AN ENHANCED HYBRID ROUTING AND CLUSTERING TECHNIQUE FOR WIRELESS SENSOR NETWORK

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 12, No. 1, February 2020
DOI: 10.5121/ijwmn.2020.12102 13
AN ENHANCED HYBRID ROUTING AND
CLUSTERING TECHNIQUE FOR WIRELESS SENSOR
NETWORK
Hasan Al-Refai
Department of Computer Science, Philadelphia University, Jordan
ABSTRACT
Wireless Sensor Networks (WSN) have extensively deployed in a wide range of applications. However,
WSN still faces several limitations in processing capabilities, memory, and power supply of sensor nodes. It
is required to extend the lifetime of WSN. Mainly this is achieved by routing protocols choosing the best
transmission path in-network with desired power conservation.This cause is developing a generic protocol
framework for WSNa big challenge. This work proposed a new routing technique, described as Hybrid
Routing-Clustering (HRC) model. This new approach takes advantage of clustering and routing procedures
defined in K-Mean clustering and AODV routing, which constituted of three phases. This development aims
to achieve enhanced power conservation rate in consequence network lifetime. An extensive evaluation
methodology utilized to measure the performance of the proposed model in simulated scenarios.The results
categorized in terms of the average amount of packet received and power conservation rate. The Hybrid
Routing-Clustering (HRC) model was determined, showed enhanced results regarding both parameters. In
the end, they are comparing these results with well-known routing and well-known clustering algorithms.
KEYWORDS
Wireless Sensor Networks, Clustering, Routing, Power conservation, Network lifetime, AODV, K-Mean,
LEACH
1. INTRODUCTION
Wireless Sensor Network (WSN) is a collection of small, self-contained electromechanical
devices that monitor the environmental conditions and be useful to employ in many applications
such as medical, automotive safety, and space applications. There are many essential priorities to
build an architectural (WSN), such as deployment, mobility, infrastructure, network topology,
network size and density, connectivity, lifetime, node addressability, data aggregation, etc. Sensor
nodes have several limitations, such as limited battery life, low computational capability, short
radio transmission range, and small memory space. Still, the primary constraint of the nodes is
their limited energy resource, which causesthe disconnection of the network.
Therefore, to reduce energy usage in wireless sensor networks, many cluster-based routings have
been proposed. Among those proposed, LEACH (Low Energy Adaptive Clustering Hierarchy) is
a well-known cluster-based sensor network architecture, which aims to distribute energy
consumption evenly to every node in a given network. This clustering technique requires a
predefined number of clusters and has been developed with an assumption that the sensor nodes
are uniformly distributed throughout the network (Sukhchandan Randhawa&Sushma Jain.
2019)(Tillapart et al. 2005). Moreover, (Maurya et a. 2014) stated that Low-Energy Adaptive
Clustering Hierarchy (LEACH) is the first significant protocol, which consumes less amount of
energy while routing the data to the base station. However, other researchers discussed this issue

14
in a different way (Gnanambigai et al. .2014) stated that LEACH proves to be an outwitting
routing scheme ishas limitations in inheritance due to the power-consuming overhead processing
and the increased number of participating nodes. Such limitations motivated the researcher to
carry out this research.
Numerous citations performed for the first paper released LEACH. These studies based their
work principles on LEACH false assumptions, which in turn results in failure throughout their
researches' works. Therefore, this research implemented and adopted a new model based on
realistic values within the use of both simple routing and clustering represented in subsections
(2.1.2) and (2.1.3), respectively, to clarify LEACH assumptions limitations.
Therefore, this work will follow the below methodology:
1- Define the problem statement by implementing LEACH and AODV assumptions on real
environments’ parameters.
2- Defining the scope of work by focusing on solving the formulated problems and issues of
LEACH assumptions on real environments.
3- Proposing a new hybrid routing and clustering technique based on integrating K-Mean
clustering as a clustering algorithm and simple routing i.e., AODV, as routing algorithm,
to be reliable in real environments.
4- Divide the proposed technique into phases, to enhance its efficiency and ease the
troubleshooting process.
5- Evaluate the proposed technique by simulating it on the real environment’s parameters
using MATLAB.
Besides this section, the next section reviewed some related works and current solutions for the
problem under study. The proposed algorithm and its phases are discussed in the third section.
The fourth section discussed the experiment and the scenarios that were implemented to prove the
algorithm and the obtained results. Finally, yet importantly, the fifth section showed the
conclusions and summarized the entire work.
2. TECHNICAL BACKGROUND
Having mentioned the problem statement of this work in addition to our proposed solution, in this
section, brief details about the techniques, schemas, and algorithms to be used in this work will be
shown.
Wireless Sensor Network (WSN) is an emerging network technology that provides reputable
monitoring of the various environmental circumstances. One of the paramount constraints in the
WSN is the scrimp energy resource. Many experimental works in WSNs are focussed on
achieving energy efficacy. Many researchers focused on routing schemes as an effective Factor to
achieve energy-efficient operation (Sukhchandan Randhawa&Sushma Jain. 2019),
(Kaswan A., Singh V., Jana P.K. 2018), ( Amit Sarkar1 & T. Senthil Murugan. 2019),
(Gnanambigai et al., 2014). designing an effective routing protocol is a critical approach for
energy conservations in wireless sensor networks (Huang, & Yen, 2009).The latest algorithm of
routing build and test under the assumption of uniformly distributed sensor node .although, the
determines of this assumption is that it needs to be installed and condensed in some area and
spares in other areas to form the scope of the ongoing monitoring process. (Baroudi et al., 2012).
Wireless Sensor Network is modern technology, and it developed rapidly. Extending the lifetime
of a wireless sensor network is highly recommended issues. Reduce energy consumption in the
network by choosing the best transmission path in a network that is responsible for the routing

15
techniques (Sharma et al.,2015). Consequently, Routing protocols schemes are the main
requirements for developing a model for WSNs, if the desired is power conservation. There are
mainly two fields of managing the WSN environment, categorized as Clustering Algorithms and
Routing Algorithms.
3. CLUSTERING OF SENSOR NODES
The clustering technique has been proved to be one of the most efficient methods in this field, due
to its scalability, the ability of aggregating data, minimized energy consumption, and robustness.
Many protocols are categorized under the umbrella of cluster-based protocols. However, some of
these protocols repeat the clustering operation at certain intervals of time, leading to a potential
waste of energy.
To name a few of clustering techniques in WSNs: Low-Energy Adaptive Clustering Hierarchy
(LEACH) (Heinzelman et al. 2000), Hybrid Energy-Efficient Distributed clustering (HEED)
(Younis & Fahmy 2004), Distributed Weight-based Energy-efficient Hierarchical Clustering
protocol (DWEHC) (Ding et al. 2005), and Position-based Aggregator Node ELection protocol
(PANEL) (Sukhchandan Randhawa&Sushma Jain. 2019), ( utty andSchaffer,2010).
Researchers suggested that grouping the sensors into clusters will provide higher scalability, and
make energy consumption more efficient, hence, prolonging the lifetime of the entire WSN. Such
clusters allow for aggregation and limiting data transmission.
Clustering means that nodes are divided into virtual groups according to some rules, where nodes
belonging to a group can execute different functions from other nodes. When forming a cluster, a
“Cluster Head”, CH, is elected and the members of such a cluster can communicate with their CH
directly; then this CH can forward the aggregated data to the central base station, eventually,
through other CHs (Kumar, et al.; 2014) and (Halder, & Ghosal, 2015), (Sukhchandan
Randhawa&Sushma Jain. 2019). Figure 1 below shows how a clustered WSN work:
Figure 1: Clustering Technique in a WSN
3.1. LEACH Algorithm
The LEACH (Low-Energy Adaptive Clustering Hierarchy) clustering protocol (Heinzelman et al.,
2000) was a pioneer in such methodology. In this clustering protocol, the head repeatedly rotates
amongst the nodes, to balance the energy consumption. This way, energy in the nodes is reduced
evenly among the nodes, and untimely battery drainage can be avoided. In this protocol, each
sensor node transmits collected data to the CH, which in turn collects them and sends them

16
directly to the base station (which is known as the data sink), regardless of its distance. It is most
efficient when the area covered by WSN is small, or when the cost of receiving data is high.
Figure 2 below shows the clustering hierarchy of a WSN applying the LEACH protocol:
Figure 2: LEACH Clustering
LEACH works over two phases:
1- The Setup Phase (where cluster-heads are chosen)
- Cluster-head selection: where the probability of a node to become a CH is calculated on timed
intervals, and the CH selection is made independently by each node without consulting other
nodes in the cluster to minimize overhead in cluster head establishment. This probability
decreases in case a node was chosen to be a CH in previous rounds; each node during cluster head
selection will generate a random number between 0 and 1. If the number is less than a calculated
threshold, the node will become a cluster head. The threshold is calculated as:
Where n is the current node, P is the a priori probability of this node to be selected as a CH, r is
the order of the current round, and G is the set of nodes that haven’tbecomeCHs within the last
1/P rounds.
- Cluster Set-Up: Each node in the WSN will broadcast a message to the rest of the nodes stating
its status. These nodes will then determine the most
suitableclustertheyprefertojoinbasedonthereceivedmessage’sstrength.CHs must keep their
receivers ON to receive surroundingnodes’decisions.
- Transmission schedule creation: the CH creates a schedule containing the number of nodes in
the cluster, and then nodes send their data to the cluster head.
2. Steady-state Phase (CH manages data transmission among connected nodes)
- Data transmission begins from cluster nodes to cluster heads.
- Validation of received signals and aggregation of the data to be transmitted to the base station. -
Data transmission from cluster heads to the base station.

17
Power Model of LEACH Protocol
To accurately define our problem, which can formulate the motivation of this research, we have
implemented the LEACH assumption on parameters from a real environment containing physical
tools. However, the impact of the power model on the LEACH protocol is illustrated in figure 3.
We have changed the network size in this configuration from 100x100 to 200x200 with the same
number of nodes without changing the location of the sink node. It can observe that in 100x100
areas, the entire nodes are connected as well as the whole cluster heads can reach a sink node
since the distance between nodes and sink node is less than 100. However, we can observe that in
area 200x200, the disconnected nodes number, which cannot even reach a cluster head is larger
than 50% of the total number of nodes.
Moreover, 4 of the 5 selected cluster heads in different rounds could not reach a sink node. This
figure shows the importance of routing in WSN. Furthermore, it explains that LEACH has many
problems, which developed a motivation to solve them by proposing this algorithm.
3.2. Network Life Time for LEACH and Simple Routing
After detailing the related parameters of LEACH and simple routing algorithm, we can evaluate
the life-time of their nodes and the entire network, by applying their assumptions on the same
parameters. Figure 4shows the CDF of the network lifetime of both of LEACH and simple
routing. We can observe that under the Wi-Fi power model. Simple routing lives more than
LEACH. We can observe that 80 % of nodes live more than 1800 rounds, unlike LEACH, which
die in less than 1800 rounds. We also can observe the distribution in dying of nodes in LEACH.
Since LEACH randomizes the dying nodes, which leads to the more fairallocation of the dead
nodes and then longer life-time. However, one thing to be mentioned is that the nodes send the
data even if these data did not receive from the sink node. In other words, nodes also sent if the
network is disconnected. We have observed that only 819 packets have been received in the area
of 300x300 using AODV, comparing with 1770 using Clustering. Moreover, to compare the
lifetime between LEACH and the Routing (AODV), the disconnected rounds have been deleted
from the LEACH CDF plot for area 300x300.
Figure 3: Impact of the power model on LEACH protocol

18
Figure (4): Routing vs. ClusteringCDF (Network Lifetime)
4. RELATED WORK
Several studies are carried out based on LEACH assumptions such as: (Tillapart et al.,
2005;Haiming, &Sikdar, 2007; Yan et al., 2009; Huang,& Yen, 2009Baroudi, et al., 2012;
Baroudi et al., 2012; Guo et al., 2012; Guo et al., 2012;Tyagi et al.; 2013;Zhang et al., 2013;
Patra, Chouhan, 2013; Zhang, & Zhao, 2014;Gambhir,& Fatima, 2014;Gnanambigai et al., 2014;
Sharma, & Choudhary, 2014; Maurya et al., 2014). Others are studied routing and clustering, such
as, while some research works achieved good performance based on the assumption of equal
initial power and/or predefined ratio in types of nodes, these mechanisms are not flexible.
Besides, existed works rarely deal with cluster head distribution (Tsai & Chen, 2015). Studies that
are concerned in hybrid models based on leaching like: (Tillapart et al., 2005;Zhang et al., 2013;
Patra, Chouhan, 2013).
(Tillapart et al., 2005) proposed a new architecture based on hybrid clustering and routing for
WSNs quite similar to the proposed research model with some differences in the main parts based
on. Their method relays on three parts: first, a technique termed modified subtractive clustering
technique. Secondly, a technique termed an energy-aware cluster head selection.
Finally, an algorithm termed coast based routing algorithm. These supposed to be performed at
the base station BS because they are considered as centralized techniques, and it is showed that
their proposed architecture outperforms the LEACH scheme in terms of energy conservations.
(Haiming andSikdar, 2007) reduced the consumption happening to the energy throughout
excluding the requirements of synchronizing TDMA. With the use of LEACH sleep-wake up
based on decentralizing MAC protocol as the first step. Then, illustration for an optimal
probability-obtaining framework is made, where the node becomes CH to reduce the energy
consumption in the network. Firstly, the small networks are illustrated in their analysis with an
assumption that the distance expected for all CHs from the sink node is similar. After that, the
larger networks represented as a complement analysis of small ones; but with a conditionwhere
the distance, as mentioned above, is altered not similar. The assumption, however,changed
because the nodes in large networks may be farther away from the sink, which in turn required
more energy to be consumed until it reaches the sink. Their resulted simulations show that their
proposed method outperforms LEACH in terms of significant energy consumption, which relates
to one of our research parameters that will be compared throughout the research proposed method
concerning LEACH.

19
(Huang & Yen.2009) proposed a protocol also to reduce power consumption efficiently. This
protocol called CB-DHRP. Their proposed protocol designed according to LEACH scheme with
identical work principle with only one difference in single LEACH assumption. This assumption
relates to the transmission range for a node and the method of formation, which used to create a
routing structure.
Regarding the restricted transmitting range of a node and the assist of the cluster member and
CHs combination CB-DHRP builds transmission paths, which can be described with an efficient
energy reservation and have the ability to send the sensed data into the sink base station BS. Their
resulted simulations show that, within an available selection to CHs, CB-DHRP outperforms
LEACH, ERA, and PEGASIS protocols in terms of improving network lifetime by 1.5 times
better than the other protocol above does. This study is critical to our research due to the network
lifetime parameter that CB-DHRP compared its performance to, concerning LEACH.
To achieve power conservation, we should use the proposed hybrid self-decisive clustering
technique based on Hierarchical Agglomerative Clustering and k-means algorithm, as earlier
mentioned before. Such algorithms are to arrive at an optimal number of clusters for a given set of
nodes distributed over the geographical area.
5. PROPOSED ALGORITHM
The proposed hybrid routing-clustering (HRC) model is described in section 3.4. However, the
demonstration of sensor nodes’ power, data rates, and transmission distances used within the
proposed model are described in section (3.1) and (3.2).
5.1. Power Consumption Demonstration for Sensor Nodes
WSN contains numerous small devices called sensor nodes, which are frequently setup randomly
over a wide area for sensing and monitoring purposes to the different physical phenomena
associated parameters together with environmental circumstances at different positions, and
communicate with each other. However, WSN devices have various resource limitations, such as
fewer memories, small clock speed, limited battery energy, and restricted computational power.
Similarly, the lifetime of the network, the effectiveness of the energy, load balance, and much
more scalability key issues, which considered the fundamental requirements of WSN applications
(Kodali&Aravapalli, 2014).
However, as explained in later sections, sensor nodes of the proposed power model are Wi-Fi
enabled nodes. See table 1, which shows the consumed power to send or receive bits. These
values can be found in (Halperin et al., 2010). Utilizing these values, equation (1) illustrates the
required power to send while equation (2) illustrates the needed power to receive and then, listen.
Table 1: Power model coefficients
Variable Value
Power to send (Ps) 17nJ/bit
Power to listen (Pl) 12nJ/bit

20
𝑆𝑒𝑛𝑑𝑃𝑜𝑤𝑒𝑟 (𝑇𝑥) = 𝑃𝑠 ∗ 𝑃𝑎𝑐𝑘𝑒𝑡𝑠𝑖𝑧e (1)
𝑅𝑒𝑐𝑒𝑖𝑣𝑖𝑛𝑔𝑃𝑜𝑤𝑒𝑟 (𝑅 𝑥) = 𝑃𝑙 ∗ 𝑃𝑎𝑐𝑘𝑒𝑡𝑠𝑖𝑧𝑒 (2)
5.2. Data Rate and Transmission Distance Demonstration
From observing the above equations, power does not depend on distance as in LEACH
assumptions. Power depends on packet size and bit rate, as shown in (Halperin et al. 2010).
In the proposed model, the first assumption of LEACH protocol, which is regarding that each
sensor nodes contain an equal amount of energy, was modified, and nodes are not required to
measure distances and direct their antennas. In other words, this model is implementable. Two
parameter values are defined in the proposed model; first, the bit rate of sensor nodes is 50Mbps.
Second, the transmission distance is constant, which is 100m and not like LEACH.
5.3. Proposed Hybrid Routing-Clustering (HRC) Model
After demonstrating the power consumption, data rate, and transmission distance, the proposed
protocol is described in this section.
Figure 5: Flow chart of the proposed algorithm.
HRC consists of three main phases. In the first phase, the sink node broadcasts a packet. Any
node receives this packet will be marked as layer 1. Subsequently, the nodes in this layer will add
the layer number to the packet and rebroadcasts it. Any node receives this rebroadcasted packet
will be identified as layer 2. The nodes in layer 2 will continue in the same procedure until all
nodes in the network receive a layer number.
These two layers, in addition to the HRC, three phases will be briefly identified in the next
subsections below, but after explaining briefly how this implemented system algorithm works in
detail. See figure 5, which illustrates the flow chart of the proposed algorithm.
Flooding Message From Sink
Nodes Broadcast Message with its Layer Number
Build Routing Table (Rout to the Sink) in each node
Calculate Clusters Utilizing parameters from Sink Flood
Message
Average Nodes in clusters accoding to Location of Centroids
Send Data to Cluster Heads and use table to route data to sink

21
First Phase: Preparing the Nodes
The sink node sends a broadcast to all nodes within its coverage area. All nodes that receive the
broadcast message resend a broadcast and declare as layer one. All nodes that receive the resend
broadcast are clarified as layer two and resend a broadcast, and so on. For more detailed steps on
what is happening during this phase, the steps below are stated as follows:
1- Each node has a physical address.
2- Sink node knows its location in the network.
3- Sink node broadcast a MSG <sink physical address, layer 0, location, distance=0, random
values = 20, 30, 40, 50>.
4- Any node receives this MSG will mark itself as Layer 1 node. They modify the received MSG
as follows < Physical address, layer 1, location “sink”, distance= can be computed from equation
(3), random values=20, 30, 40, 50>.
5- Each node in this layer will broadcast this MSG. If a layer one node receives this MSG, it will
add the node physical address into a neighbor table like the one shown in table 2.
6- Any node which has no layer number receives these broadcast MSG will update its status to be
in layer= MSG.layer+1. Subsequently, it will broadcast the MSG by modifying the following
fields <Physical address, layer 2, location “sink”, distance=MSG.distance+ computed new
distance, random values=20, 30, 40, 50>
7- If a node in the current layer (e.g. layer 2) receives multi-packets from nodes with layer
number less than its layer number (e.g., layer 1). It will use the first packet it received to calculate
the distance, and it will add the physical address in the new MSG into a gateway table for
possible routing nodes. See table 3 for an example for a gateway list.
8- Any nodes receive this MSG will do the following.
A- If the node in Layer 1 discards the MSG.
B- If the node in Layer2 adds the Physical address into the neighbor table.
C- If it has, no layer will repeat step 6 and so on.
D- If multi packets from the layer add the physical address to the gateway table
9- When nodes in the last layer broadcast the MSG, no new nodes will receive the MSG;
however, they will consider itself as a leaf node.
Table 2: Neighbor table
Physical address Mark
1
4
8
Etc.

22
Table 3: Gateway List
Node 4
Node 7
Etc.
All nodes will receive these MSGs. If a node does not receive any MSG, this means it cannot
connect with any other nodes, and it should be relocated since it is disconnected from the
Network. Equation (3) illustrates the Free-space propagation model (Schmidt-Eisenlohr, 2010).
Any propagation model can be used. However, this one is the easiest.
𝑃𝑇(𝑑) =
𝑃𝑡 𝐺𝑡 𝐺 𝑇 𝜆2
(4𝜋)2 𝑑2 𝐿
………………….. (3)
Where:
𝑃𝑡Is the transmitted power
𝑃𝑇(𝑑)Is the received power
𝐺𝑡Is the transmitter antenna gain mostly equal to one
𝐺 𝑇Is the receiver antenna gain mostly equal to one
𝑑Is the distance between transmitter and receiver
𝜆Is the wavelength in which the communication takes place.
𝐿Is the system loss factor dependent upon line attenuation, and its equal to one.
Second Phase: Clustering and Centroid
In this phase, the clusters are constructed. The WSNs are consist of a number of clusters (# CHs).
In this work, 100 nodes are distributed randomly in the different coverage area dimensions;
100x100, 200x200 ... 500x500. Moreover, all the nodes in this phase each (nodes in any layer
greater than layer one) will perform the following steps:
1. First, compare that node’s Physical ID with all Physical IDs in its neighbor table. If it has the
highest ID number, it will wait.
2. If it is the smallest Physical ID number, it will select the first random value “20” and add it to
its calculated distance to be selected as a centroid.
3. It will broadcast this Physical ID number<MSG.Cluster>. Each neighbor node will receive this
number anduse it to calculate their distances to the new centroid.
4. If the ID number of the node is in the middle of other nodes in the neighbor list (it is ID
number is greaterthan some node’s ID and smaller than some node’s ID), it will perform the
procedures described in (b) and (c) of step (8) in the first phase. If the node received
<MSG.Cluster> before, it will use the second random value. Else, it will use the first one.
5. When a node receives MSG.Cluster messages from multiple nodes, it will save the distance and
the node ID of this MSG. Subsequently, it will calculate the distance between its distance and the
value of centroid in these MSGs. It will choose the smallest distance.
6. Each node will broadcast the cluster name it will follow. The names will be the physical
address of the node sent to the MSG.cluster.
7. Nodes in the same cluster will hear these broadcasts. They will build the cluster list “table” of
all nodes in the same cluster with them.
8. The node with the smallest ID “Physical address” will be the first cluster head for the first
round.

23
9. After each round, the nodes will change to the second cluster head in the list. Thesecond higher
physical address and so on.
Third Phase: Data Routing
After using-mean clustering, each node is referred to as a particular cluster. The following steps
are responsible for routing process accomplished in the third phase of HRC:
1- Any cluster head will wait to receive data from all nodes in its cluster “round” it will add them
to one packet and rout them.
2- To route the data, the cluster head will choose one node of its gateway table randomly as the
next-hop address and sends the packet to that node.
2- Cluster head can randomly choose any node from the gateway table for load balancing.
6. HRC EVALUATION (EXPERIMENTAL ENVIRONMENT)
The proposed HRC model was evaluated through a set of simulation studies conducted using
Matlab. Three different protocols have been considered; LEACH, simple routing, and proposed
HRC. Moreover, the static distance has been embedded in LEACH to determine its impact on
network connectivity. Simple routing has been simulated to be a ground truth of simple
configuration without clustering. Table (4) shows the configuration used for the simulation
environment. Five different scenarios were used, each with different coverage area
dimensionsstaring from 100x100 and ending with 500x500. These scenarios were repeated 30
times and then averaged for accuracy.
Table (4): research configuration of the simulation Environment
6.1. Research Results
Results are categorized according to the average received packets and power conservation
achieved by using Proposed HRC, LEACH, and Simple Routing. The comparison was based on
adifferent number of rounds and several area sizes.
6.2 The Average Received Packets in HRC and Simple Routing
First of all, the values of the below table have resulted. Table (5) describes the average received
packets in HRC and using a simple routing approach. The average received packets are used to
clarify the advances achieved in the proposed HRC in receiving a greater amount of received
packets comparing to simple routing. The maximum average of the received packets was1834
packets if all nodes can reach the sink node without any routing. This can be calculated according
Component Value
Area Size 5 scenarios. (100x100, 200x200 …. 500x500)
Number of nodes 100
Initial power in J 0.5J
Number of Sink nodes 1
Packet Size 2000 Bytes
Sink Node Location (50,50)

24
to equation (4). This value is considered as the maximum boundary of the number of transmitted
packets.
𝒎𝒂𝒙𝒊𝒎𝒖𝒏𝒂𝒗𝒆𝒓𝒂𝒈𝒆𝑹𝒙𝒑𝒂𝒄𝒌𝒆𝒕𝒔 =
𝒊𝒏𝒊𝒕𝒊𝒂𝒍𝒑𝒐𝒘𝒆𝒓 ×𝒑𝒂𝒄𝒌𝒆𝒕𝒔𝒔𝒊𝒛𝒆
𝒕𝒓𝒂𝒏𝒔𝒎𝒊𝒕𝒕𝒆𝒅𝒑𝒐𝒘𝒆𝒓
……... (4)
The average packet received for simple routing was 551. The reason behind this small number of
received packets is the death of nodes in the first layer. These nodes died after a few rounds
because of the routing process. However, the number was increased in HRC because of clustering
and compressing processes of data.
Table (5): the average received Packets and simple routing:
6.3 Comparing HRC and Simple Routing
A comparison between simple routing and HRC in the lifetime of the network is illustrated in
figure 6. This comparison has been conducted with an area size of 200x200 m.It can be noticed
that 50% of the nodes consumed their power in less than 600 rounds within the use of simple
routing; however, these nodes survived until approximately 1200 rounds using HRC. Moreover, it
is quite noticeable that 35% of the nodes in simple routing lived until 1800 rounds; these nodes
are leaf nodes, which do not route any traffic in the network. Therefore, this provides evidence
that the network dies from meddling and not in a random manner as it is assumed in LEACH.
Figure 6: CDF of LEACH and Proposed HRC
Figure7, the y-axis indicates Cumulative Distribution Function (CDF) measured by percentage
(1=100%, 0.9 = 90% … and so on). The stability in simple routing graph between rounds (400-
600, 600-800, and 800-1800), respectively, indicate useless nodes, which still alive, but it does
not route anymore. Such nodes are considered leaf nodes whose location is way far from the sink
node. After round 1800, those nodes are considered to die due to lack of electricity.
Simple Routing 551
HRC 1304

25
Figure 7: Network Lifetime
6.4 The Relation between HRC Layers and Area Size
Figure 8: Relation between Layers and Network Size
The relation between the number of layers in HRC and the area size is illustrated in figure 8. It is
clear that this relationship is linear. In addition, within an area of 500x500, 8 layers are needed.
This shows the limitation in the LEACH protocol, which has no routing at all.
Figure (8) shows the Cumulative Distribution Function. LEACH nodes lived shorter connected
time than the proposed HRC method. We also can observe that 50% of nodes died in less than
800 rounds. However, in the proposed model, 50% of nodes lived 1100 rounds. The improvement
can be measured as average rounds in proposed over average rounds in LEACH, which is equal to
57.3%. As shown in the below figure.
6.5 The Power Conservation Comparison between Proposed Method, LEACH and
Simple Routing
A comparison of the remaining power after each round in all threeproposed methods HRC,
LEACH, andsimple routing is illustrated in figure 9.
This power conservation comparison has been conducted with an area size of 300x300 m. The
average power after each round can be calculated as the sum of the remaining power in all nodes
divided by the number of nodes.

26
Figure (9): illustrates the power conservation comparison.
It is noticed that the remaining average power in routing decreases faster in both methods, simple
routing and LEACH, comparing to the proposed HRC method. Also, the average number of
rounds passed in routing before reaching zero power while running LEACH and simple routing
was smaller than the proposed HRC. Accordingly, it can be determined that the proposed HRC
outperformed both methods in terms of power conservation, taking into consideration the number
of rounds.
Here you need to add a table summarizing all scenarios results considering all five-simulation
area sizes. All results should be the average of the 30 iterations.
7. CONCLUSIONS
Several researchers have been carried out to overcome the shortcomings in LEACH protocol
while other researches based their work on LEACH assumptions, but none of the related research
works builds a WSN regarding what’s available in the real world. Moreover, none of the
researchers concentrated on using a hybrid routing clustering model, most of the researchers
focused on using a routing protocol, or hybrid routing, and/or clustering protocol.
In LEACH, protocol, two assumptions were proposed as it first described in (Liao, & Zhu, 2013);
power consumption is depending on the distance, and all nodes can reach the sink. Where the
distance mentioned here indicates how far is that nodes from the sink, CH as a normal nod. One
another assumption is about the antenna defined for the LEACH protocol. There are two types of
transmitting antenna; directed antenna, and undirected antenna. In the directed antenna, the sink
can send a message to the node with a very far distance in the beam but ata high cost.
On the other hand, in the undirected antenna, the sink can send a message to the node with a
limited distance. However, the antenna used in LEACH is the directed antenna. Theses
assumptions mentioned above which are cited and used in numerous papers such as; (Chen. et al,
2007; Kang, & Nguyen, 2012; Li. et al, 2011; Ran. et al, 2010; Heinzelman. et al. 2000; Ramesh,
& Somasundaram, 2011; Zhao. et al, 2012 … etc.) shows the unrealistic assumptions used within
its simulation which leads to the inapplicability that face LEACH.
Therefore, this research can be considered as the first attempt in studying WSNs based on real,
valid, and implementable values and facts within the use of HRC protocol to obtain such results
that outperform any resulted ones using other WSNs’ protocols in any related works.

27
Due to that end, this research has adopted a new model proposed in this research work based on
realistic values rather than LEACH assumptions to overcome thelimitation found in LEACH and
to introduce a new model represented in HRC.
REFERENCES:
[1] Al Basset Almamou, A., Wrede, R., Kumar, P., Labiod, H., & Schiller, J. (2009). Performance
evaluation of routing protocols in a Real-World WSN. Information Infrastructure Symposium, 2009.
GIIS'09.Global, 1-5.Retrievedfrom
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5307052&isnumber=5307032
[2] Baroudi, U., Al-Roubaiey, A., Mekid, S., &Bouhraoua, A. (2012). The Impact of Sensor Node
Distribution on Routing Protocols Performance: A Comparitive Study. IEEE 11th International
Conference on Trust, Security and Privacy in Computing and Communications (TrustCom),1714-
1720.
[3] Chen, H., Wu, C., Chu, Y., Cheng, C., & Tsai, L. (2007). Energy Residue Aware (ERA) Clustering
Algorithm for Leach-based Wireless Sensor Networks. Conference: Systems and Networks
Communications.
[4] Gajjar, S., Dasgupta, K., Pradhan, S., &Vala, K. (2012). Lifetime improvement of LEACH protocol
for Wireless Sensor Network. 2012 Nirma University International Conference on Engineering
(NUiCONE),1-6.Retrievedfrom
[5] Gambhir, S., & Fatima, N. (2014). Op-LEACH: An Optimized LEACH Method for Busty Traffic in
WSNs. Fourth International Conference in Advanced Computing & Communication Technologies
(ACCT), 222-229.
[6] Gnanambigai, J., Rengarajan, N., &Prarthana, R. (2014). An analytical approach for quadrant based
leach: An energy efficient routing protocol for WSNs. 2nd International Conference on Devices,
Circuits and Systems (ICDCS), 1-6. Retrieved from
[7] Guo, L., Xie, Y., Yang, C., & Jing, Z. (2010). Improvement on LEACH by combining Adaptive
Cluster Head Election and Two-hop transmission.2010 International Conference on Machine Learning
and Cybernetics (ICMLC), 4, 1678-1683.
[8] Halperin, D., Greenstein, B., Sheth, A., &Wetherall, D. (2010). Demystifying 802.11n Power
Consumption. Proceeding HotPower'10 Proceedings of the 2010 International Conference on Power
Aware Computing and Systems, 1(1).
[9] Haiming, Y., &Sikdar, B. (2007). Optimal Cluster Head Selection in the LEACH Architecture. IEEE
International in Performance, Computing, and Communications Conference, 93-100.
[10] Heinzelman, W., Chandrakasan, A., &Balakrishnan, H. (2000). Energy-Efficient Communication
Protocol for Wireless MicrosensorNetworks.Proceedings of the 33rd Hawaii International Conference
on System Sciences.
[11] Huang, J., & Yen, P. (2009). CB-DHRP: A Cluster-Based Directed Hierarchical Routing Protocol for
wireless sensor networks.International Symposium in Autonomous Decentralized Systems, 1-6.
[12] Javaid, N., Waseem, M., Khan, Z., Qasim, U., Latif, K., &Javaid, A. (2013). ACH: Away cluster
heads scheme for Energy Efficient Clustering Protocols in WSNs. Saudi International in Electronics,
Communications and Photonics Conference (SIECPC), 1-4. Retrieved from
[13] Kang, S., & Nguyen, T. (2012). Distance Based Thresholds for Cluster Head Selection in Wireless
Sensor Networks. IEEE COMMUNICATIONS LETTERS, 16(9), 1396-1399.
[14] Kumar, G., Mehra, H., Seth, A., Radhakrishnan, P., Hemavathi, N., &Sudha, S. (2014). An hybrid
clustering algorithm for optimal clusters in Wireless sensor networks. IEEE Students' Conference on
Electrical, Electronics and Computer Science (SCEECS), 1-6. Retrieved from
[15] Li, Y., Yu, N., Zhang, W., & Zhao, W. (2011). Enhancing the performance of LEACH protocol in
wireless sensor networks. IEEE Conference: Computer Communications Workshops.
[16] Liao, Q., & Zhu, H. (2013). An Energy Balanced Clustering Algorithm Based on LEACH Protocol.
Proceedings of the 2nd International Conference On Systems Engineering and Modeling.

28
[17] Maurya, P., Kaur, A., &Choudhary, R. (2014). Behavior analysis of LEACH protocol. International
Conference on Parallel, Distributed and Grid Computing (PDGC), 68-71. Retrieved from
[18] Ramesh, K., &Somasundaram, K. (2011). A COMPARATIVE STUDY OF CLUSTERHEAD
SELECTION ALGORITHMS IN WIRELESS SENSOR NETWORKS. International Journal of
Computer Science & Engineering Survey, 2(4), 153-164.
[19] Ran, G., Zhang, H., & Gong, S. (2010). Improving on LEACH Protocol of Wireless Sensor Networks
Using Fuzzy Logic. Journal of Information & Computational Science, 7(3), 767-775.
[20]Sharma, I., Singh, R., &Khurana, M. (2015). Comparative study of LEACH, LEACH-C and PEGASIS
routing protocols for wireless sensor network. International Conference on Advances in Computer
Engineering and Applications (ICACEA), 842-846. Retrieved from
[21] Sharma, S., &Choudhary, S. (2014). Heterogeneous multi-hop LEACH routing protocol. International
Conference on Green Computing Communication and Electrical Engineering (ICGCCEE), 1-6.
Retrieved from
[22] Schmidt-Eisenlohr, F. (2010). Interference in Vehicle-to-vehicle Communication Networks: Analysis,
Modelling, Simulation and Assessment (p. 149). KIT Scientific Publishing. ISBN: 9783866445710
[23] Tillapart, P., Thammarojsakul, S., Thumthawatworn, T., &Santiprabhob, P. (2005). An Approach to
Hybrid Clustering and Routing in Wireless Sensor Networks. 2005 IEEE in Aerospace Conference, 1-
8.
[24] Tran, S., & Yang, T. (2007). OCO: Optimized Communication & Organization for Target Tracking in
Wireless Sensor Networks. University of Houston.
[25] Tsai, M., & Chen, Y. (2015). A Virtual Cluster Head Election Scheme for Energy-Efficient Routing in
Wireless Sensor Networks. 3rd International Conference in Future Internet of Things and Cloud
(FiCloud), 341-348.
[26] Tyagi, S., Gupta, S., Tanwar, S., & Kumar, N. (2013). EHE-LEACH: Enhanced heterogeneous
LEACH protocol for lifetime enhancement of wireless SNs. International Conference on Advances in
Computing, Communications and Informatics (ICACCI), 1485-1490. Retrieved from
[27] Yan, L., Pan, W., Luo, B., Liu, J., &Xu, M. (2009). Communication protocol based on optical low-
energy-adaptive-clustering-hierarchy (O-LEACH) for hybrid optical wireless sensor networks.
Communications and Photonics Conference and Exhibition (ACP), 1-6. Retrieved from
[28] Zhang, J., Mao, Z., &Jia, R. (2013). Study on hybrid modelling approach for gold cyanidation
leaching process. 25th Chinese in Control and Decision Conference (CCDC), 1557-1561. Retrieved
from http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6561176&isnumber=6560882
[29] Zhao, F., Xu, Y., & Li, R. (2012). Improved LEACH Routing Communication Protocol for a Wireless
Sensor Network. International Journal of Distributed Sensor Networks, 1, 1-6.
[30] Zhao, W., & Zhang, L. (2014). Analysis and research of improved LEACH routing protocol for
pressure sensor based on WSN. Cyberspace Technology (CCT 2014), International Conference, 1-6.
[31] Sukhchandan Randhawa & Sushma Jain (2019). MLBC: Multi-objective Load Balancing Clustering
technique in Wireless Sensor Networks. Journal of Applied Soft Computing. ELSEVIER. Volume 74,
January 2019, Pages 66-89.
[32] Kaswan A., Singh V., Jana P.K. (2018). A multi-objective and PSO based energy efficient path design
for mobile sink in wireless sensor networks. Pervasive Mob. Comput., 46 , pp. 122-136
[33] Amit Sarkar1 & T. Senthil Murugan. (2019). Cluster head selection for energy efficient and delay-less
routing in wireless sensor network. Wireless Netw (2019) 25:303–320. Springer.

AN ENHANCED HYBRID ROUTING AND CLUSTERING TECHNIQUE FOR WIRELESS SENSOR NETWORK

More Related Content

What's hot (18)

Similar to AN ENHANCED HYBRID ROUTING AND CLUSTERING TECHNIQUE FOR WIRELESS SENSOR NETWORK (20)

Recently uploaded (20)

AN ENHANCED HYBRID ROUTING AND CLUSTERING TECHNIQUE FOR WIRELESS SENSOR NETWORK