1 \documentclass[14]{article}
7 %\usepackage[T1]{fontenc}
8 %\usepackage[latin1]{inputenc}
10 \renewcommand{\labelenumii}{\labelenumi\arabic{enumii}}
11 %\titleformat*{\section}{\Large\bfseries}
13 %\title{Response to the reviewers of \bf "Perimeter-based Coverage Optimization to Improve Lifetime in Wireless Sensor Networks"}
14 %\author{Ali Kadhum Idrees, Karine Deschinkela, Michel Salomon and Raphael Couturier}
22 \vspace{-0.5cm}\hspace{-2cm}FEMTO-ST Institute, UMR 6714 CNRS
24 \hspace{-2cm}University Bourgogne Franche-Comt\'e
26 \hspace{-2cm}IUT Belfort-Montb\'eliard, BP 527, 90016 Belfort Cedex, France.
31 Detailed changes and addressed issues in the revision of the article
33 ``Multiround Distributed Lifetime Coverage Optimization \\
34 Protocol in Wireless Sensor Networks''\\
37 by Ali Kadhum Idrees, Karine Deschinkel, Michel Salomon, and Raph\"ael Couturier
42 Dear Editor and Reviewers,
44 First of all, we would like to thank you very much for your kind help to improve
45 our article named: `` Multiround Distributed Lifetime Coverage Optimization
46 Protocol in Wireless Sensor Networks
47 ''. We highly appreciate the detailed valuable
48 comments of the reviewers on our article. The suggestions are quite helpful for
49 us and we incorporate them in the revised article. We are happy to submit to you
50 a revised version that considers most of your remarks and suggestions to improve
51 the quality of our article.
53 As below, we would like to clarify some of the points raised by the reviewers
54 and we hope the reviewers and the editors will be satisfied by our responses to
55 the comments and the revision for the original manuscript.
59 \section*{Response to Reviewer No. 1 Comments}
61 The paper entitled "Multiround Distributed Lifetime Coverage Optimization
62 Protocol in Wireless Sensor Networks" introduces MuDiLCO, a distributed protocol
63 to enhance the use of WSN by splitting the network lifetime into periods, and by
64 breaking each sensing activity of a period into a series of rounds. A sensor
65 called the leader is elected, and based on local data, solves a Mixed Integer
66 Linear Program in order to schedule the activity of its neighbors over the
67 sensing rounds of the current period.
69 The contribution of this paper is interesting, but probably needs to be deepened
70 in order to reach the standards of a publication in Ad Hoc Networks.\\
73 \noindent\textcolor{black}{\textbf{MAJOR COMMENTS:}} \\
75 \noindent {\bf 1.} Page 6, Section 3.2
76 The author didn't explain how subregions are created. This is an important
77 point, as clustering may have a significant impact on solution quality. Not only
78 the size of the subregion should be discussed and analyzed, but also the
79 clustering strategy.\\
81 \textcolor{blue}{\textbf{\textsc{Answer:} In the study, we assume that
82 the deployment of sensors is almost uniform over the region. So we only
83 need to fix a regular division of the region into subregions to make the
84 problem tractable. The subdivision is made such that the number of hops
85 between any pairs of sensors inside a subregion is less than or equal
86 to~3. In particular, we discuss the number of subregions in......}}\\
89 \noindent {\bf 2.} Page 8
90 The objective function (5) of the Mixed Integer Linear Program appears to be
91 very questionable. Indeed overcoverage and undercoverage may compensate each
92 other, so the same objective value may represent two incomparable situations. It
93 seems that the semantic of the objective function is not well defined, as one
94 may wonder what exactly is (quantitatively speaking) the problem objective.
95 Coverage breach is obviously an issue in WSN, but why penalizing overcoverage? A
96 two-phase approach where breach is minimized first, and then overcoverage is
97 minimized would probably make more sense.
100 \textcolor{blue}{\textbf{\textsc{Answer:} As mentioned in the paper the integer program is based on the model proposed by () with some modifications. Their initial approach consisted in first finding the maximum coverage obtainable from available sensors to then use this information as input to the problem of minimizing the overcoverage. But this two-steps approach is time consuming. The originality of the model is to solve both objectives in a parallel fashion. Nevertheless the weights }}\\
104 \noindent {\bf 3.} Page 9
105 In the MILP formulation, it is possible that some point p is never covered at
106 all, which means that some part of the area to monitor may never be monitored by
107 the WSN. The authors are referred to "alpha-coverage to extend network lifetime
108 on wireless sensor networks", Optim. Lett. 7, No. 1, 157-172 (2013) by Gentilli
109 et al to enforce a constraint on the minimum coverage of each point. \\
111 \textcolor{blue}{\textbf{\textsc{Answer:} }}\\
114 \noindent {\bf 4.} Page 13
115 The criterion "Energy Consumption" is the average consumption per round. But the
116 duration of a round is a feature that can be arbitrarily set in the algorithm.
117 Computing the average energy consumption per unit of time over the network
118 lifetime would be better, as it is independent from the number and duration of
121 \textcolor{blue}{\textbf{\textsc{Answer :} }}
125 \noindent {\bf 5.} Page 15-18
126 Figures 2-6 mention four different versions of MuDiLCO. The performance of these
127 different versions should be analyzed with more details for each figure.
128 Alternatively, the authors may remove some versions of MuDiLCO if they do not
129 bring any valuable insight. \\
132 \textcolor{blue}{\textbf{\textsc{Answer :} }}
135 \noindent {\bf 6.} Page 19 (most major point)
136 The authors state that solving the Mixed Integer Linear Program is
137 time-consuming, and use this point for explaining why MuDiLCO-7 is not as
138 efficient as other versions. Why not using a heuristic (or a metaheuristic) for
139 addressing the problem instead of using an exact solver? A straightforward and
140 easily implementable idea to cut CPU time would be to return the best feasible
141 solution found by the solver after a given time threshold for example. Doing so
142 is very likely to save time and energy, and to improve the results of MuDiLCO-7
146 \textcolor{blue}{\textbf{\textsc{Answer :} }}
151 \noindent\textcolor{black}{\textbf{MINOR COMMENTS:}} \\
153 \noindent {\ding{90} Page 2
154 The paragraph that begins with "The remainder of the paper is organized as
155 follows" should mention the content of Section 5. } \\
157 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed. Section 5 is included as a subsection 4.4 within section 4. }}\\
159 \noindent {\ding{90} Page 3
160 The sentence "the centralized approaches usually suffer from the scalability
161 problem, making them less competitive as the network size increase" should
162 probably be tempered: heuristics and metaheuristics can handle very large and
163 centralized problems (even if exact approaches can't), and these approaches are
164 very popular in WSN. } \\
166 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}\\
168 \noindent {\ding{90} Page 5
169 "The choice of number and locations of primary points is the subject of another
170 study not presented here". The authors should provide at least one reference
171 about the aforementioned study. } \\
173 \textcolor{blue}{\textbf{\textsc{Answer:} }}\\
175 \noindent {\ding{90} Page 6, Figure 1:
176 All rounds seem to have the same duration. This should be stated explicitly, and
177 justified (in column generation based approaches, "rounds" to not have the same
180 \textcolor{blue}{\textbf{\textsc{Answer:} All rounds have the same duration. It is explicitly explained
181 in paragraph ... in section .... This assumption leads to an integer formulation of the optimization problem. The decision variables are binary variables, $X_{t,j}$ for the activation ($X_{t,j}=1$) or not ($X_{t,j}=0$) of the sensor $j$ during the round $t$. Column generation based approaches can be applied when the decision variables of the optimization problem are continuous. In this case the variables are the time intervals during which the sensors of a cover set (not necessarily disjoint) are active. The time intervals are not equal. Concerning the choice of round duration of equal length, it is correlated
182 with the types of applications, with the amount of initial energy in sensors
183 batteries, and also with the duration of the exchange phase. All
184 applications do not have the same Quality of Service requirements. In our
185 case, information exchange is executed every hour, but the length of the
186 sensing period could be reduced and adapted dynamically. On the one hand, a
187 small sensing period would allow the network to be more reliable but would
188 have higher communication costs. On the other hand, the choice of a long
189 duration may cause problems in case of nodes failure during the sensing
192 \noindent {\ding{90} Page 11 in Table 1
193 $W_\Theta$ should be replaced with $W_\theta$
196 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}.\\
198 \noindent {\ding{90} Page 12
199 Don't italicize "mW" in "equal to 02575 mW" } \\
201 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}.\\
203 \noindent {\ding{90} Page 14
204 ML is not defined (see the denominator of the large, unnumbered formula that
207 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}.\\
209 \noindent {\ding{90} Page 19 Section 6
211 "Sensing phase itself divided into T rounds" -> T should be italic. }
214 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}.\\
216 \noindent {\ding{90} Page 18 Section 5.5
217 The name of the solver used to solve the Mixed Integer Linear Program should be
220 \textcolor{blue}{\textbf{\textsc{Answer:} Right, fixed }}.\\
223 We are very grateful to the reviewers who, by their recommendations, allowed us
224 to improve the quality of our article.