IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART C:APPLICATIONS AND REVIEWS, VOL. 42, NO. 6, NOVEMBER 2012 1093
An Evolutionary Multiobjective Sleep-Scheduling
Scheme for Differentiated Coverage in Wireless
Sensor Networks
Soumyadip Sengupta, Swagatam Das, Member, IEEE, Md. Nasir, Athanasios V. Vasilakos, and Witold Pedrycz
Abstract—We propose an online, multiobjective optimization
(MO) algorithm to efficiently schedule the nodes of a wireless sen-
sor network (WSN) and to achieve maximum lifetime. Instead of
dealing with traditional grid or uniform coverage, we focus on the
differentiated or probabilistic coverage where different regions re-
quire different levels of sensing. The MO algorithm helps to attain
a better tradeoff among energy consumption, lifetime, and cover-
age. The algorithm can be run every time a node failure occurs
due to power failure of the node battery so that it may reschedule
the network. This scheduling is modeled as a combinatorial, multi-
objective, and constrained optimization problem with energy and
noncoverage as the two objectives. The basic evolutionary multi-
objective optimizer used is known as decomposition-based multi-
objective evolutionary algorithm (MOEA/D) which is modified by
integrating the concept of fuzzy Pareto dominance. The perfor-
mance of the resulting algorithm, which is called MOEA/DFD, is
compared with the performance of the original MOEA/D, which
is another very well known MO algorithm called nondominated
sorting genetic algorithm (NSGA-II), and an IBM optimization
software package called CPLEX. In all the tests, MOEA/DFD is
observed to outperform all other algorithms.
Index Terms—Density control, differentiated coverage, evo-
lutionary multiobjective optimization (MO), node deployment,
wireless sensor networks (WSNs).
I. INTRODUCTION
W
IRELESS sensor network (WSN) [1]–[3] is a spe-
cial kind of mobile network composed of hundreds or
even thousands of autonomous and compact devices, which are
called sensor nodes. The nodes can perform sensing, processing,
and wireless communication tasks. In a typical WSN applica-
tion, the source nodes collect data from a phenomenon and
Manuscript received August 13, 2011; revised December 22, 2011, and
February 20, 2012; accepted April 13, 2012. Date of current version October
12, 2012. This paper was recommended by Associate Editor J. Lazansky.
S. Sengupta and Md. Nasir are with the Department of Electronics and
Telecommunication Engineering, Jadavpur University, Kolkata 700032, India
(e-mail: senguptajuetce@gmail.com; nasir795@gmail.com).
S. Das is with the Electronics and Communication Sciences Unit, Indian
Statistical Institute, Kolkata 700108, India (e-mail: swagatam.das@isical.ac.in).
A. V. Vasilakos is with the Department of Computer and Telecommunication
Engineering, University of Western Macedonia, Kozani 50100, Greece (e-mail:
vasilako@ath.forthnet.gr).
W. Pedrycz is with the Department of Electrical and Computer Engineering,
University of Alberta, Edmonton, AB T6G 2R3, Canada, and also with the
Systems Research Institute of the Polish Academy of Sciences, 01-447 Warsaw,
Poland (e-mail: wpedrycz@ualberta.ca).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TSMCC.2012.2196996
disseminate them toward the sink node using multihop com-
munication. One of the important problems of any sensor node
design is the deployment of the sensor nodes [4] in the area to
be monitored. Sensor nodes usually have limited energy storage
along with low processing and communication capabilities [5].
Monitoring of the area may be based on uniform event detec-
tion or differentiated event detection, where the probability of
appearance of the event in the areas concerned varies both geo-
graphically and temporally. Finally, it is important to maintain
connectivity among the nodes so that the data collected by any
individual sensor node can flow through other nodes to the sink
node. Deployment of the sensor nodes taking all such objectives
together is a very challenging problem.
Both mathematical programming and evolutionary comput-
ing techniques have been applied over past few years to
achieve some of the objectives mentioned previously. Younis
and Akkaya [6] presented an overview of various strategies for
coverage problems. The main focus is often to determine an
optimal sensor placement to cover a grid area (sometimes under
uncertainty [7]) and minimize the cost or prolong the network
lifetime [8]. Peng et al. [9] studied the impact of sensor node
distribution on the network coverage for WSNs. An important
way to deal with coverage, connectivity, and lifetime maximiza-
tion is to deploy a densely distributed sensor network randomly
in area concerned and to use active and inactive nodes. After
certain period of time, active and inactive nodes are rearranged
to prolong the network lifetime maintaining connectivity and
coverage as described in [10] and [11].
In this study, we propose an online density control-based
sleep-scheduling method for lifetime maximization, energy
minimization, and coverage maximization, where coverage is
modeled as probabilistic event detection. First, we randomly de-
ploy a large number of nodes in a given area to be monitored and
establish the event probability distribution. The optimization al-
gorithm then schedules the active nodes. As the time progresses,
node failure may occur due to mechanical problems as well as
exhaustion of the battery. As each node failure occurs, the main
optimization algorithm is executed to reschedule the network
unless all the nodes are broken or the network connectivity is
lost or the nodes do not have sufficient energy to continue sens-
ing and communication. We use a multiobjective framework to
achieve energy minimization and coverage maximization with
connectivity as constraint. Every time a decision maker is imple-
mented to choose the solution with minimum energy and with
a fixed coverage threshold. Maximizing coverage means that
the nodes must be placed far apart from the sink node, which
1094-6977/$31.00 © 2012 IEEE