔ,t be the
family of connected graphs G with ρ(G) = Δ(G) = Δ and order t (Δ + 1), for some
integer t ≥ 1. Then using Theorem 1.3, li = t, for i = 0, 2, 4, . . . , λ2,1 (G) = 2Δ.
Note that GΔ,1 = {K Δ+1 }.
Lemma 1.2 [16] If G is a connected Δ-regular graph of order 2(Δ + 1), then
G ∈ GΔ,2 or λ2,1 (G) = 2Δ + 1.
vr,1 vs,2
Remark 1.1 The above result says about the radio 2-chromatic number of G if G is
a connected Δ-regular graph of order 2(Δ + 1) and G ∈/ GΔ,2 .
Theorem 1.5 [16] Let G be a connected Δ-regular graph of order 2(Δ + 1). Then
G ∈ GΔ,2 if and only if there is a set S ⊆ E(K Δ+1 ) such that G is isomorphic to
X S (K Δ+1 + K Δ+1 ).
We shall now address the more general issue of connected Δ-regular graphs having
order t (Δ + 1), for any t ≥ 2. Král’ et al. [24] identified a special property of
connected Δ-regular graph with order t (Δ + 1). A connected Δ-regular graph with
order t (Δ + 1) is said to have the t-neighborhood property if for any two disjoint
sets U and W of vertices of G, the following holds: If no vertex of U is adjacent to
any vertex of W and any pair of vertices in U or W are at least three distance apart,
then |U | + |W | ≤ t.
We now have the following result.
14 U. Sarkar and A. Adhikari
v/ v v
1 1 2
v v/ v v
3 1 1 2
L(v2 ) = 2i + 1, by Lemma 1.3. Then the vertices in the right component of Hm are
colored with 0, 2, 4, . . . , 2i, 2i +1, 2i +3, . . . , 2m −1. Therefore, {L(u 1 ), L(v1 )} ⊆
{1, 3, 5, . . . , 2i − 1, 2i + 2, 2i + 4, . . . , 2m − 2} as L satisfies the distance two
condition. Hence, |L(u 1 ) − L(v1 )| ≥ 2. Therefore, {L(u 1 ), L(v1 )} = {2 j, 2 j + 1},
for some j ∈ {0, 1, . . . , m − 1} and so the vertices in the left component are colored
by 0, 2, 4, . . . , 2 j, 2 j + 1, 2 j + 3, . . . , 2m − 1. But as i = j and L(u 2 ) = 2i,
L(v2 ) = 2i + 1, so this leads to a contradiction. Hence, λ2,1 (Hm ) = 2m.
We now construct a graph G m which consists of three components: one is K m+1 −
{e1 , e2 }, where e1 = u 1 v1 and e2 = u 1 v1 are two non-incident edges (called middle
component); the other two components are K m+1 − ei , where ei = u i vi , i = 2, 3
(called left and right components, respectively). Now, add edges u 1 u 2 , v1 v2 , u 1 u 3
and v1 v3 (See Fig. 1.8).
Lemma 1.6 [31] λ2,1 (G m ) = 2m, for m ≥ 3.
Proof Since Hm is a subgraph of G m , λ2,1 (G m ) ≥ 2m, by Lemmas 1.4 and 1.5.
Again, label the vertices u 1 , v1 , u 1 , v1 , u 2 , v2 , u 3 , v3 with 0, 2, 2m − 2, 2m, 2, 0, 2m,
2m − 2, respectively; the remaining (m − 1) vertices of the left (right, respectively)
component by distinct colors from the set {0, 2, . . . , 2m −4} ({4, 6, . . . , 2m}, respec-
tively). Lastly, label the remaining (m − 3) vertices in the middle component by the
distinct colors from {4, 6, . . . , 2m − 4}. This is an L(2, 1)-coloring of G m and hence
λ2,1 (G m ) = 2m.
Now we are in a position to discuss the following result.
Theorem 1.7 [31] For each integer m ≥ 3, there exists a connected m-regular graph
G of order 3(m + 1) without 3-neighborhood property such that λ2,1 (G) = 2m and
ρ(G) = m.
Sketch of proof. Clearly G m is a connected m-regular graph of order 3(Δ + 1).
By Lemma 1.6, λ2,1 (G m ) = 2m. Consider the set of vertices U = {u 2 , u 3 } and
W = {v2 , v3 } in G m . Clearly, each pair of vertices of U or W is three distance apart,
and no vertex of U is adjacent to any vertex of W . Moreover, |U | + |W | = 4 > 3.
Hence, G m does not satisfy the 3-neighborhood property.
Now, it can be shown (see [31]) that for any λ2,1 -coloring L of G m , if u, v are any
two vertices in the same component, i.e., in the middle or left or right component,
of G m , then L(u), L(v) cannot be consecutive integers. Hence, 1, 3, 5, . . . , 2m − 1
are holes in L. So, ρ(G m ) = m.
16 U. Sarkar and A. Adhikari
a0 a
1
a2
In this subsection, we study a conjecture regarding Gr,t and its disproof. But first we
need to understand the backdrop in which the conjecture shall find its relevance. For
this, we will study the following two families of connected r -regular graphs.
We begin with a class Ωr of connected r -regular graphs, for any r ≥ 1, introduced
in [16]. Let X = r K r and Y = r K 1 . Let V (Ωr ) = V (X ) ∪ V (Y ), where
−1
(i) V (X ) = ∪ri=0 Bi , Bi = {bi, j | 0 ≤ j ≤ r − 1} and (ii) V (Y ) =
{a0 , a1 , . . . , ar −1 }.
−1
Let E(Ωr ) = R ∪ S where (i) R = ∪ri=0 Ri , Ri = {bi, j bi,k : 0 ≤ j < k ≤ r − 1}
r −1
and (ii) S = ∪i=0 Si , Si = {ai bm,i : 0 ≤ m ≤ r − 1}. Clearly, Ωr is a connected
r -regular graph with order r (r + 1) (See Fig. 1.9).
It can be easily verified that Ω1 and Ω2 are isomorphic to K 2 and C6 , respectively.
We now observe the following facts regarding distances in Ωr .
For 0 ≤ i < j ≤ r − 1, d(ai , a j ) = 3.
For 0 ≤ i, j, k, l ≤ r − 1,
⎧
⎨ 1, if i = k and j = l,
d(bi, j , bk,l ) = 2, if i = k and j = l,
⎩
3, otherwise;
Also for 0 ≤ i, j, k ≤ r − 1,
1, if i = k,
d(ai , b j,k ) =
2, otherwise;
Using these facts, we can say that Ωr satisfies the r -neighborhood property.
Now let Bk = {bi, j | ( j − i) ≡ k (mod r )}, 0 ≤ k ≤ r − 1. Clearly, |Bk | = r
and each pair of vertices of Bk and Y are at distance three. Hence, we can assign an
L(2, 1)-coloring on Ωr as follows:
1 Hole: An Emerging Character in the Story of Radio k-Coloring Problem 17
b b b
0,0 b 1,0 1,1
0,1
b b0,2 b b
0,3 1,3 1,2
2k, if v ∈ Bk ,
L(v) =
2r, if v = ai , where i = 0, 1, . . . , r − 1.
"He had nearly died, and he was all she had," replied Mrs.
Evans. "Heir to a fine property, I believe. I scarcely think that
Mrs. Caruth could be the one Joyce mentioned. Was it likely
there would be any intimacy between the daughter of a poor
country clergyman and people of position like those Caruths?"
"But not always their wives and daughters," said Mrs. Evans.
"Did you notice the name of the place Joyce's friend lived at? I
have the address of those we met at Mentone; it was Ferns—
something—crag, probably."
"I can hardly think that," replied the girl; "I could not imagine
Joyce saying an untrue word. She is not that sort of girl. And,
mamma, she is my cousin and a lady, though she is not rich. I
cannot help feeling sorry for her. If these friends of hers
should turn out to be the Caruths we met, and at some future
time we should see them again, what will they think of us for
letting Joyce go?"
"Think, you foolish girl! What can they think? Just that as she
was too proud to go to Fernsclough, she was too headstrong
to be guided by us, and went her own wilful way. You need not
trouble your head about that."
CHAPTER IV.
THOUGH Joyce Mirlees' twenty-first birthday brought some
clouds and storms, it was not wholly without peace and
brightness. More than a dozen letters reached her from
various quarters. Her uncle did not forget Joyce, but wrote
warmly and lovingly, and promised to be at The Chase before
she left it.
Other letters were from old friends at Welton, who did not fail
to send birthday greetings and simple gifts to their former
pastor's daughter. One packet, containing some beautiful
fancy articles, came to her from her Sunday scholars, who had
worked them for the dear teacher whose absence they
regretted more and more, they said. Yet it was plain that one
and all pictured Joyce amongst loving kinsfolk, and amid
luxuries of every kind, for they seemed half afraid that their
simple tokens of love would look very poor and mean amongst
her birthday gifts in her new and splendid home.
There was one more letter not named hitherto, which, though
full of kindness, brought some disappointment. The writer,
Mrs. Caruth, said all that could be expected from an old and
true friend. But there was no other message, though she
mentioned casually that her son, being quite well, had
rejoined his regiment instead of availing himself of the longer
leave at his disposal.
It was still early evening, and Joyce was in her own room,
when she heard a light tap at the door, and the words, "May I
come in, Cousin Joyce?"
"Yes, my uncle has always been kind, and I have had Sarah
Keene. Besides, I was but a stranger who had to win the
affection of strangers, though they might be relatives; and I
really believe you care for me after all!" cried Joyce, looking
up into Adelaide's face, and smiling through the tears which
her cousin's words had brought to her eyes. "Forgive me,
Adelaide. I want forgiveness, too, for I have judged you rather
hardly, I am afraid."
"No, you have not; I have never been kind, but I want to be
now." And two pairs of arms went out, and two girls' lips met
for the first time in mutual affection and forgiveness. Then
they sat down side by side, each encircling the other with one
embracing arm.
"I should not have minded about working early and late if you
had wanted help and we had worked together," said Joyce. "If
any one here had been ill, I should have thought nothing too
much to do for them, night or day. Supposing that my uncle
had been poor, and had given me a home with his children, I
would have slaved for him and them most cheerfully, and
taken care that his kindness should have cost him nothing in
the end. But you are all rich, and every wish can be gratified;
and the thought of being sent to sew under the orders of
Russell was—"
"Do not ask me, dear; I cannot take it," said Joyce.
"Yes, and I will take a gift, too, and prize it. Spend ten
shillings on a little brooch in cut steel, and I will wear it, and
never part with it while I live. And give me your likeness; I
should like to have it, though I shall always picture your face
as it looks to-night."
"You shall have these trifles, Joyce, and I will keep this, no
matter how long, until you are willing to wear it." And
restoring the ring to its case she put it into her pocket. "Now
what else can I do for you?" she asked.
"My uncle breakfasts earlier than you and the rest do. I have
been used to pour out his coffee and join him at table. I think
he will miss me at first. Will you sometimes breakfast with
him?"
"How selfish I have been not to notice this, or care for his
loneliness! Rely on me, I will breakfast with him always,
unless by some special chance I have been up very late the
night before."
But she smiled as she said it, for she was glad to think that
the father and daughter would be brought together by her
own departure.
"Then what?"
"Her son came home for a time, and she had him, and I
became more of a companion to my father."
"I believe I have seen both Mrs. Caruth and her son. Does she
call him Alec?"
"He is very fond of his mother, and she of him," said Adelaide.
"Indeed, he seems a good, noble-minded man altogether.
Augusta thought there was no one like him during the eight
weeks we spent at Mentone."
"He neither troubled himself about her nor any other girl. I
mean so far as paying special attention went. He was
everything that was kind and courteous, but the elder ladies
and the children absorbed the larger share of his time—
somewhat, I think, to the disgust of the grown-up girls. If
hazarded a guess, it would be that he had no heart left to
give, and that he was far too noble and true a man to pay
unmeaning attentions, which could lead to nothing but regrets
and pain for another. I suppose he has no sister, or he would
be a model 'brother of girls.'"
"Not now, dear. But if ever the time should come when I can
be sure you all wish for me, or if I am needed by any, I will
return."
Adelaide was obliged to be satisfied with this. The girls parted
with expressions of affection and pledges of future friendship;
and Joyce laid her head on her pillow with a lighter heart than
she had done for months past.
Mr. Evans was expected home the evening before his niece
was to leave The Chase, but in place of him came a telegram
—
Joyce was to leave the station at nine, so her uncle would not
arrive till after she was gone. Mrs. Evans declined to see her,
but sent word that when Joyce came to her senses, and was
prepared to submit and acknowledge she had done wrong,
she might write and say so.
Moved still further by the new and better feelings just born in
her heart, Adelaide declined to drive with Mrs. Evans and
Augusta, and went instead to meet her father on his return at
noon.
"That first kiss was poor Cousin Joyce's," she said. "She left it
for you, and I promised to deliver it."
"Joyce's! She is surely not gone? I thought you would all have
joined to keep her until my return. My only sister's only child
to leave The Chase in such haste!"
"She could not stay. I tried hard to persuade her, for, papa, I
am sorry I have not been kinder to Joyce. We are friends now,
dear friends, and I hope we shall always be so. I cannot blame
Joyce for going. How could she stay? But you do not know all
yet. I trust things will turn out better than they seem to
promise. I think I ought to tell you all about Joyce's birthday
and what was said, only you must promise to say nothing to
mamma. I cannot help thinking she is a little sorry now, and
she is more likely to feel regret about Joyce's going if no one
speaks of it."
Then Adelaide told her father all that had passed, and Mr.
Evans listened, not altogether sadly, for his daughter made
the most of all that had been bright for Joyce on her birthday
—the loving letters and souvenirs from Welton, Mrs. Caruth's
consideration for her cousin's safe convoy, the opening of
hearts between themselves, and the new-born friendship,
which was to bind them more closely than the ties of
relationship had done.
There was much to cheer Mr. Evans in what he heard from his
daughter, and acting upon her suggestion, he made no
allusion to Joyce's departure. His silence was both a relief and
a reproach to his wife, who expected a scene, and was
conscious that, in spite of her desire to free herself from a
sense of responsibility, she could not even excuse herself for
her treatment of Joyce.
"Now another, papa," she said. "That is for Joyce. You must
give me one every morning for her, as I am her deputy."
CHAPTER V.
JOYCE reached the station nearest to Springfield Park at three
o'clock, having had a change of trains, and a stoppage of an
hour and a half on the road. Mrs. Caruth's maid, returning
direct to Fernsclough, would rejoin her mistress before six.
Her train would not, however, start for twenty minutes, so she
was able to tell her mistress that she had seen Miss Mirlees in
charge of a grey-haired coachman, who, with two little girls,
awaited her arrival.
"Are you the young person for Springfield Park?" asked the
man.
"No doubt Mrs. Ross feels that the children are safest with
you," said Joyce.
"Just so. She has had time to know what I am, for I drove her
when she was no bigger than the least of them, and I was in
her father's service. Now you step in next the eldest one—
Miss Mary. She should have been a boy by rights, but nobody
would like to change her for one now. Your things will be
brought by that lad, who has a trap close at hand. They are all
together, I suppose?"
"Come in," cried the elder child. "We wanted to see you, so
mamma let Price bring us. I am Mary, 'papa's Molly,' they call
me, and that is Alice. She turns her face away because she is
shy, but she will be friends soon. Mamma said we must be
very good and not make you sorry, because you have no
father and mother."
Tears sprang into Joyce's eyes, which the child noted instantly,
and her own face grew sorrowful.
"Why do you cry?" she said. "Let me kiss the tears away, as
mamma does mine, if I am only sorry, not naughty."
"There, that is right. Now look how pretty the park is, and see
the deer under the trees. They feed out of our hands, and
they will know you very soon, because you will be with us."
The little girls had been taken charge of by Mrs. Ross's maid,
Paterson, and the housekeeper told Joyce that when she was
ready she was to come to her own room for refreshment.
"Here are your boxes in good time," said Mrs. Powell; and
thus Joyce was able to make the needed change in her dress.
She was about to go down, when, recollecting her new
position, she turned back for one of the aprons, ironed so
carefully by Sarah Keene's hands, and over which, as badges
of coming servitude for her darling, she had shed many a tear.
"I always have an early cup of tea," said Mrs. Powell, "and I
thought it would be the best for you, along with something
more substantial, after a journey. Your future meals will be
taken upstairs with the children. Mrs. Ross will see you in the
morning; but she and the master are away—only for the day;
they will be back to-night. My mistress trusted you to me, and
I promised to make you comfortable," said Mrs. Powell, with a
look of great kindness in her motherly face.
To Joyce this was like having her old friend Sarah Keene by
her side, and she thanked the kindly housekeeper most
heartily and gratefully for her encouraging words.
But the tea was being neglected, and Mrs. Powell turned
Joyce's attention in that direction; so, impelled by a healthy
girlish appetite, she made a hearty meal, much to her new
friend's satisfaction.
One hour after she spent with the children, of whom, however,
she was not to take formal charge until the morning. Then the
housekeeper, being at leisure, showed her through the house
and a portion of the gardens, and finally left her to indulge in
happier thoughts than she could have imagined would be
possible to her under her new circumstances.
Joyce rose early and dressed the children, the little one having
overcome her shyness, and being now willing to make friends.
She was sitting, telling them a baby-story, when Mrs. Ross
entered the nursery after breakfast, and greeted her with the
utmost kindness.
"
Springfield Park, Sept. 6th."
"After three months, dear old nurse and friend,
I can say that I am glad I came here. Every one
is good to me; the children are so sweet that it is
delightful to work for them; and I do work,
Sarah."