International Journal of Informatics and Communication Technology (IJ-ICT)

Vol.10, No.3, December 2021, pp. 159~170

ISSN: 2252-8776, DOI: 10.11591/ijict.v10i3.pp159-170  159

An efficient coverage and maximization of network lifetime in

wireless sensor networks through metaheuristics

A. Nageswar Rao, B. Rajendra Naik, L. Nirmala Devi

Department of Electronics and Communication Engineering, Osmania University, India

Article Info ABSTRACT

Article history: In wireless sensor networks (WSNs), energy, connectivity, and coverage are
the three most important constraints for guaranteed data forwarding from
Received Feb 17, 2021 every sensor node to the base station. Due to continuous sensing and
Revised Sep 14, 2021 transmission tasks, the sensor nodes deplete more quickly and hence they
Accepted Oct 11, 2021 seek the help of data forwarding nodes, called relay nodes. However, for a
given set of sensor nodes, finding optimal locations to place relay nodes is a
very challenging problem. Moreover, from the earlier studies, the relay node
Keywords: placement is defined as a non-deterministic polynomial tree hard (NP-Hard)
problem. To solve this problem, we propose a multi-objective firefly
Energy consumption algorithm-based relay node placement (MOFF-RNP) to deploy an optimal
Energy harvesting number of relay nodes while considering connectivity, coverage, and energy
Network lifetime constraints. To achieve network lifetime, this work adopted energy
Relay node placement harvesting capabilities to the sensor nodes and backup relay strategy such
Wireless sensor network that every sensor node is always connected to at least one relay to forward
the data. The optimal relay placement is formulated as an objective function
and MOFF is applied to achieve a better solution. Extensive Simulations are
carried out over the proposed model to validate the performance and the
obtained results are compared with state-of-art methods).
This is an open access article under the CC BY-SA license.

Corresponding Author:
L. Nirmala Devi
Department of Electronics and Communication Engineering, University College of Engineering
Osmania University, Hyderabad, India
Email: [email protected]

1. INTRODUCTION
In this paper, we have from past few years, wireless sensor networks (WSNs) have gained a huge
interest in various fields like Industrial monitoring, home automation, forest and fire detection, environmental
control, intensive agriculture, and among others [1]. Generally, the WSNs are composed of a group of sensor
nodes (SNs) and a sink node (also can be called as base station). In WSNs, the main responsibility of SNs is
to capture the information about the environment and the responsibility of sink is to accumulate the entire
data from SNs. The SNs have some prominent features by which the WSN technology has gained a huge
demand. For example, the SNs are cheap, small, autonomous power enabled and can capture several kinds of
dataeven with same node. Mainly, the utilization of wireless technology has facilitated the network topology
organization with very much less cost. These features have allowed the WSNs to deploy in environments
where the wired technology is highly expensive or almost impossible [2].
Generally, the sensor nodes are battery operated devices which has limited lifespan due to the
limited battery capacity. The entire information collected by SNs is forwarded to the sink node, consumes the
energy resources of SNs. Moreover, this sensitive behavior of sensor nodes will affect the network
performance. For example, if the topology of a network is assumedas start topology, and then all the senor

Journal homepage: http://ijict.iaescore.com

160  ISSN: 2252-8776

nodes are assumed to have an equal energy distribution which is a very challenging issue. Even though
various energy efficient mechanisms are derived, eventually the battery will drain out after a particular span
of time [3]. A promising solution to this problem is making the SNs to harvest the required energy through
some super capacitors and energy harvesters [4]. Instead of providing the required energy from external
power supplies like batteries, if the sensor node’s power supply unit is replaced with a renewable power
generation unit (ex. solar panels), an increased network lifetime can be observed. With this strategy, the
network lifetime can be increased up to some extent but the network demise is not avoidable, because the
sensor node has to harvest energy in the doze off period. In the doze off period, the sensor node will harvest
energy required for both information capturing and information transmission. Moreover, the time SN node to
harvest energy is more than the time taken for depletion of energy. In such scenario, the SN is disconnected
with other nodes, which makes the network to get demise. Hence there is a need to keep alternative nodes or
back up nodes which takes the responsibility of doze off nodes which prevents the network from demising.
In WSNs, the SNs not only monitor and capture the information from environment, but also help to
other senor nodes to forward the sensed data to sink node, which can be termed as relaying [5]. This type of
multiple responsibilities (sensing and relaying) makes the sensor nodes to deplete more quickly. Generally,
after sensing the information, sensor node goes into sleep mode to minimize the energy consumption, but due
to the additional task of relying, they need to be awakened for all time. As a result, the senor nodes face
additional computational tasks which results in faster energy depletion. Once the residual energy (RE) left at
a SN is less than the energy threshold, then it will stop all tasks and moves into harvesting mode. Hence there
is a need of responsibility distribution in which the entire nodes are grouped into two groups such as source
senor nodes (SSNs) and relay nodes (RNs). The main responsibility of SSNs is to sense the information and
the responsibility of RN is to forward the sensed data to sink node. However, the problem is to
determineaproper number of RNs through which both network coverage and connectivity can be achieved
along with less energy consumption.
From past few years, relay node placement (RNP) has gained a noteworthy research interest in the
field of WSNs due to its effectiveness in achieving an improved network lifetime. In [6], the authors focused
on two different deployments such as grid-based deployment and random deployment. In random
deployment, the nodes are placed in a random manner and are structured in an AdHoc manner whereas in the
grid deployment, the nodes are placed at the apexes of grid. Compared to the random deployment, the grid-
based deployment has achieved more accurate positioning. Likewise, in [7], [8], a grid based RNP is modeled
to connect the disjointed WSN segments by dividing the total area into equal-sized cells. The network size is
optimized by the selection of an optimal cell count to be distributed by RNs such that all the segments are
linked. However, this is considered as a “non-deterministic polynomial tree hard (NP-hard)” problem.
Generally, to make the NP-Hard problem as a more realistic one, a two-layer procedure is employed [9],
[10]. In these methods, the first layer is occupied by sensors in which they have to collect the information and
transmit to the cluster head (CH) or RN. In this manner, the SNs will gain a less energy consumption rate as
they will move into sleep mode immediately after the completion of data transmission to CH.
Recently, artificial intelligence has also been used for the optimization of network lifetime
throughthe placement of optimal number of relay nodes at optimal locations. “genetic algorithm (GA)”,
“particle swarm optimization (PSO)”, “artificial bee colony (ABC)”, Firefly, “ant colony optimization
(ACO)” etc., are some of the most common meaheuristic algorithms employed for the optimization problem.
Khosrowshahi and Shakeri [11] addressed the RNP in a multiple disjoint network through GA. In this
method, initially an upper bound of relay nodes is measured to set up the initial chromosome length. Next,
the GA iteratively reduced the RNs count and discovers the optimal locations at the same time. Next, an
improved version of GA, called as “genetic simulated annealing hybrid algorithm (GA-H-SA)” is considered
by Yang et al. [12] to solve the relay node cover problem. This paper mainly focused over the optimization
of three aspect such as energy consumption, number of relay nodes and connectivity. Next, focusing over the
k-connectivity of the sensor nodes, Gupta et al. [13] developed two independent algorithms for RNP in which
every SN will maintain at least kRNs. The first algorithm is based on GA and another one is based on Greedy
mechanism. However, this approach focused over only the connectivity but not focused about the problem
when the energy of a RN is depleted. George and Sharma [14] considered a modified version of GA for RNP
in WSNs by following a “constrained RNP problem (CRNPP)”to reduce the RNs count while providing
maximum connectivity.
Dandekar and Deshmuk [15] also focused on thek-connectivity of the sensor nodes and
accomplished PSOthat places optimalRNs to achieve a required connectivity between the SNs of a
homogeneous WSN. In this approach, the homogeneity is adopted by assuming that all the SNshave same
range of communication. Xu et al. [16], the RNP problem is articulated as “steiner tree problem with
minimum steiner points and bounded edge length (STP-MSPBEL)” which is NP-hard. Here, a variable meta-
heuristic based PSO called as “multi-space (MSPSO)”is proposed to attain an optimal number of RNs.

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170
Int J Inf & Commun Technol ISSN: 2252-8776  161

In addition, some authors employed ABC algorithm also to solve the RNP problem. These
approaches simultaneously focused over the Q-coverage and K-coverage scenario by which the network life
and coverage are enhanced simultaneously [17]-[18]. Further, in [19], Liu and He aimed at the maximization
of connectivity and minimization of cost incurred, and proposed a hybrid optimization algorithm called,
ACO-Greedy. This approach especially focused on the Grid based networks. This approach is based on the
ACO-Greedy through which the communication or sensing radius is dynamically adjusted to alleviate the
energy hole issue and also prolong the Network Lifetime. Next, to lessen the deployment cost and also to
recover the partitioned WSN, Sentukara et al. [20] proposed two distributed RNP methods based on Game
theory and Virtual Force based relay movements.
Next, by combining the advantages of both ABC and PSO, an optimal RNP method is proposed by
Mini et al. [21] which achieve an enhanced network lifetime with a pre-specified sensing range. Further,
Lanza and Pulido [22] combined the ABC with Firefly algorithm to solve the RNP problem through multiple
objectives. In this approach, totally three objectives such as network reliability (NR), average sensitivity Area
(ASA) and average energy consumption (AEC) are considered and an optimized solution is derived through
multi-objective ABC and multi-objective firefly algorithms. Unlike these methods which focus on theRNs
placement, anenhanced RNs deployment method is proposed by Hamim et al. [23] based on ABC algorithm.
This approach mainly focused on the extension of network lifetime with the deployment of optimal count
RNs. This method has main focus on the extension of network lifetime through the optimization of network
related parameters those are relatedto the constrained RNP problem.
Deployment of the SNs with energy harvesting capabilities is a new research direction in WSN
which has more efficiency in network survivability and sustainability. Based on this inspiration, Misra et al.
[24] focused to deploy minimum RNs at constrained locations, by guaranteeing that the RNs can harvest a
huge volume of energy. For both issues, this approach had proven that the problem is NP-hard and solved
through an algorithm based on polynomial time approximation. Next, as an extension, a “unified mixed
integer linear program (MILP)” based RNP is developed by Misra et al. [25] to measure the lower range for
the optimal solution of minimum RNP.
In this paper, we have proposed a new relay node placement strategy based on multiple objectives,
called as multi-objective-oriented relay node placement (MORNP). MORNP totally considers three
objectives such as energy, connectivity and coverage and the nodes which satisfy all these objectives are only
chosen as relay nodes. The energy constraint is derived based on the novel relationship between energy
harvesting rate and depletion rates of relay nodes. Next, the connectivity is ensured through the availability of
paths to the sink node. The coverage is ensured based on the Euclidean distance between SSNs and RNs.
Finally, this work also proposed a multi-objective firefly (MOFF) Algorithm to determine the optimal
number of RN. Simulations are conducted over the proposed MORNP and the performance is measured
through several performance metrics. Remaining paper is ordered as; the details of proposed MORNP are
explored in section 2. Section 3 explores the details of simulation experiments and performance evaluation.
Finally, the concluding remarks are given in section 4.

2. RESEARCH METHOD
2.1. Overview
In this paper, we have developed a novel optimization technique which reduces the number of relay
nodes required to maintain an efficient connectivity and coverage in the WSN. This technique is a multi-
objective-oriented relay node placement (MORNP) strategy based on firefly algorithm. Under this technique,
we have considered three objectives such as energy, connectivity and coverage and accomplished multi-
objective firefly algorithm to attain an optimal count of RNs. Moreover, this method also considered the
problem of energy harvesting during the RNP. Under the energy harvesting problem, the RNs are assumed to
have energy harvesting capabilities and for any source sensor node. This method derives at least one
connected relay to the sink node during the doze off period of remaining relay nodes. Here, to ensure the
energy constraint, we have developed a relation based on the harvesting rate and depletion rate of a relay
node. Next, the coverage constraint is ensured based on the Euclidean distance between SSNs and RNs.
Finally, the connectivity constraint is ensured by maintaining at least one path between any RN and sink
node. Further the complete details of network model, energy constraint, connectivity constraint, coverage
constraint, firefly algorithm are deliberated in the subsequent subsections.

2.2. Network model

In this paper, we have assumed a randomly deployed network with N number of SNs and only one
sink B. Next, the SNs and RNs range of communication is assumed as r whereas for sink node, it is
considered as R, where R>>r. In this model, the RNs are assumed to have energy harvesting capabilities.

An efficent coverage and maximization of network lifetime in WSN through … (A. Nageswar Rao)
162  ISSN: 2252-8776

During the data transmission, after particular span of time, if the energy level of any RN is reduced below the
energy threshold, then that RN doze off (turns off all tasks such as transmitting and receiving) and energy
harvesting starts. The RN will become active only after gaining sufficient amount of energy, called as
Activation energy 𝐴𝐸 . In this model, we have supposed that the SSNs and sink node placementis known to a
prior. Further assumed that the process of energy harvesting is a stochastic one in which the rate of
harvesting is varied with environmental conditions and hence it can also model as Spatio-Temporal Process.
In the temporal model, the gain varies with different time instances where as in the spatial model, the gain is
varied with different locations. Moreover, the energy harvesting rate is less than the energy depletion rate.
We have consideredonly one single sink node and its location is purely randomized. Further the
locations of SNs are also randomized over a spatial location and let’s let it be S. For any source node, if the
sink node found to be located within its range of communication, then it forwards the sensed data directly.
Otherwise, the source node seeks the help of any RN and forwards the sensed data through that RN. Two
nodes A and B located at locations i and j can communicate only if the Euclidean distance between them is
less than or equal to communication ranger as (1).

𝐸𝐷(𝐴, 𝐵) ≤ 𝑟 (1)

Where ED is the Euclidean distance and r is the range of communication. Simply the (1) states that
the RNlocated at position j can help to the source node located at position i, to forward the data packets on
satisfying that they have located at a distance which is less or equal to communication range r of node A.
Next, the sink node is supposed to have an infinity power and it is associatedwith main power supply.

2.3. Energy constraint

For a WSN with larger network area, usually the most of the SNs can’t lie inside the communication
range of sink node. Hence the SNs not only perform the sensing task and also executes relaying task [5].
Under this, the sensor nodes work as relays and helps to the source sensor nodes to forward the data packets
to the sink node. Due to these multiple responsibilities, all sensor nodes need to be awakening for almost all
the time which results in faster energy depletion [29]. Hence the energy levels of SNs are dropped into the
level below the energy threshold, which makes the nodes to drive into the mode of energy harvesting to
procure sufficient energy. Thus, making the SNs to have energy harvesting capabilities can enhance the
lifetime of network.
Proposed to develop an energy constraint assisted RNP. Here the energy constraint is defined with
respect to two parameters such as harvesting rate and depletion rate. As already discussed, that the harvesting
rate is always less than the depletion rate, i.e., the time taken for procuring the sufficient energy through
energy harvesting is greater than the time taken for depleting the energy. This is due to the issue of both tasks
such as transmitting and receiving. Compared to the depletion rate of relay node, the depletion rate of a
source node is less due to the single responsibility. In this work, for energy harvesting model, we have
followed the features of energy harvesting sensors [26] in which the sensor node will active only after
gaining sufficient amount of energy. After gaining such amount of energy, it can be activation energy 𝐴𝐸 , the
relay node can spend and harvest simultaneously. Hence there exists a relationship between harvesting rate
and depletion rate. Let 𝑑𝑟 be the depletion rate and ℎ𝑟 be the harvesting rate, a relay node is formulated into
three modes based on the relationship between 𝑑𝑟 and ℎ𝑟 , as (2);

𝑑𝑟 = ℎ𝑟 , 𝑁𝑒𝑢𝑡𝑟𝑎𝑙 𝑀𝑜𝑑𝑒 (𝑁𝑚 )

𝑅𝑚 = { 𝑑𝑟 < ℎ𝑟 , 𝑆𝑎𝑣𝑖𝑛𝑔 𝑀𝑜𝑑𝑒 (𝑆𝑚 ) (2)
𝑑𝑟 > ℎ𝑟 , 𝐷𝑒𝑝𝑙𝑡𝑖𝑛𝑔 𝑀𝑜𝑑𝑒 (𝐷𝑚 )

Where 𝑅𝑚 is the relay mode.

Let𝑅𝐸 be the residual energy left at a RN after particular span of time t, if it is less than the energy
threshold 𝑇𝐸 , then it stops all the communications and switches to the harvesting mode. During this mode the
transceiver of RN is turned off. Once the RN has turned off the transceiver, the SSN cannot forward the data,
by which the network becomes disconnected. Hence there is a need of backup RNs to perform the
responsibilities of current disconnectedRN. However, the problem is to discover the locations at where the
backup RNshave to place. One possible solution is to keep the backup RNs at the location approximately
nearer to the departed RNs. However, it is not a viableanswer because there is no knowledge about the rate of
harvesting of a particular location. Moreover, locating the backup RN at the same location of departed RN
may causes serious effects like physical destruction, location damage, barriers, and shades. Moreover, the
inaccurate prediction about the energy availability causes a serious effect over the lifetime of network, due to

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170
Int J Inf & Commun Technol ISSN: 2252-8776  163

the unpredictable relationship between rat of depleting and rate of harvesting. As deliberated above, the
energy harvesting from natural resources like wind, sun, heat and vibration is totally uncontrollable and
unreliable. Hence the energy harvesting process can be modeled as Spatio-temporal process [27]. Further the
energy harvesting process follows a periodic pattern [28] as the maximum energy is harvested in the day time
and minimum energy is harvested in the night time. All these constraints make the RNP highly unstable.
Hence this paper addresses the backup RNs which are independent of energy harvesting constraints and also
to the energy obtainability of RNs in the future.
Under the energy constraint modelling, we have developed an efficient backup RN deployment
method through which every SSN has an availability of ‘z’ backup relays in its neighborhood. The constraint
to pick up the backup relay node is modeled as (3) and (4).
𝑄
𝐴𝐸 1 𝑅𝐸(𝑞)
𝑃 ≤ ∑𝑄∈𝑸 𝑄 𝑄 (3)
ℎ𝑟(𝑝) 𝑄 𝑑𝑟(𝑞) −ℎ𝑟(𝑞)

and
𝑝
𝑅𝐸(𝑝)
𝑝 = 𝑎𝑟𝑔 𝑚𝑖𝑛 ( 𝑃 −ℎ𝑃 ) , ∀ 𝑝 ∈ 𝑃 𝑎𝑛𝑑 𝑞 ∈ 𝑄 (4)
𝑝 𝑑𝑟(𝑝) 𝑟(𝑝)

𝑃 𝑃
Where ℎ𝑟(𝑝) is the harvesting rate of a relay node 𝑝 ∈ 𝑃, 𝑑𝑟(𝑝) is the depletion rate of a relay node
𝑄 𝑄
𝑝 ∈ 𝑃, 𝑑𝑟(𝑞) is the depletion rate of a backup relay node 𝑞 ∈ 𝑄, ℎ𝑟(𝑞) is the harvesting rate of a backup relay
𝑄
node 𝑞 ∈ 𝑄 and 𝑅𝐸(𝑞) is the residual energy of the backup relay node 𝑞 ∈ 𝑄. The constraint is (3) declares
that the time consumed to gain a sufficient amount of energy such that the RN will get activeshould be less
than the average provision time of short-lived RNs from all the remaining sets of RNs. Further, the constraint
shown in (4) finds the short-lived RNs based on residual energy, harvesting rate and depletion rate. The
shortest-lived RN 𝑝 ∈ 𝑃 must harvest the enough amount of energy 𝐴𝐸 to get activate in the time the
shortest-lived RNs of other relay sets 𝑞 ∈ 𝑄 get depleted. Thus, the SSNs would have always a RN to
transfer the data to sink node. Through this constraint, the source sensor node keep connected to at least one
relay node even though the ‘z-1’ relay nodes are moved to harvesting mode.

2.4. Connectivity constraint

In the WSN, most of the sensor nodes have no direct communication with sink. Hence, they will
depend on the relay nodes to forward the sensed data. In the conventional relay node placement, if any relay
node is disconnected due to reasons like limited energy, out of range communication, buffer overflow etc.,
the source node will get disconnected from the network. Hence the relay node needs to be positioned in such
a manner that all the SSNs can found a communication link either through one relay or through set of relays.
RN needs to be placed based on the several constraints and if not the deploying cost and maintenance
overhead will be too high. Hence there is a need of an optimal RNs deployment to guarantee the connectivity.
The major hurdle in the deployment is the total number of RNs to be positioned such that the network can
achieve maximum connectivity. If less relay nodes are placed then there is a problem of connectivity, means
all source sensor nodes are not covered. On the other hand, if a greater number of RNs are located, then there
will be a problem of high deployment cost and also the huge maintenance overhead.
Hence, we developed an optimal RNP strategy which ensures a maximum connectivity with optimal
number of RNs. For a given set of relays, the proposed method first discovers the possible pathsthen the relay
node checks whether the sink is within the communication range or not. If it is found that the sink node is
within the communication range, then the relay node connects to the sink directly. On the other hand, if the
sink is found that it is not in the communication range butis in the communication range of neighbor relay
node, then the current ray node establishes a link to the sink through the neighbor relay node. For instance, if
we consider the path 𝑖 → 𝑗 → 𝐵, here i is the current relay node, j is the neighbor relay node and B is sink. To
ensure the connectivity there will be at least one relay node which have a path to the sink either directly or
through forwarding relays. The connectivity constraint is formulated as (5).

𝑝𝑎𝑡ℎ = 𝐶𝑟 → 𝑁𝑟 → 𝐵, ∀ 𝐶𝑟 ∈ 𝑃 (5)

Where 𝐶𝑟 is the current relay node,𝑁𝑟 is the neighbor relay node and B is the sink.
Note: to establish a connection between 𝐶𝑟 and 𝑁𝑟 , the Euclidean distance between 𝐶𝑟 and 𝑁𝑟 must be less
than or equal to the communication range of each other.

An efficent coverage and maximization of network lifetime in WSN through … (A. Nageswar Rao)
164  ISSN: 2252-8776

2.5. Coverage constraint

The coverage constraint is defined as the maximum number of SNs those were covered by one RN.
As the number of SNs increases covered by a RN, the total number of RNs to get deploy will get minimized.
Here the coverage constraint is evaluated based on the Euclidean distance between the SNs and RNs. For a
SSN,𝑠 ∈ 𝑆 and RN,𝑝 ∈ 𝑃, the coverage constraint is defined as (6);

𝐸𝐷(𝑠, 𝑝) ≤ 𝑟, ∀ 𝑠 ∈ 𝑆 𝑎𝑛𝑑 𝑝 ∈ 𝑃 (6)

The source sensor node 𝑠 can seek the help of a RN,𝑝 when it is within the communications range.
The coverage constraint shown in (6) denotes that every SN,𝑠 ∈ 𝑆 must be covered by at least one RN from
each of ‘z’ relay sets. The demonstration of connectivity and coverage constraint is depicted as Figure 1.

Figure 1. Connectivity and coverage constraint

As exposed in the Figure 1, the RN,1, 𝑅1 covers four SSNs such as 𝑆1 , 𝑆2 , 𝑆3 and 𝑆4 . Next the RN,2,
𝑅2 covers five SSNs such as 𝑆1 , 𝑆2 , 𝑆3 , 𝑆4 , 𝑆5 , 𝑆6 and 𝑆7 .And, the RN,3, 𝑅3 covers three SSNs such as 𝑆5 ,
𝑆6 and 𝑆7 . RN,4, 𝑅4 covers two SSNs such as 𝑆2 , and 𝑆3 . Finally, the RN,5, 𝑅5 covers three SSNs such as 𝑆4 ,
𝑆5 , and 𝑆6 . We noticed that the 𝑅2 has maximum coverage and if we consider the 𝑅1 and 𝑅2 , then the total
number of RNs required to cover the entire set of SSNsare 2. And, for every SSN there is an alternative relay
node, for example, if SSN,3, 𝑆3 is disconnected with 𝑅1 , then the responsibility of data forwarding of 𝑅1 is
done with 𝑅2 . Similarly, if the 𝑆5 is disconnected with 𝑅2 , then the responsibility of 𝑅2 is taken by 𝑅3
because the 𝑆5 is simultaneously connected with 𝑅3 . Hence the optimal number of RN sare 3, i.e.,𝑅1 , 𝑅2 and
𝑅3 .

2.6. Optimization of RNP by MOFF

In order to optimize the RNP, we have adopted MOFF. Since the firefly (FF) is simple metaheuristic
algorithm with less computational complexity, we have considered it. With optimal number of RNs, the
maximum connectivity is achieved with less energy consumption. Through the optimization, we will get an
optimal number of RNs through which we can achieve maximum connectivity along with less energy
consumption.
The FF is developed by Yang [30] based on the inspiration of fireflies idealized behavior. Generally,
the fireflies produce a flash light to draw the attention of breeding partners and also for impending prey. The
flash lights produced by fireflies are visible to only for limited distance. FF is constructed based on three
rules:
(a) Entire fireflies are unisex, i.e., the fireflies draw the attention of other fireflies irrespective of sex.
(b) The attention of draw is relative to the intensity of flashlights produced by fireflies, i.e., for a given two
fireflies, the firefly with less brightness is attracted towards the firefly which has high brightness.
(c) The variations in the light intensity are completely dependent on the quality of firefly.
Our main objective is to attain an optimal number of relay nodes based on three constraints such as
energy, connectivity and coverage. Generally, in firefly algorithm, the firefly which has higher brightness is
chosen as an optimal solution. Similarly in our work, the relay node is selected which can satisfy all the three

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170
Int J Inf & Commun Technol ISSN: 2252-8776  165

constraints. For a given set of source sensor nodes, we have to derive an optimal number of RNs which have
maximum connectivity and coverage. For this purpose, the objective function 𝑓(𝑧) defined as (7).

𝑓(𝑧) → min ∑𝑃𝑝=1|𝑃𝑝 | (7)

Subjected to
Energy constraint: (3) and (4)
Connectivity Constraint: (5)
Coverage Constraint: (6)

Here ‘z’ denotes the set of minimum number of relays through which we can gain maximum
coverage and connectivity followed by less energy consumption. Resembling with FF algorithm,
attractiveness is related to Energy, Connectivity, and Coverage. Let 𝐸𝐷(𝐴, 𝐵) be the Euclidean distance
between sensor node A and relay node B in the solution space, given by (8)

2
𝐸𝐷(𝐴, 𝐵) = √(𝐴𝑥 − 𝐵𝑥 )2 + (𝐴𝑦 − 𝐵𝑦 ) (8)

Where (𝐴𝑥 , 𝐴𝑦 ) be the co-ordinates of SSN, A and (𝐵𝑥 , 𝐵𝑦 ) be the location coordinates of a RN, B.
Grounded on this constraint, the next instance is formulated as (9):
2
𝑓(𝑥𝑗 , 𝑦𝑗 ) = 𝑓(𝑥𝑖 , 𝑦𝑖 ) + 𝛽0 𝑒 𝛾𝐸𝐷(𝐴,𝐵) (𝐴𝑗 − 𝐵𝑗 ) + 𝛼𝑡 (9)

Here 𝑓(𝑥𝑗 , 𝑦𝑗 ) is the objective function at jth instant and 𝑓(𝑥𝑖 , 𝑦𝑖 ) denotes the objective function at ith
instant, 𝛽0 is an initial distance (attractiveness) at ED = 0, 𝛾 is the distance variation between i and jth instants
and 𝛼𝑡 is a randomization parameter

3. RESULTS AND DISCUSSION

This section presentssimulation experiments conducted over the proposed model along with two
conventional models. The simulations are carried outusing MATLAB and the comparative analysis is
presented. This section includes simulation setup, performance evaluation through several performance
metrics.

3.1. Simulation set up

In simulation set up we have created a random network with varying node count as N = [20, 30, 40,
50, and 60] and with different network areas, Area = [300 × 300, 500 × 500,700 × 700,900 × 900,1100 ×
1100] and the sink is placed at a random location. The transmission range of a sensor nodes is considered as
one fourth of the network area, for example, if Area is 300 × 300 𝑚2 , then the communication range is kept
as 75 m. The communication range of both senor node and relay nodes is assumed as same. To adopt the
energy harvesting capabilities to the SNs, we have followed the energy distribution model explored in [31].
The details of simulation parameters are presented in Table 1.

Table 1. Simulation set up

Parameter Value
Number of nodes [20, 30, 40, 50, and 60]
Network Area 300 × 300, 500 × 500, 700 × 700,
[ ]
900 × 900, 1100 × 1100
Communication range ⁄
1 4 (Network Area)
Sink node location Random
Data Rate 125 kbps
Data type Constant Bit Rate (CBR)
Energy threshold (TE)) 10% of initial energy
Activation Energy (AE ) 75% of total energy capacity
Packet size 512 bytes
Harvesting energy (mW)- Uniform 0.3, 0.6, 0.9, 1.2, 1.5
Harvesting energy (mW)- Random 0.2, 0.4, 0.6, 0.9, 1.3

An efficent coverage and maximization of network lifetime in WSN through … (A. Nageswar Rao)
166  ISSN: 2252-8776

3.2. Performance metrics

The performance metrics we have considered are Total number of RNs, average energy
consumption (AEC) and network lifetime (NL). The definitions of performance metrics are done as:
− Total number of relay nodes
This metric is defined as the number of RNsessential to deploy in the network such that every SN will
have at least one RN. Lesser the total number of RNs, better the performance.
− Network lifetime
This metric is defined as the time taken by the process from the starting of a network to the time until any
SSN is failed to send the data to the sink node. In this paper, the network is assumed to be disconnected
when any of the SN is not connected to even one RN thereby it can’t send the sensed data to sink. Higher
the network lifetime, better the performance.
− Average energy consumption
This metric is defined as the average amount of energy consumed by a node (both sensor and relay) to
sense and forwards the data to sink node. Lesser the Average energy consumption, better the
performance.

3.3. Results
In this section, we represent the particulars of performance metrics evaluated after the simulation of
proposed model over varying network characteristics. Simultaneously, a detailed comparison of the proposed
MOFF-RNP and conventional approaches. Under the conventional approaches, we have compared the
MOFF-RNP with GA-RNP [13] and ABC-RNP [23]. Under the first case, we have varied the number of
nodes and measured the total number of relay nodes and network lifetime. Under the second case, we have
varied the network size and measured the total number of relay nodes and network lifetime. Under third case,
we have varied the renewable power supply and measured the network lifetime and finally the average
energy consumption is measured with varying number of RNs.
Figure 2 illustrates the details of total number of RNs deployed for varying number of SNs. From
this figure, we can observe that the number of RNs increases gradually with an increase in the number of
SNs. As the SSN count increases, they need more assistance (i.e., more umber of RNs) to forward the sensed
data to the sink. After a particular level, it becomes constant because the entire network area has been
covered by the deployed RNs and adding additional RNs consequences to more complexity. Next, the
proposed MOFF-RNP is observed to have a smaller number of RNs compared to the conventional
approaches at every count of SNs.
The GA-RNP and ABC-RNP require more relay nodes. In GA-RNP, the RNP is constrained to
fixed positions and the placement is accomplished through Genetic Algorithm. GA-RNP didn’t focus on the
connectivity but not on the coverage and energy due to which there is a need of a greater number of RNs to
cover the entire set of SSNs.Next, the ABC-RNP followed ABC algorithm for the selection of optimal
locations of RNs by which the NL is maximized while the restrictions on cost and connectivity are fulfilled.
However, this approach didn’t focus on the energy depletion rate by which the additional RNs are required
because the source sensor node will get disconnected if all the relay nodes are depleted. On an average the
proposed approach has located 8 relay nodes while the ABC-RNP and GA-RNP located 16 and 19 relay
nodes respectively. This result shows that the MOFF-RNP required only half of the relays of ABC-RNP and
this great achievement is due to the energy harvesting capabilities of SNs.
In this work, the NL is defined by the time at which any of the SSN can’t found a RN to connect.
Figure 3 illustrates the details of NL for varying number of SNs. From this figure, we can observe that the
network lifetime decreases gradually with an increase in the number of SNs. As the SSN count increases, the
network results in more activities like packet retransmissions, synchronizations and communications between
nodes by which the energy of nodes will get depleted faster rate. However, the proposed MOFF-RNP shows
a more network lifetime compared to the conventional approaches.
In the conventional approaches, the relay nodes will not have energy harvesting capabilities and if
their energy level is below the network will fail. Further, there is no back up off relay nodes and the source
sensor nodes are disconnected completely.
The GA-RNP algorithm willnot ensure that every SN is covered by RN. Hence the performance is
poor. In proposed method there exists always at least one backup RN for every SSN. Moreover, our approach
ensures a guaranteed connectivity and never makes the SSNs to accomplish the RNs responsibility. Hence
the network lifetime is more compared to the conventional approaches. On an average the proposed approach
has a network lifetime of 10,600 minutes while the ABC-RNP and GA-RNP has 7,500 minutes and 6,300
minutes respectively.

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170
Int J Inf & Commun Technol ISSN: 2252-8776  167

35 12000
MOFF-RNP MOFF-RNP
ABC-RNP[23] ABC-RNP[23]
30 10000 GA-RNP[13]
GA-RNP[13]

Network Lifetime (minutes)

25
8000
Number of Realy nodes

20
6000

15
4000

10

2000
5

0
20 30 40 50 60
0
20 25 30 35 40 45 50 55 60 Number of Sensor Nodes
Number of Sensor Nodes

Figure 2. Number of relay nodes vs number of sensor Figure 3. Network lifetime vs number of sensor
nodes nodes

Figure 4 describes the details of number of RNsobtained by the proposed and conventional
approaches as a function of network size. From this, we can observe that the number of RNs increases
gradually with an increase in the network size. For instance, in the above figure, the proposed approach
requires 2 RNs for a network size of 700x700 whereas it was increased to 8 when the network size is
increased to 1100x1100. As the network size increases, for a fixed communication range node, additional
RNs are required to deploy for covering the entire area. Since the RN can cover only a fixed coverage area,
the remaining area is covered by additional RNs. The proposed MOFF-RNP deployed a smaller number of
RNs and the increment is also gradual. This reveals that the proposed approach has high coverage capability
than the conventional approaches. Becauseit is simple; MOFF-RNP searches from the sink and deploys RNs
such that every source node is connected to sink through either one or multiple RNs. This fact benefits to the
proposed and hence the total number for RNs is less compared to the conventional approaches. On an
average the proposed approach has located 4 RNs while the ABC-RNP and GA-RNP located 7 and 10 relay
nodes respectively.
Figure 5 describes the details of network lifetime for varying network size. From this, we can
observe that the network lifetime decreases gradually with an increase in the network size. In a network with
smaller size, the relay nodes around the sink have less burden but in the case of network with larger size, the
relay nodes around sink will suffer with great burden. In an elaborated way, the relay nodes nearer to the sink
will carry the data packets of all nodes in the network thereby the energy will get depleted more quickly,
resulting in anearly failure of the network. With small networks (i.e.,300x300) the GA-RNP has lifetime of
12500 minutes while for the larger networks (1100x1100) the lifetime is 6500 and it is approximately twice.
On the other hand, the for small scale networks, the proposed approach has gained a network lifetime of
15,000 minutes while it is of 11,500 for large scale networks. This shows a greater scalability of the proposed
MOFF-RNP.

14 15000
MOFF-RNP MOFF-RNP
ABC-RNP[23] ABC-RNP[23]
12
GA-RNP[13] GA-RNP[13]
Network Lifetime (minutes)

10
Number of Realy nodes

10000

5000
4

0 0
300x300 500x500 700x700 900x900 1100x1100 300x300 500x500 700x700 900x900 1100x1100
Network Size Network Size

Figure 4 Number of RNs vs. network size Figure 5 Network lifetime vs. network size

An efficent coverage and maximization of network lifetime in WSN through … (A. Nageswar Rao)
168  ISSN: 2252-8776

In the proposed model, the relay nodes are assumed to have energy harvesting capabilities and once
their energy is below energy threshold, they will turn off all communication and moves into harvesting mode.
In the harvesting mode, the relay nodes will harvest sufficient amount of energy and will get activate. In th
simulation model, the harvesting energy is modeled with respect to time.
Figure 6 reveals the obtained network lifetime values after the simulation of proposed approach for
constant supply of renewable incremental energy. In this case, we allocate supply a constant and incremental
power supply and the network lifetime is measured. From the obtained results in Figure 6, we can observe
that the proposed approach has gained a linear increment in the network lifetime. The linearity is due to the
constant and same power supply in an incremental fashion. Furthermore, we can notice that the proposed
approach has higher network lifetime than the conventional approaches. On an average the proposed
approach has a network lifetime of 14000 minutes while the ABC-RNP and GA-RNP has 11000 minutes and
9000 minutes respectively.

4
x 10 18000
2
MOFF-RNP
MOFF-RNP 16000 ABC-RNP[23]
1.8
ABC-RNP[23] GA-RNP[13]
1.6 GA-RNP[13] 14000

Network Lifetime (minutes)

Network Lifetime (minutes)

1.4 12000

1.2 10000

1
8000
0.8
6000
0.6
4000
0.4
2000
0.2

0
0 0.2 0.4 0.6 0.9 1.3
0.3 0.6 0.9 1.2 1.5
Renewable Energy Supply (mW)
Renewable Energy Supply (mW)

Figure 6. Network Lifetime vs. Constant Renewable Figure 7. Network Lifetime vs. Random Renewable
power supply power supply

Unlike the above simulation, here we have simulated with random renewable power supply, the
observed network lifetime is shown in Figure 7. In this simulation, initially the senor nodes are sated with
random number of energies and also harvested random number of energies. From this figure, we can notice
that the observed network lifetime through random renewable power supply is lesser than the network
lifetime obtained through constant renewable power supply (as shown in Figure 6). For a detailed analysis,
we have incremented the renewable power supply in random intervals. Initially it was increased by 0.2 mW
and further it was increased by 0.3 mW and 0.4 mW and observed a maximum network lifetime of 16500
minutes (1.3 mW) whereas in the constant power supply it is of 20,000 minutes (at 1.5 mW). This is a more
realistic one because in real time there is no possibility of constant power supply due to unpredictable and
unreliable environments. On an average the proposed approach has a network lifetime of 11000 minutes
while the ABC-RNP and GA-RNP has 9200 minutes and 8300 minutes respectively

4. CONCLUSION
This paper deals with the optimization of traditional WSNs by integrating the energy harvesting
Relay Nodes. The central idea of this paper is to optimize three factors such as Average Energy
Consumption, Connectivity and Coverage. This problem is noticed as NP-hard optimization from the earlier
studies. However meta-heuristics are found fairly better performance towards such problems. Based on this
inspiration, we have developed a Multi-Objective Firefly Algorithm based Relay Node Placement. In this
algorithm, the sensor nodes are assumed to have an energy harvesting capability and also focused to maintain
a set of backup relays always for every senor node. Further, this approach also considered connectivity and
coverage constraints to attain an optimal number of relay nodes through which every sensor node will have a
connection at any time. Finally, the optimal relay node placement is formulated as an objective function and
solved through firefly algorithm. Through computer simulations we have proved that proposed method
outperformed the state-of-art methods.

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170
Int J Inf & Commun Technol ISSN: 2252-8776  169

ACKNOWLEDGEMENTS
This works is supported by the Wireless Sensor Networks Lab, Dept of ECE, University College of
Engineering, Osmania University, funded by Ministry of Electronics & Information Technology (MeiTY),
New Delhi

REFERENCES
[1] J. Y. B. Mukherjee, B. Mukherjee and D. Ghosal, “Wireless sensor network survey,” Comput. Netw., vol. 52, no.
12, pp. 2292–2330, 2008, doi: Biswanath Mukherjee.
[2] I. F. Akyildiz, Weilian Su, Y. Sankarasubramaniam and E. Cayirci, "A survey on sensor networks," in IEEE
Communications Magazine, vol. 40, no. 8, pp. 102-114, Aug. 2002, doi: 10.1109/MCOM.2002.1024422.
[3] W. Dron, S. Duquennoy, T. Voigt, K. Hachicha and P. Garda, “An Emulation-Based Method for Lifetime
Estimation of Wireless Sensor Networks,” 2014 IEEE International Conference on Distributed Computing in
Sensor Systems, 2014, pp. 241-248, doi: 10.1109/DCOSS.2014.10.
[4] M. Magno, D. Boyle, D. Brunelli, B. O'Flynn, E. Popovici and L. Benini, “Extended Wireless Monitoring Through
Intelligent Hybrid Energy Supply,” in IEEE Transactions on Industrial Electronics, vol. 61, no. 4, pp. 1871-1881,
2014. doi: 10.1109/TIE.2013.2267694.
[5] M. A. Razzaque, M. H. U. Ahmed, C.S. Hong, and S. Lee, “QoS-aware distributed adaptive cooperative routing in
wireless sensor networks,” Ad Hoc Netw., vol. 19, pp. 28–42, 2014, doi: 10.1016/j.adhoc.2014.02.002.
[6] M. Younis and K. Akkaya, “Strategies and techniquesfor node placement in wireless sensor networks:a survey,”
Ad Hoc Network, vol. 4, no. 4, pp. 621–655, 2008, doi: 10.1016/j.adhoc.2007.05.003.
[7] J. Lee, T. Kwon and J. Song, “Group Connectivity Model for Industrial Wireless Sensor Networks,” in IEEE
Transactions on Industrial Electronics, vol. 57, no. 5, pp. 1835-1844, May 2010, doi: 10.1109/TIE.2009.2033089.
[8] S. Lee and M. Younis, “Optimized relay placement to federate segments in wireless sensor networks,” in IEEE
Journal on Selected Areas in Communications, vol. 28, no. 5, pp. 742-752, June 2010, doi:
10.1109/JSAC.2010.100611.
[9] A. A. Abbasi, M. Younis and K. Akkaya, “Movement-Assisted Connectivity Restoration in Wireless Sensor and
Actor Networks,” in IEEE Transactions on Parallel and Distributed Systems, vol. 20, no. 9, pp. 1366-1379, Sept.
2009, doi: 10.1109/TPDS.2008.246.
[10] F. Al-Turjman, , H. Hassanein, and A. Ibnkahla, “Efficient deployment of wireless sensor networks
targetingenvironment monitoring applications,” Computer Communications, vol. 36, no. 2, pp. 135–148, 2013, doi:
10.1016/j.comcom.2012.08.021.
[11] H. T. Khosrowshahi, M. Shakeri, “Relay Node Placement for Connectivity Restoration in Wireless Sensor
Networks Using Genetic Algorithms,” International Journal of Electronics and Communication Engineering, vol.
12, no. 3, pp.161-170, 2018, doi: 10.5281/zenodo.1315907.
[12] Y. Wang, L. Tan, and Y. Zhang, “Relay Node Placement in Hierarchical Wireless Sensor Networks,” Journal of
Advances in Computer Networks, vol. 5, no. 1, 2017, 41-46.
[13] S. K. Gupta, P. Kuila and P. K. Jana, “Genetic Algorithm for k-ConnectedRelay Node Placement in WirelessSensor
Networks,” Proceedings of the Second International Conference on Computer and Communication Technologies,
vol. 379, 2016, doi: 10.1007/978-81-322-2517-169.
[14] J. George and R. M. Sharma, “Relay node placement in wireless sensor networks using modified genetic
algorithm,” 2016 2nd International Conference on Applied and Theoretical Computing and Communication
Technology (iCATccT), 2016, pp. 551-556, doi: 10.1109/ICATCCT.2016.7912061.
[15] D. R. Dandekar and P. R. Deshmuk, “Fault-Tolerant Relay Placement in Wireless Sensor Networks Using Particle
Swarm Optimization,” Proceedings of the International Conference on SocProS 2011, AISC, vol. 130, pp. 749–
757, 2012, doi: 10.1007/978-81-322-0487-9_71.
[16] Y. Xu, Y. Xiao and Q. Sun, “A Swarm-Based Meta-Heuristic for Relay Nodes Placement in Wireless Sensor
Networks,” International Journal of Innovative Computing, Information and Control, vol. 15, no. 2, pp.551-567,
2019, doi: 10.24507/ijicic.15.02.551.
[17] S. K. Udgata, S. L. Sabat and S. Mini, “Sensor deployment in irregular terrain using Artificial Bee Colony
algorithm,” 2009 World Congress on Nature & Biologically Inspired Computing (NaBIC), 2009, pp. 1309-1314,
doi: 10.1109/NABIC.2009.5393734.
[18] A. R. Ajayan, and S. Balaji, “A modified ABC algorithm& its application to wireless sensor networkdynamic
deployment,” IOSR Journal of Electronicsand Communication Engineering, vol. 4, no. 6, pp. 79–82, 2013.
[19] X. Liu, and D. He, “Ant colony optimization withgreedy migration mechanism for node deploymentin wireless
sensor networks,” Journal of Network and Computer Applications, vol. 39, no. 1, pp. 310–318, 2014, doi:
10.1016/j.jnca.2013.07.010.
[20] I. Senturka, K. Akkayaa, and S. Yilmaz, “Relay placement for restoring connectivity in partitioned wireless sensor
networks under limited information,” Ad Hoc Networks, vol. 13, pp. 487–503, 2014, doi:
10.1016/j.adhoc.2013.09.005.
[21] S. Mini, S. K. Udgata and S. L. Sabat, “Sensor Deployment and Scheduling for Target Coverage Problem in
Wireless Sensor Networks,” in IEEE Sensors Journal, vol. 14, no. 3, pp. 636-644, March 2014, doi:
10.1109/JSEN.2013.2286332.

An efficent coverage and maximization of network lifetime in WSN through … (A. Nageswar Rao)
170  ISSN: 2252-8776

[22] J. M. Lanza-Gutierrez and J. A. Gomez-Pulido, “Assuming multiobjectivemetaheuristics to solve a three-

objectiveoptimization problem for Relay Node deployment in Wireless SensorNetworks,” Applied Soft Computing,
vol. 30, pp. 675–687, 2015, doi: 10.1016/j.asoc.2015.01.051.
[23] H. A. Hashim, B. O. Ayinde, and M. A. Abido, “Optimal placementof relay nodes in wireless sensor network using
artificial bee colony algorithm,” Journal of Network and Computer Applications, vol. 64, pp. 239-248, 2016, doi:
10.1016/j.jnca.2015.09.013.
[24] S. Misra, N. E. Majd and H. Huang, “Constrained Relay Node Placement in Energy Harvesting Wireless Sensor
Networks,” 2011 IEEE Eighth International Conference on Mobile Ad-Hoc and Sensor Systems, 2011, pp. 25-34,
doi: 10.1109/MASS.2011.137.
[25] S. Misra, N. E. Majd and H. Huang, “Approximation Algorithms for Constrained Relay Node Placement in Energy
Harvesting Wireless Sensor Networks,” in IEEE Transactions on Computers, vol. 63, no. 12, pp. 2933-2947, Dec.
2014, doi: 10.1109/TC.2013.171.
[26] W. K. G. Seah, Z. A. Eu and H. Tan, “Wireless sensor networks powered by ambient energy harvesting (WSN-
HEAP) - Survey and challenges,” 2009 1st International Conference on Wireless Communication, Vehicular
Technology, Information Theory and Aerospace & Electronic Systems Technology, 2009, pp. 1-5, doi:
10.1109/WIRELESSVITAE.2009.5172411.
[27] V. Raghunathan and P. H. Chou, “Design and Power Management of Energy Harvesting Embedded Systems,”
ISLPED'06 Proceedings of the 2006 International Symposium on Low Power Electronics and Design, 2006, pp.
369-374, doi: 10.1145/1165573.1165663.
[28] S. Liu, J. Lu, Q. Wu and Q. Qiu, “Harvesting-Aware Power Management for Real-Time Systems With Renewable
Energy,” in IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 20, no. 8, pp. 1473-1486,
Aug. 2012, doi: 10.1109/TVLSI.2011.2159820.
[29] C. Renner, S. Unterschütz, V. Turau, K. Römer, “Perpetual data collection with energy-harvesting sensor
networks,” ACM Trans. Sen.Netw., vol. 11, no. 1, pp. 12:1–12:45, 2014, doi: 10.1145/2566675.
[30] X. S. Yang, “Firefly algorithms for multimodal optimization, in: Stochastic Algorithms: Foundations and
Applications,” Lecture Notes in Computer Science, Springer, Berlin/Heidelberg, vol. 5792, pp. 169–178, 2009, doi:
10.1007/978-3-642-04944-6_14.
[31] V. Raghunathan, A. Kansal, J. Hsu, J. Friedman and M. Srivastava, “Design considerations for solar energy
harvesting wireless embedded systems,” IPSN 2005. Fourth International Symposium on Information Processing in
Sensor Networks, 2005., 2005, pp. 457-462, doi: 10.1109/IPSN.2005.1440973.

BIOGRAPHIES OF AUTHORS

A. Nageswar Rao received B.E and M.E degrees in Electronics and Communication
Engineering in 2000 and 2004 from Nagarjuna University and IIT Kahragpur and he is currently
working with Strategic Electronics& Research and Design Center (SLRDC) –HAL, Hyderabad
as a Manager (Design) and pursuing his PhD under Visveswaraiah Scholarship sponsored by
DeitY and Media Asia Labs in the Wireless Sensor Networks at Osmania University

B. Rajendra Naik received his B. Tech in ECE from Nagarjuna University -2000, M.E from
Osmania University-2005 and Ph. D from Osmania University-2013. He has rich experience in
Teaching and Research.Currently he is working as Head & Professor in Dept of ECE, University
College of Engineering, Osmania University, Hyderabad. He worked as a Visiting Researcher at
Tokyo Metropolitan University, Tokyo, JAPAN. He published many research papers at National
and International conferences and Journals. He is a recipient of Young Faculty research
Fellowship DeitY Govt.of India, and he is a member of IEEE, IETE.

L. Nirmala Devi received B.E and M.E degrees in Electronics and Communication Engineering
in 1997 and 2005 from Osmania University Hyderabad, India. She joined the Department of
Electronics and Communication Engineering Osmania University as a Senior Assistant professor
in 2007. She is completed her PhD in 2014, currently she is working towards the Ministry of
Electronics and information Technology(MeitY), University Grant Commission(UGC),
sponsored projects and she completed Department of Science Technology DST sponsored
projects “Security issues in MANET” .Her research interest includes Optical networks, Mobile
AdHoc networks and wireless sensor networks, cognitive radio networks, She is delivered
lectures on Wireless Sensor network and Optical Networks and MANET on national universities
and published various national and international IEEE conferences and journals. She is visited
various countries for her paper presentation and she is a member of IEEE, OSI, OSA, FIEI.

Int J Inf & Commun Technol, Vol. 10, No. 3, December 2021 : 159–170