International Journal of

INTELLIGENT SYSTEMS AND APPLICATIONS IN

ENGINEERING
ISSN:2147-67992147-6799 www.ijisae.org Original Research Paper

A Dynamic Fuzzy Based High-Efficiency Clustering and Intrusion

Prevention Routing Protocol for IOT Based Wireless Sensor Networks.
1
Mr. J. Devasundarraj, 2Dr. M. Arasakumar, 3Dr. C. N. S. Vinoth Kumar
Submitted:14/03/2024 Revised: 29/04/2024 Accepted: 06/05/2024

Abstract: The rapid expansion of the Internet of Things necessitated the development of numerous services, programs, electronic
devices with built-in sensors, and associated protocols, all of which are currently active. With the help of the internet of things, real-life
objects can now see, hear, think, and do remarkable things thanks to improved communication, data sharing, and decision-making
capabilities. The growth of the Internet of Things (IoT) relies heavily on wireless sensor networks (WSNs), which are made up of
inexpensive smart devices that collect data. But there are limits to the amount of processing power, memory, and energy that these smart
gadgets can use. Achieving dependability while protecting transmitted data in a susceptible environment using energy-saving strategies is
one of the main problems for WSN. Wireless sensor networks have a major issue with energy efficiency. Due to their dependence on
batteries, sensor networks eventually fail. In order to alleviate these issues, this article presents the DFHC-IPR protocol for WSN-assisted
IoT systems, which stands for dynamic fuzzy-based high-efficiency clustering and intrusion prevention routing protocol. An energy-
efficient clustering technique called dynamic fuzzy multi-criteria decision making is used, which is a combination of fuzzy AHP with
TOPSIS to find the best cluster heads. An intrusion prevention optimization technique is employed for routing purposes to improve data
delivery security. The proposed method's first step is to use nodes' inherent properties to form various energy-efficient clusters. In
addition, the (k,n) threshold-based Shamir secret sharing system ensures the confidentiality and dependability of the sensory information
between the cluster head and the base station (BS). Performance metrics, including throughput, energy economy, and data security, have
all been enhanced by implementing and comparing this routing protocol with two other existing protocols.

Keywords: necessitated, optimization, Achieving, Performance metrics

Introduction: devices. Radio frequency identification (RFID) systems
primarily consist of tag readers and radio signal
In recent years, the Internet of Things (IoT) has grown
transponders (tags). Typically, radio frequency
rapidly [1], providing limitless applications in numerous
identification tags have data stored electronically,
domains such as smart cities, smart buildings, smart
allowing individuals to categorize, follow, and keep tabs
grids, intelligent transportation systems, healthcare,
on the items. Radio frequency identification tags may be
education, industry, entertainment, and more.[2][3]
affixed to virtually anything in order to track its precise
Internet of Things (IoT) refers to the network of
location and collect data.
interconnected physical devices, services, and
information that will enable communication among A Wireless Sensor Network (WSN) is a system of
machines (M2M) and constitute the "next generation" of interconnected, autonomous sensors that can detect and
the Internet [4]. The fundamental idea behind the Internet report changes in parameters including sound level,
of Things is to enable communication between any two temperature, pressure, and motion. In order to send the
homogenous items, regardless of their location. One of data that they've gathered to a central server or computer,
the first technologies for the Internet of Things (IoT), these sensors often use radio waves for wireless
radio-frequency identification (RFID) [5][6][7] uses communication. Wireless sensor networks (WSNs) have
electromagnetic fields to automatically transmit several uses, such as in healthcare, smart homes,
identifying data to a reader over wireless connection industrial automation, and environmental monitoring [8].
Their versatility and resilience in the face of extreme
1
Research Scholar, Department of Information Technology conditions make them an important component of many
Annamalai University, Chidambaram, Tamilnadu, India. cutting-edge technological systems. The Internet of
Mail jdevasundarraj@gmail.com Things (IoT) and wireless sensor networks (WSNs) are
2
Department of Information Technology Annamalai becoming feasible solutions that are widely employed in
University, Chidambaram, Tamilnadu, India. real-time data gathering and monitoring applications [9].
3
Department of Networking and Communications,School of These applications include automatic irrigation, target
Computing SRM institute of Science and Technology, monitoring, viewing clinical records, tracking landslides,
Kattankulathur, Tamilnadu, India.

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2928
and making forecasts for forest fire and catastrophe Here are the key points of the paper:
management.
In addition, wireless sensor networks (WSNs) that are In order to identify the cluster head, our proposed
based on the Internet of Things are utilized in both intrusion prevention routing protocol for IoT-based
monitored and unmonitored settings, such as monitoring wireless sensor networks combines the TOPSIS
water quality, air pollution, and smart cities, among other approach with dynamic fuzzy based high efficiency
applications [10]. In addition to providing reliable data clustering. Data security is improved with the usage of
forwarding, another important concern is to enhance intrusion prevention routing protocol.
energy efficiency [11]. In the past, a number of
The following is the outline of the paper: Section 2
researchers have developed a cluster-based approach for
details the relevant literature. The approach that was
wireless sensor networks (WSN) in order to successfully
presented is detailed in Section 3. Section 4 provides the
achieve energy efficiency [12]. Clustering techniques
experimental outcomes.
include the nodes being partitioned into distinct regions,
with a single cluster head that is referred to as the leader The conclusion and future scope are included in Sect. 5,
node between the various regions. The purpose of the the last section.
cluster head is to collect the data from the nodes that are Related works:
members of the cluster, aggregate it, and then send it on
to the base station (BS). Either a single hop or a multi- The energy conservation aim is a demanding criterion in
hop technique can be utilized to fulfil the task of most large-scale networks, including the Internet of
transmitting data from cluster heads to BS [13]. Things (IoT) and sensor nodes with limited resources
[18][19]. Cluster heads often serve as a symbolic
In addition, the ESR protocol takes advantage of a governing entity during data transmission and collection
constrained clustering practice to update the cluster in cluster-based networks [20][21].
heads' status with minimal communication cost, as each
cluster is inundated with the least amount of re-election This means that, particularly in terms of network
packets. Secondly, ESR protocol uses a lightweight homogeneity and stability, picking the right cluster heads
secret sharing technique between cluster heads and BS to has a meaningful impact on network performance
provide trustworthiness and safe data routing over the [22][23]. Due to resource constraints, safe data routing
whole network [14]. Each member of the set of cluster is an important component of large-scale IoT-based
heads shares a portion of the secret key that the BS WSNs. The lack of safeguards against malevolent attacks
creates under this method. The suggested secret sharing causes most of the current systems' [24][25] network
approach encrypts incoming packets as they travel communication unreliable and unsafe.
between cluster heads and BS. Likewise, the suggested In WSNs, layered and tiered data transmission is one of
data security solution initiates decryption upon packet the most popular and reliable methods. Following
receipt at the BS from the cluster heads prior to onward extensive testing, this method is today considered to be
transmission to the end user [15]. one of the most effective routing systems for delivering
Early on, the Internet is shifting its focus from data packets to their final destination, such as BS. Layers
connecting people to connecting things, in order to such as CMs, CLHs, and CLGs include all of the sensor
encourage a new way of thinking about the Internet of nodes, and each layer is responsible for a certain
Things (IoT). According to Farman et al. (2018) [16], function.
and Mahajan et al. (2016) [17], this new pattern In large-scale networks with limited resources, like the
introduces items or objects to the Internet and opens up Internet of Things (IoT) and sensor networks (SNs),
new economic opportunities. Everything is operational energy saving is usually the most important
and can become a new source of data for the internet, consideration. Typically, CH is the controlling entity in
from interior wearable appliances to outside natural cluster-based networks [26]. Its primary function is to
sensors. Raising the online components with a collect and send data. Additionally, due to resource
heightened awareness of the here and now. Internet of constraints, securely routing data is the most important
Things (IoT) is an exciting new field with a fresh, component of large-scale WSNs connected with the IoT.
cutting-edge pattern. It is possible to get items online
using a variety of approaches. Two methods to improve Due to a lack of a threat prevention schema, some
WSN energy efficiency in the IoT are routing and current methods fail to deliver dependable and safe data
clustering. routing in the network [27]. There are several steps to a
LEACH protocol, which stands for low-energy adaptive
clustering hierarchy [28]. Using a multi-hop foundation

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2929
for data transmission improves the performance of the A protocol for source location protection in WSNs based
LEACH protocol [29]. on dynamic routing has been proposed for the social
internet of things (Han et al. 2018) [33]. In order to solve
A method for energy-limited heterogeneous fog-
the problem of source zone security, this study suggests
supported wireless sensor networks has been designated
doing research on source region assurance conventions
by Naranjo et al. (2017) [30]. This study presents an
using dynamic coordination. We provide an adaptable
alternative method for selecting CHs in WSNs and
course correction strategy for data transfer. An
organizing progressive nodes. Both the basic and
underlying hub is initially selected at random from the
advanced nodes' energy levels depend on prolong SEP
system's limit by the introduced plot. In order to reach
(P-SEP), which was presented in this study. When it
the sink, the packet will first take an avaricious path and
comes to choosing CHs from among the hubs, P-SEP is
then a coordinated one.
totally logical as it means that every node has an equal
chance of being picked. Our plan can guarantee source territory security and
combat various security disclosure assaults without
The data gathering technique and load balancing that the
hurting the system lifespan, according to both theoretical
authors suggested in [31] are an attempt to reduce energy
and exploratory outcomes.
consumption and boost the network's performance in
terms of data aggregation and forwarding. The suggested A Dynamic fuzzy based high-efficiency clustering and
approach builds a stack of layers according on the Intrusion Prevention routing protocol:
number of hops between nodes and the sink node. On the
The use of clustering and data aggregation is a crucial
other hand, communication costs rise due to the hop-by-
approach to improving energy efficiency using software-
hop nature of the next-hop selection operation. There are
based protocols, which is a critical issue in WSN
more route discoveries and more transmission delays as
applications for energy efficiency. In order to reduce
the network size increases.
energy consumption and increase service quality, this
According to the EEBCDA method [32], which stands article suggests a DFHEC-IPR protocol. We first
for energy-efficient balanced cluster-based data assumed that WSNs have a variety of nodes with
aggregation, the network field is divided into rectangular different types of data. In every single one There are
grids of uneven sizes, each containing one cluster head. various degrees of nodal energy at each node. The
Within each grid, the position of the cluster head is battery life of each sensor node is limited, and they
moved among the member nodes. The knowledge about cannot be recharged. The nodes may merge all the
residual energy is used to choose a small number of duplicate data, which reduces power consumption.
nodes within each grid to serve as cluster heads. Figure 1 explains the general technique that was
However, creating a grid is a more involved process that introduced.
necessitates additional overhead.

Fig.1 Flow of DFHC-IPR Methodology

• Minimize the cost of the intra-cluster communication. The phrase "intra-cluster communication cost" refers to
the total amount of all the individual expenses that are
• Minimize the cost of the inter-cluster communication.
related with the links that connect all of the members of
the cluster to their respective CHs. The overall cost of
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2930
the tree that was formed, which is also referred to as the of the costs of the links that connect with the CHs that
inter-cluster communication cost, is defined as the total comprise that tree.

Fig 2. Decision Making Criteria for Cluster Head Selection

Implementation of the recently proposed system will be has been described, the overall functionality of the
carried out with the assistance of the MATLAB platform. protocol is split down into two major components. These
As far as quality-of-service parameters are concerned, components will be discussed in the subsequent sections
the performance is evaluated and compared with an of this article. In order to guarantee secure data routing
earlier clustering and routing scheme. These parameters inside the first component, the ESR protocol is in charge
include the ratio of packet delivery to packet loss, of organising the optimal hierarchical topology
throughput, network lifetime, end-to-end delay, channel construction and the thresholding-based secret sharing
load, jitter, bit error rate (BER), buffer occupancy, and scheme (SSS) [34]. This is done with the intention of
energy consumption. Concerning the performance, this guaranteeing that the data is routed safely within. The
evaluation and comparison is carried out in relation to optimal cluster heads are determined in conjunction with
the performance. the scattered clusters in a manner that is both balanced
and efficient in terms of energy consumption. In addition
The following section provides a concise summary of the
to the restrictions imposed by the quality of service, this
energy-efficient and secure routing (ESR) protocol that
assessment is informed by a number of other
has been suggested for wireless sensor networks that are
considerations.
based on the Internet of Things. In the ESR protocol that

Fig 3. Block diagram of the energy-efficient and secure routing protocol (ESR) protocol
Optimise Clusters: to function normally. In terms of boundless resources,
the BS 180 is limitless. At first, all of the nodes in the
The initialization step begins by randomly dispersing the
monitoring field were able to receive the 181 hop-by-hop
number of nodes in a square-sized network field. With a
broadcasts of BS's location. In addition, each node's
fixed identifier and few restrictions, each node continues
routing 182 table is updated with the neighbour’s
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2931
information. After that, ESR protocol 183 makes a and stays put. In terms of boundless resources, the BS
dispersed announcement in the network field about the 180 is limitless. At the outset, all the nodes in the field
procedure for selecting the cluster leader. The were able to receive the 181 hop-by-hop broadcasts of
competitive value 185 is calculated for all the nodes by BS's location.
taking into account the residual energy 𝑒𝑖, the Received
𝑛 𝜀𝑓𝑠 𝐷
Signal Strength Indicator (𝑅𝑆𝑆𝐼𝑖)184, the distance from 𝑘=√ √𝜀 ,
2𝜋 𝑚𝑝 𝑥𝐵𝑆𝑆2
the base station 𝑑𝑖 𝑡𝑜 𝐵𝑆, and the queue length
𝑄𝐿𝑖𝑖 variables. In the process of exchanging control XBS is the average distance from all of the sensor nodes
messages, every node learns about its 186 neighbours. to the base station, where n is the number of nodes in the
Firstly, the maximum residual energy of a node is given network and D is its dimension.
188 greater significances since node energy is the most The distances between the sensor nodes and the cluster
important aspect in the network's survivability. Second, centres are calculated using the following equation,
if the RSSI value is higher than a specific threshold, it which makes use of the concept of Euclidean distance.
indicates that the wireless link is performing well and
𝑁
can receive packets at a good pace (189). The RSSI
threshold, as indicated in Equation 1, is calculated by the 𝑋𝑛2𝐶𝐶 = √∑ (𝑋𝑗 − 𝑋𝐶𝐶 ),
𝑗=1
ESR protocol and is equal to the average reception rate
of beacon packets from N neighbours at a specific time
In this case, Xn2CC is the total distance from the node to
period (Δt). A number higher than the 192 calculated
the cluster centre, Xj is the position of the jth sensor
threshold is required for the node's RSSI. Low
node, and XCC is the total of all cluster centre
connection quality, as indicated by an RSSI value below
coordinates.
the threshold, increases the chance of packet loss ratio
(193). To get the value of R, we apply the following Dynamic fuzzy based high-efficiency clustering
equation, where X is the reception rate of beacon 194 This work presents a dynamic clustering strategy that
packets. combines fuzzy PAH and CNN techniques to achieve
𝑋 optimum cluster head selection, resulting in great
RSSI threshold =
𝑁
efficiency. Within In this method, three criteria and six
As a third point, minimising energy consumption and sub-criteria are taken into account. it instantly stimulates
maximising network lifetime are achieved by using the the network's lifespan. The text is indistinct or unclear.
shortest path of node towards BS. Ultimately, the node- PAH is used to determine the weights of each criterion.
level congestion is measured by the queue length factor The CNN technique is used for the purpose of
𝑄𝐎𝑖𝑖, which also increases data delivery performance prioritising the options. There are three. The factors
197. Equation (2) may be used to calculate the queue include energy status, influence on quality of service
length 𝑄𝐿𝑖 of a node i, where 𝑅𝑅𝑖 is the packet receipt in (QoS), and node placement. The number is six. The sub-
bytes and TB is the total buffer size in bytes. criteria consist of residual energy and communication
𝑅𝑅𝑖 cost between. The information includes a node and its
𝑄𝐿𝑖 = neighbouring nodes, the quality of the links between
𝑇𝐵
them, the number of times the node has been restarted,
Last but not least, as demonstrated in Equation (3), all the number of neighbours it has, and the marginality of
factors are averaged using weights, and the nodes with the node. Fig. 2 provides explanations for many criteria.
the greatest competitive value 𝐶v are chosen to serve as
the first cluster heads. The suggested ESR technique Risk Assessment
chooses the best cluster heads according to their intrinsic In this part, we conduct the weighting factor sensitivity
properties, making the resulting clusters more flexible. A analysis for the cluster head selection process in the
normalisation series of [0,1] is applied to the calculated dynamic network architecture. The equation optimises
value of 𝐶v . the cluster heads selection process by applying weighting
1 factors w1, w2, w3, and w4. These variables take into
𝐶v = w1 × 𝑒𝑖 + w2 × 𝑅𝑆𝑆𝐼𝑖 + w3 × ( ) account the values of residual energy, RSSI, proximity
𝑑𝑖 to 𝐵𝑆
+ w4 × 𝑄𝐿𝑖 from BS, and the queue length factor. When calculating
the relative importance of each component in an
Cluster formation: assessment, all of the relevant criteria are considered and
The initialization step begins by randomly dispersing the weighted. For all weighted factors to be included, the
number of nodes in a square-sized network field. With total of their impact fractions must be 1, or 100%.
few exceptions, every node has its own unique identifier
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2932
In order to assess the improved performance of cluster In the initial configuration, w2 = 0.1, w3 = 0.5, and w4 =
head selection, four distinct weighted factor values are 0.1.
employed.
All designs are evaluated based on energy usage, route
Configuration-1 consists of w1 = 0.5, w2 = 0.3, w3 = breakages, and transmission.
0.1, and w4 = 0.1.
Distance and packet delivery performance are used as
Configuration-2 is characterised by w1 = 0.1, w2 = 0.5, benchmarks to get the optimal values for w1. w2, w3,
w3 = 0.3, and w4 = 0.1. and w4.
Configuration-3 is defined by w1 = 0.3. Furthermore, error bars are used as a visual depiction to
illustrate the
Configuration-4, the values are as follows: w1 = 0.2, w2
= 0.1, w3 = 0.1, and w4 = 0.6. Measurement errors or uncertainties in the experimental
findings
NUMBER OF CLUSTER HEADS
0.24
0.22
ENERGY CONSUMPTION

0.20
PER ELECTION(J)

0.18
0.16
0.14
0.12
ESR 0.15 0.17 0.19 0.2 0.21
CCBRP 0.17 0.18 0.19 0.21 0.23
LEACH-MAC 0.18 0.19 0.21 0.23 0.24
LEACH-HR 0.19 0.2 0.22 0.24 0.25

Fig 4. Energy Consumption while cluster head selection

Figure 4 shows that when comparing configurations 2, 3, You can see the results of route breakages for all the
and 4 in terms of energy usage, configuration 1 yields configurations in Figure 5, however configuration-2
superior results, with an average improvement of 31%. performs better than configuration-1, configuration-3,
When sending and receiving cluster head election and configuration-4 by an average of 38%.
packets, the average power consumption ratio between
sensor nodes is used to calculate energy usage.

NUM B ER OF
CLUSTER HEADS
NUMBER OF ROUTE

30 22 24
22
25 19 19
18 20
17
BREAKAGES

16
15
14 16 14 15
20 13
12 12
15 10 8
6
10
5
0
1 2 3 4 5
ESR 6 8 12 14 15
CCBRP 10 14 16 18 20
LEACH-MAC 12 15 17 19 22
LEACH-HR 13 16 19 22 24

ESR CCBRP LEACH-MAC LEACH-HR

Fig 5. Performance of route breakages for all given configurations using different weighting factors
Finally, Figure 6 shows that compared to other
configurations, configuration-4 yields superior results, on
average, in terms of packet delivery ratio, by 37%.

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2933
 NUM B ER OF
CLUSTER HEADS
90

PACKET DELIVERY
88
86

RATIO (%)
84
82
80
78
76
74
72
70
ESR 81 76 75 73 69
CCBRP 87.5 83 80 78 75
LEACH-MAC 85 80 77.5 75 73
LEACH-HR 81 76 75 72 68

Fig 6. Packet delivery ratio for given all configurations with different weights
Experiment findings showed that under different IoT- improve the dependability of data transmission, this
based WSN topologies, each weighted component article takes two goals into account. The goal is to keep
significantly affects the selection of cluster head, the cost of communication inside and between clusters as
according to the performance sensitivity analysis shown low as possible. Data reliability is achieved via an
in figures 4 – 6. optimisation process that is inspired by the immune
system. This expression also optimises the cost of
Remaining power
communication between clusters as well as within them.
When thinking about wireless sensor networks, energy is |𝑉| |𝐶𝑘 |
the most crucial resource to consider. When it comes to
∑ ∑ 𝑤𝑐𝑚𝑚,𝑘→𝐶𝐻𝑘
processing, routing, and aggregating data, nodes that
𝑘=1 𝑚=1
serve as cluster leaders use more power than other
|𝑉|
members of the cluster. This equation calculates the
remaining energy. ∑ 𝑤CH 𝑘→ Next Hop CH ,
𝑘
𝑘=1
𝐸𝑟 = 𝐸0 − 𝐸c
Where, CHk → CH number k; k → Total number of
How much it costs for a node to communicate with its selected CHs; NextHop CH𝑘 → Next hop for
neighbours CH𝑘 ; cm𝑚,𝑘 → Cluster
The energy used by the sending message is exactly member number m of cluster ⬚k ; V→ The vector
proportional to the square of the distance between the comprising the selected CHs; Ck →The vector covering
candidate nodes and the source node. The the cluster members in the cluster that corresponds to
communication cost is defined as the numerical amount CHk
that represents the expense associated with transmitting
information. Experimental results and discussion
2
𝑑avg The presented technique is integrated into the MATLAB
𝐶= ,
𝑑02 program. It determines how well the new method
performs in comparison to the current clustering and
The energy used by the sending message is exactly
routing protocols of MOBFO-EER, FRLDG, TCBDGA,
proportional to the square of the distance between the
and HEED. Using one hundred nodes, we can calculate
candidate nodes and the source node. The
the quality-of-service metrics for the following: end-to-
communication cost is defined as the numerical amount
end latency, energy consumption, channel load, jitter, bit
that represents the expense associated with transmitting
error rate (BER), packet delivery ratio, throughput, and
information.
packet loss ratio. Table 2 displays the remaining
Emergence of issues with routing protocols simulation settings. As shown in Table 3, the parameters
One way to put it is that the clustering and routing issue of the newly implemented routing and clustering
is a problem of minimising many objectives. In order to protocol are examined.
In order to determine how effective the proposed routing
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2934
mechanism is, we use the following performance Delay = Queuing delay + Processing delay + Propagation
measures in conjunction with the simulation results. delay
Packet Delivery Ratio =
𝑁𝑜 𝑜𝑓 𝑝𝑎𝑐𝑘𝑒𝑡𝑠 𝑡𝑟𝑎𝑛𝑠𝑚𝑖𝑡𝑡𝑒𝑑 𝑠𝑢𝑐𝑐𝑒𝑠𝑠𝑓𝑢𝑙𝑙𝑦
X 100
𝑁𝑜 𝑜𝑓 𝑝𝑎𝑐𝑘𝑒𝑡𝑠 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑

PARAMETER VALUE
Area 1000X1000 m
BS Location 1000-1150 m
Number of nodes 100
Initial Energy 0.1J
Bandwidth 20 kbps
Packet size 500 bytes
Node distribution Random values
Total number of nodes in the region 500
Antenna direction Omni-directional
Table 1: Simulation Parameters
No of nodes vs Qos parameters
QoS No Energy Networ Packe End Throughp Packe Bit Channe Buffe Jitte
parameter of consumptio k life- t to ut (Mbps) t loss erro l load r r
s node n (𝑴𝒋 ) time deliv- end ratio r (%) occu- (𝑀𝑆 )
s (rounds ery dela (%) rate pancy
) ratio y (%) (%)
(%) (sec)
DFHC- 100 45 5600 100 2.2 0.98 0 28 28 26 0.38
IPR
200 70 5300 99 3.8 0.88 1 25 24 24.5 0.36
(Proposed
) 300 97 5100 98 4.9 0.75 2 22.5 22 21 0.35
400 110 4950 97 5.5 0.62 3 18 21 19 0.33
500 145 4800 96 6.1 0.59 4 20 20 18 0.31
MOBFO- 100 55 5200 99 2.9 0.77 1 24 26 23 0.55
EER
200 80 4900 98 4.8 0.65 2 26 24 15 0.49
300 107 4800 97 6.0 0.60 3 18 23 20 0.49
400 140 4500 96 7.5 0.54 4 17 20 19 0.47
500 165 4300 95 8.9 0.51 5 17.5 17 20 0.44
TCBDGA 100 65 4900 98 4.0 0.88 2 19 15.5 19.5 0.65
200 105 4800 97 5.3 0.71 3 18 25 25 0.62
300 140 4600 96 6.2 0.68 4 11 19 26 0.61
400 158 4200 95 8.2 0.62 5 10.5 18.5 27 0.6
500 183 4000 94 9.1 0.52 6 14 14 22 0.55
HEED 100 135 4500 97 5.8 0.44 3 8 14 24.5 0.6

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2935
 200 159 4300 96 6.2 0.75 4 9 24 25 0.59
300 180 3900 95 7.4 0.69 5 14 19 19 0.57
400 210 3500 94 8.0 0.60 6 20.5 18 17.5 0.56
500 250 3300 93 9.5 0.50 7 24 16 15 0.5
FRLDG 100 143 4200 96 6.3 0.44 4 17 14 14.5 0.7
200 170 4000 95 7.0 0.65 5 15.2 20 11 0.69
300 195 3800 94 8.8 0.58 6 14 21 9 0.65
400 240 3500 93 9.2 0.44 7 11 18 8 0.6
500 275 3200 92 10.1 0.32 8 12 17 9 0.53
Table 2. Results obtained for Number of nodes versus QoS Parameters

NETWORK SIZE (Number of Nodes)

DFHC-IPR (Proposed) MOBFO-EER TCBDGA HEED FRLDG

300
ENERGY CONSUMPTION

250
200
150
100
50
0
No of
100 200 300 400 500
nodes
DFHC-IPR (Proposed) 45 70 97 110 145
MOBFO-EER 55 80 107 140 165
TCBDGA 65 105 140 158 183
HEED 135 159 180 210 250
FRLDG 143 170 195 240 275

Fig 7. Performance of Energy Consumption

The results of the energy usage are shown in Figure 7. MOBFO-EER (55 𝑀𝑆 ), FRLDG (65 𝑀𝑆 ), TCBDGA (135
There is a range of 0 to 300 for energy usage. Compared 𝑀𝑆 ), and HEED (143 𝑀𝑆 ) all use a considerable amount
to previous techniques, the one that was proposed utilises of energy in a network of 100 nodes.
45 𝑀𝑆 of energy in 100 nodes. Present methods

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2936
 1.2
NETWORK SIZE
1

THROUHPUT
0.8
0.6
0.4
0.2
0
5 6 7 8 9
No of nodes
100 DFHC-IPR (Proposed) 0.98 0.88 0.75 0.62 0.59
200 MOBFO-EER 0.77 0.65 0.6 0.54 0.51
300 TCBDGA 0.88 0.71 0.68 0.62 0.52
400 HEED 0.44 0.75 0.69 0.6 0.5
500 FRLDG 0.44 0.65 0.58 0.44 0.32

Fig 8. Throughput Performance

The comparison of network lifespan is shown in Figure rounds, FRLDG uses 4800 rounds, TCB-DGA uses 4300
8. As can be seen from the graph, the proposed Fuzzy rounds, and HEED uses 4100 rounds, the presented
logic method outperforms the alternatives in terms of method makes advantage of a wide network lifespan of
network lifespan. Whereas MOBFO-EER uses 5000 5500 rounds over 100 nodes.

30 PERFORMANCE OF PACKET DELIVERY RATIO

PACKET DELIVERY RATIO

25
20
15
10
5
0
No of
100 200 300 400 500
nodes
DFHC-IPR (Proposed) 28 25 22.5 18 20
MOBFO-EER 24 26 18 17 17.5
TCBDGA 19 18 11 10.5 14
HEED 8 9 14 20.5 24
FRLDG 17 15.2 14 11 12

Fig 9. Performance of Packet delivery ratio

Figure 9 displays the results of the packet delivery ratio method achieves an impressive packet delivery ratio of
for both the current method and the newly introduced 99%. There are four current methods that provide packet
method (DFHC-IPR). As the number of nodes increases, delivery ratios of 98% in 100 nodes: MOBFO-EER, 97%
the simulation shows that the packet delivery ratio in FRLDG, 95% in TCBDGA, and 93% in HEED.
decreases. In a network with 100 nodes, the presented

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2937
 Network size

30

BUFFER OCCUPANCY (%)

25
20
15
10
5
0
No
of
100 200 300 400 500
node
s
DFHC-IPR (Proposed) 26 24.5 21 19 18
MOBFO-EER 23 15 20 19 20
TCBDGA 19.5 25 26 27 22
HEED 24.5 25 19 17.5 15

Fig 10. Performance of Buffer Occupancy

Figure 10 depicts the buffer occupancy performance of is superior in terms of efficiency compared to existing
the new (DFHC-IPR) and current techniques. The energy-efficient protocols. The buffer occupancy
simulation demonstrates that the suggested method percentages of the current techniques MOBFO-EER,
successfully achieved a high buffer occupancy of 26% FRLDG, TCBDGA, and HEED are 23%, 20%, 18%, and
over 100 nodes. Hence, our newly implemented protocol 15% respectively.

Netwo r k Size (Nu m b er of n od es)

0.8
0.7
0.6
JITTER (MS)

0.5
0.4
0.3
0.2
0.1
0
100 200 300 400 500
DFHC-IPR (Proposed) 0.38 0.36 0.35 0.33 0.31
MOBFO-EER 0.55 0.49 0.49 0.47 0.44
TCBDGA 0.65 0.62 0.61 0.6 0.55
HEED 0.6 0.59 0.57 0.56 0.5
FRLDG 0.7 0.69 0.65 0.6 0.53

Fig 11. Performance of Jitter (𝑀𝑆 )

Figure 11 illustrates the comparison of jitter between our algorithm for the Internet of Things (IoT). An advanced
newly developed DFHC-IPR protocols and the old sensor network has been implemented. A novel
protocols. The findings indicate that our introduced clustering technique, called dynamic fuzzy-based high
method achieved a low jitter value of 0.38 𝑀𝑆 over 100 efficiency clustering, has been proposed. This approach
nodes. Hence, the newly developed protocol surpasses selects a cluster head among the most eligible nodes
previous current energy protocols in terms of efficiency. based on critical parameters related to energy status. ESR
The jitter values for the current protocols MOBFO-EER, enhanced the selection process of cluster heads and used
FRLDG, TCB-DGA, and HEED are 0.55 𝑀𝑆 , 0.65 𝑀𝑆 , a distributed strategy to create clusters that ensure equal
0.6 𝑀𝑆 , and 0.7 𝑀𝑆 , respectively. distribution of energy usage. Furthermore, the cluster
head selection strategy has the capability to decrease the
Conclusion
amount of traffic and save energy. Most of the earlier
This work presents a dynamic fuzzy-based high- solutions used a greedy algorithm to design the routing
efficiency clustering and intrusion prevention routing
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2938
route, but failed to consider intrusions in an with Items in a Controlled area. U.S. Patent
infrastructure-less and unsupervised environment. This 9,773,136, 26 September 2017.
leads to a significant increase in the number of route [7] Miorandi, D.; Sicari, S.; De Pellegrini, F.;
discoveries and re-transmissions, especially in situations Chlamtac, I. Internet of things: Vision, applications
when there are a large number of malicious nodes and a and research challenges. Ad Hoc Netw. 2012, 10,
high network load. The intrusion prevention routing 1497–1516. [CrossRef]
algorithm is used for the purpose of routing. In addition, [8] Hong Z, Pan X, Chen P, Su X, Wang N, Lu W
the ESR protocol included a lightweight secret sharing (2018) A topology control with energy balance in
system between cluster heads and BS to provide safe underwater wireless sensor networks for IoT-based
network-wide data routing and protect against rogue application. Sensors 18:2306
nodes. The MATLAB software package is used for [9] Sharma, N.; Sharma, A.K. Cost analysis of hybrid
simulating scenarios, and the data is obtained via real- adaptive routing protocol for heterogeneous
time data transfer. The performance results are graphed wireless sensor network. Sadhan ¯ a¯ 2016, 41,
based on QoS parameters, including packet delivery ratio 283–288. [CrossRef]
(99%), packet loss ratio, throughput (0.95Mbps), [10] Wang, K.; Wang, Y.; Sun, Y.; Guo, S.; Wu, J. Green
network lifetime (5500 rounds), end-to-end delay, industrial Internet of things architecture: An energy-
channel load, jitter, bit error rate (BER), buffer efficient perspective. IEEE Commun. Mag. 2016,
occupancy, and energy consumption (45 𝑴𝒋 ), in 54, 48–54. [CrossRef]
comparison to existing routing protocols. In the future, [11] Abo-Zahhad, M.; Ahmed, S.M.; Sabor, N.; Sasaki,
the goal is to discover the most efficient path with a S. Mobile sink-based adaptive immune energy-
significant amount of remaining energy for multi-hop efficient clustering protocol for improving the
inter-cluster communication. This involves strategically lifetime and stability period of wireless sensor
deploying many base stations in close proximity to a networks. IEEE Sens. J. 2015, 15, 4576–4586.
single base station in order to reduce the workload on [CrossRef]
cluster heads. This approach will enhance energy [12] Batra, P.K.; Kant, K. LEACH-MAC: A new cluster
efficiency and prolong the total lifespan of the network head selection algorithm for wireless sensor
in a wireless sensor network (WSN). networks. Wirel. Netw. 2016, 22, 49–60. [CrossRef]
[13] Chen, G.; Li, C.; Ye, M.; Wu, J. An unequal cluster-
References:
based routing protocol in wireless sensor networks.
[1] Kharrufa H, Al-Kashoash HA and Kemp AH. Wirel. Netw. 2009, 15, 193–207. [CrossRef]
RPLbased routing protocols in IoT applications: a [14] Hassanabadi, B.; Shea, C.; Zhang, L.; Valaee, S.
review. IEEE Sens J 2019; 19(15): 5952–5967. Clustering in vehicular ad hoc networks using
[2] Lin JW, Chelliah PR, Hsu MC, et al. Efficient affinity propagation. Ad Hoc Netw. 2014, 13, 535–
faulttolerant routing in IoT wireless sensor 548. [CrossRef]
networks based on bipartite-flow graph modeling. [15] Xiong, Z.; Guo, T.; Xue, Z.; Cai, W.; Cai, L.; Luo,
IEEE Access 2019; 7: 14022–14034. N. Online energy-efficient deployment based on
equivalent continuous DFS for large-scale web
[3] Ercan AO¨, Sunay MO and Akyildiz IF. RF energy
cluster. Clust. Comput. 2018, 22, 583–596.
harvesting and transfer for spectrum sharing
[CrossRef]
cellular IoT communications in 5G systems. IEEE
[16] Farman H, Jan B, Javed H, Ahmad N, Iqbal J,
T Mobile Comput 2017; 17(7): 1680–1694.
Arshad M, Ali S (2018) Multi-criteria based zone
[4] Ahmed E, Yaqoob I, Gani A, et al. Internet-of- head selection in internet of things based wireless
thingsbased smart environments: state of the art, sensor networks. Future Gener Comp Syst 87:364–
taxonomy, and open research challenges. IEEE 371
Wirel Commun 2016; 23(5): 10–16. [17] Mahajan S, Dhiman PK (2016) Clustering in WSN:
a review. Int J Adv Res Comput Sci 7(3):198–201
[5] Downie, J.D.; Nederlof, L.; Sutherland, J.S.;
Wagner, R.E.; Webb, D.A.; Whiting, M.S. Radio [18] Batra, P.K.; Kant, K. LEACH-MAC: A new cluster
Frequency Identification (RFID) Connected Tag head selection algorithm for wireless sensor
Communications Protocol and Related Systems and networks. Wirel. Netw. 2016, 22, 49–60.
Methods. U.S. Patent No. 9,652,707, 16 May 2017. [CrossRef]
[6] Koch, M.J.; Swope, C.B.; Bekritsky, B.J. System
[19] Hassanabadi, B.; Shea, C.; Zhang, L.; Valaee, S.
for, and Method of, Accurately and Rapidly
Clustering in vehicular ad hoc networks using
Determining, in Real-Time, True Bearings of Radio
affinity propagation. Ad Hoc Netw. 2014, 13, 535–
Frequency Identification (RFID) Tags Associated
548. [CrossRef]
International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2939
[20] Bian, X.; Liu, X.; Cho, H. Study on a cluster-chain [27] Deebak, B.D.; Al-Turjman, F. A hybrid secure
routing protocol in wireless sensor networks. In routing and monitoring mechanism in IoT-based
Proceedings of the 2008 Third International wireless sensor networks. Ad Hoc Netw. 2020, 97,
Conference on Communications and Networking in 102022. [CrossRef]
China, Hangzhou, China, 25–27 August 2008; pp.
[28] Yang, T.; Xiangyang, X.; Peng, L.; Tonghui, L.;
964–968.
Leina, P. A secure routing of wireless sensor
[21] Enam, R.N.; Qureshi, R.; Misbahuddin, S. A networks based on trust evaluation model. Procedia
uniform clustering mechanism for wireless sensor Comput. Sci. 2018, 131, 1156–1163. [CrossRef]
networks. Int. J. Distrib. Sens. Netw. 2014, 2014,
[29] Safara, F.; Souri, A.; Baker, T.; Al Ridhawi, I.;
1–14. [CrossRef]
Aloqaily, M. PriNergy: A priority-based energy-
[22] Das, S.K.; Tripathi, S.; Burnwal, A. Intelligent efficient routing method for IoT systems. J.
energy competency multipath routing in wanet. In Supercomput. 2020, 76, 8609–8626. [CrossRef]
Information Systems Design and Intelligent
[30] Selvakumar, K., & Revathy, G. (2018). Channel
Applications; Springer: Heidelberg, Germany,
assinment using tabu search in wireless mesh
2015; pp. 535–543.
networks. Wireless Personal Communications, 100,
[23] Latif, K.; Ahmad, A.; Javaid, N.; Khan, Z.; Alrajeh,
1633-1644.
N. Divide-and-rule scheme for energy efficient
[31] Guan, X.; Guan, L.; Wang, X.G.; Ohtsuki, T. A
routing in wireless sensor networks. Procedia
new load balancing and data collection algorithm
Comput. Sci. 2013, 19, 340–347. [CrossRef]
for energy saving in wireless sensor networks.
[24] Lou, C.; Zhuang, W. Energy-efficient routing over Telecommun. Syst. 2010, 45, 313–322. [CrossRef]
coordinated sleep scheduling in wireless ad hoc [32] Yuea, J.; Zhang, W.; Xiao, W.; Tang, D.; Tang, J.
networks. Peer-to-Peer Netw. Appl. 2016, 9, 384– Energy efficient and balanced cluster-based data
396. [CrossRef] aggregation algorithm for wireless sensor networks.
Procedia Eng. 2012, 29, 2009–2015. [CrossRef]
[25] Rani, S.; Talwar, R.; Malhotra, J.; Ahmed, S.H.;
[33] Han G, Zhou L, Wang H, Zhang W, Chan S (2018)
Sarkar, M.; Song, H. A novel scheme for an energy
A source location protection protocol based on
efficient Internet of Things based on wireless
dynamic routing in WSNs for the social internet of
sensor networks. Sensors 2015, 15, 28603–28626.
things. Future Gener Comp Syst 82:689–697
[CrossRef]
[34] Shamir, A. How to share a secret. Commun. ACM
[26] Airehrour, D.; Gutierrez, J.; Ray, S.K. Secure 1979, 22, 612–613. [CrossRef]
routing for internet of things: A survey. J. Netw.
Comput. Appl. 2016, 66, 198–213. [CrossRef]

International Journal of Intelligent Systems and Applications in Engineering IJISAE, 2024, 12(4), 2928–2940 | 2940