Performance Analysis of Routing Protocol Using Trust-Based Hybrid

FCRO-AEPO Optimization Techniques

S. MOHAN1*, P. VIMALA2

1*Department of Electronics and Communication Engineering, FEAT, Annamalai University,

Annamalai Nagar, Chidambaram, Tamil Nadu, INDIA.

2Department of Electronics and Communication Engineering, Faculty of Engineering and

Technology, Annamalai University, Annamalainagar-608 002, INDIA.

Abstract: Mobile Ad hoc Networks (MANETs) offer numerous benefits and have been used in

different applications. MANETs are dynamic peer-to-peer networks that use multi-hop data transfer

without the need forpre-existing infrastructure. Due to their nature, for secure communication of

mobile nodes, they need unique security requirements in MANET. In this work, a Hybrid Firefly

Cyclic Rider Optimization (FCRO) algorithm is proposed for Cluster Head selection, it efficiently

selects the cluster head and improves the network efficiency. The Ridge Regression Classification

algorithm is presented in this work to detect the malicious nodes in the network and the data is

transmitted using trusted Mobile nodes for the QoS performance metric improvement. A trust-based

routing protocol is introduced using the Atom Emperor Penguin Optimization (AEPO) algorithm, it

identifies the best-forwarded path to moderate the routing overhead problem in MANET. The

proposed method is implemented using Matlab software and the performance metrics are packet

delivers ratio, packet loss ratio, routing overhead, throughput, end-to-end delay, transmission delay,

network lifetime, and energy consumption. The proposed AEPO algorithm is compared with the

existing PSO-GA, TID-CMGR, and MFFA. The AEPO algorithm’s performance is approximately

1.5%, 3.2%, 2%, 3%, and 4% higher than the existing methods for packet delivers ratio, packet loss

ratio, end-to-end delay, and throughput and network lifetime. This evaluation enables the sender

nodes to improve their data transmission rates and minimizes the delay. Additionally, the suggested

technique has a clear benefit in terms of demonstrating the genuine contribution of distinct nodes to

trust evaluation.

Keywords: Mobile Ad Hoc Networks, Cluster Head Selection, Hybrid Firefly Cyclic Rider

Optimization, Malicious Node Detection, Ridge Regression Classification Algorithm, Atom Emperor

Penguin Optimization.

Received: May 29, 2022. Revised: May 13, 2023. Accepted: June 22, 2023. Published: July 20, 2023.

1. Introduction

Mobile Ad Hoc Networks (MANETs) are self-

contained networks made up of wireless

mobile nodes that operate independently of

any infrastructure [1]. Over radio frequencies,

Nodes in a MANET can dynamically and

freely interact with another node. In the

absence of fixed infrastructure, MANETs

enable mobile users to interact with one

another [2]. Due to the transmission

interference, mobility, and external noise to

work correctly, the routes in the MANETS are

frequently unable. In a diversity of vital

applications, the advancement of the internet

in recent years has greatly increased the use of

MANETs. In order to construct Internet of

Things (IoT)-based smart networks, MANETs

have recently been used [3]. Consequently, for

these networks, the consistency and safety

standards should be thoroughly re-evaluated.

MANET is widely used in the military,

commercial, and private sectors. MANETs are

subject to the variability of security threats,

including rushing attacks, black holes, and

wormholes [4]. In MANET, secure

communication is achieved by using several

traditional approaches. Nevertheless, in terms

of network security, QoS, wireless time-

varying networks and frequent node mobility

are insufficient to ensure efficient transmission

[5]. For instance, the presence and

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

75

Volume 22, 2023

collaboration of malicious nodes in the

network may disrupt the routing process,

leading to a malfunctioning of the network

operations. In order to ensure efficient

transmission and to avoid the malicious attack,

a trust-based routing protocol is used.

MANET Security

In MANETs, security often entails protecting

and maintaining data integrity and

confidentiality, as well as each network node

supplied the availability of network services

and legitimate usage [6]. In the route

discovery processes, the ability of nodes to

actively participate and to honestly forward

data packets to other nodes in the network is

essential to the MANET's feasibility. In order

to deal with data forging and hacking attacks,

a range of security measures have been

developed, such as message encryption

algorithms, secure authentication, message

integrity verification, and so on [7]. Moreover,

in many other attacks like denial of service,

node capture attacks, etc., these solutions are

ineffective. The resistance efficiency is lower

but by node captured causes the internal

attacks and external attacks are effectively

resisted by the traditional security mechanisms

[8]. Data must be transmitted through a node

inside the communication scope of forwarding

nodes to ensure communication security.

Consequently, the data transmission will be

more secure. Accordingly, for effective

transmission, a clustering approach based on

trust evaluation was implemented in [9] for the

identification of the malicious node. The

proposed technique does not take into account

QoS measurements. Subsequently, by

detecting malicious nodes, the secure network

communications performance is improved by

suggesting a new trust formation technique

[10]. Accordingly, based on trust level, link

quality, and geographical position, three

different measures for next-hop selection are

added in this system, allowing nodes to choose

more experienced next-hop forwarders. In

order to find the trusted nodes, the developed

approach ignores node characteristics.

Trusted Node Communication Security

The most difficult issue in MANET is

providing efficient and secure communication

[11]. Nodes in a geological site developed

clusters to provide good communication.

MANET can be managed more efficiently by

dividing the entire network into clusters.

Clustering is the process of forming clusters.

In the network, the communication among the

constrained nodes provides better with the

help of clustering. Gateway nodes, Cluster

Heads (CH), and cluster members made each

cluster [12]. Numerous sorts of research are

being conducted in this area and many

clustering strategies have been developed;

nevertheless, the MANET sustainable

clustering methodology has yet to be

established. Many properties of nodes can be

clustered, including the weight of the nodes,

the trustworthiness of the nodes, mobility of

the node, Node ID, spatial and temporal

locations, and so on [13]. According to the

numerous properties of the node considered at

once, clustering based on weight will be the

most efficient of all the strategies. The node’s

weight is made up of its velocity, transmission

ranges, residual energy, degree, and other

factors [14]. Among all nodes in the cluster,

the chosen CH will be the most efficient

because almost all node characteristics are

taken into account when calculating weights.

In the future of ubiquitous devices, MANET

routing protocols are a critical component.

Increased bandwidth utilization, better

throughput, longer network life, and lower

end-to-end delay are the number of benefits of

the multipath approach. Consequently, route

failures are protected and network congestion

is reduced [15]. There are routing protocols

utilized for better energy efficiency and delay

management in Mobile Adhoc Networks.

Classification of these protocols can be done

based on their method of routing from source

to destination. Depending on the congestion of

traffic, node density and mobility rate such

routing protocols are categorized.

Accordingly, by maintaining both trust and

energy efficiency, a trust-based routing

protocol is developed is the most difficult task

in developing a trust evaluation method for

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

76

Volume 22, 2023

MANET [16]. The remaining part of the work

is organized as follows, section 2 portrays the

literature survey of the study, and section 3

exhibits the problem definition and motivation

of the research. Section 4 illustrates the

proposed research methodology, section 5

elucidated the experimentation and result in a

discussion, and section 6 exposes the research

conclusion.

2. Literature Survey

The literature survey is based on the study of

secure communication of MANET using

different algorithms. The survey portrays the

routing methods for secure transmission and in

this research, an Atom Emperor Penguin

Optimization (AEPO) algorithm is proposed

for finding the best forwarding path,

respectively.

In order to safeguard transmitted packets

from malicious nodes, a dual cluster head-

based trust aware mechanism is proposed by

Aruna Subramanian et al [17] as an alternative

to cryptographic techniques. TWCBRP is a

proposed protocol that splits the network into

one-hop overlapping clusters with primary and

secondary CH in charge of all routine

activities. Replacing primary CH with

secondary CHs, also guarantees the CH’s

trustworthiness, once the former turns

malicious. Cluster members ensure a secure

channel by routing packets exclusively

through gateway nodes and trustworthy CH.

When compared to a distributed weighted

cluster-based protocol (CBPMD), TWCBRP

shows improved performance in terms of

control overhead, packet delivery ratio (PDR),

throughput, and delay when tested with

Network Simulator.

The Hybrid Ant Colony Optimization with

DSR protocol (H-ACO-DSR) method is

presented by M. Anugrahaet al [18] for better

resource allocation in the MANET context to

increase safe data transfer. In the MANET

system, Hybrid Trust Cluster-based Multiple

Routing (H-TCMR) produces trustworthiness

and efficient clusters. Accordingly, to

determine trusted nodes and the malicious

node in the system utilizing the Adaptive

Booting Technique (ABT). The data

transmission, as well as the MANET’s trusted

nodes, are carried out with the trustworthy MN

to enhance the QoS metric’s efficiency.

Finally, modelling results produce Routing

overhead, Throughput, Average Delay,

enhanced PDR, and low-cost Energy when

compared to different techniques.

M. Venkat Das et al [19] advocated using

the “Node Authentication and Trusted Routing

method (NATR)” to improve security.

Through output data delivery and better

security, NATR strives to eliminate aberrant

node interference in MANET. Additionally,

by examining the three most typical actions

taken by a node throughout the connection

process, the predictability of a node is

determined. When it comes to custom network

security, node licencing is crucial. Monitor the

data success rate node trust, RREQ, and

RREP’s success rates using this method. The

loss or drop of packets and the successful

delivery of packets are used to assess data

delivery dependability. Routing overhead

decreased by 40% and package delivery

increased by 25% as depicted based on the

experimental results. In order to evaluate

Adhoc network efficiency, NATR is compared

to AODV and SAR TMS.

Sirajuddinet al [20] proposed a trust-based

multipath routing protocol called TBSMR to

enhance the MANET’s overall performance.

The main strength of the proposed protocol is

that it considers multiple factors like

congestion control, packet loss reduction,

malicious node detection, and secure data

transmission to intensify the MANET’s QoS.

The performance of the proposed protocol is

analyzed through the simulation in NS2. Our

simulation results justify that the proposed

routing protocol exhibits superior performance

to the existing approaches.

Hassan Jariet al [21] examine the effects of

a black hole attack on MANET routing

systems. The goal of trust management is to

keep a network safe from hostile activity.

Accordingly, a new trust-based MANET

routing protocol called ITAODV is presented

in this study, which is developed from the

normal AODV protocol. Accordingly, for

packet forwarding, the proposed protocol

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

77

Volume 22, 2023

employs an indirect trust mechanism that

considers the reliability of each node. The

network simulator NS-3 was utilized for the

ITAODV and AODV performance evaluation.

The ITAODV protocol’s effectiveness against

the Blackhole attack is demonstrated by the

experimental results.

C. Gopala Krishnan et al [22] concentrated

on detecting unstable CH and replacing them

with nodes that use the self-configurable

cluster technique. To successfully designate

CH, a self-configurable cluster method is

presented with the k-means protocol

technique. Periodic irregular CH rotations or

changing the number of clusters are used in

the suggested k-means approach. Additionally,

to avoid and detect MANET vulnerabilities,

this research, offers a trust management

technique. Accordingly, the trust management

method should only employ local data because

of the limited resources (computing, power,

and bandwidth) and the constantly changing

topology. Consequently, the suggested

approach with the k-means procedure and its

experimental findings use less power than

previous protocols and provide an optimal

system for trust management.

In order to train the trust prediction model,

JasleenKauret al [23] suggested using the

Adaptive neuro-fuzzy inference system

(ANFIS) trust management. The hyper-

parameters of the ANFIS model are then tuned

using a non-dominated sorting genetic

algorithm-III (NSGA-III). A fitness function

with many objectives is designed using root

means squared error, precision, and recall

measurements. Subsequently, for comparative

analysis, the optimized link state routing

(OLSR) protocol is used. In order to collect

the dataset, three separate attacks are used on

the designed network: jellyfish, link spoofing,

and grey hole attacks. According to a

comparison of performance measures, the

suggested trust assessment model beats

competitive trust evaluation models in terms

of routing overheads, throughput, average end-

to-end latency, and PDR. Consequently, the

suggested protocol is more resistant to

different security risks.

In MANET, J. Anitha Josephine et al [24]

proposed Tanimoto Support Vector

Regression Based Corrective Linear Program

Boost Classification (TSVR-CLPBC) as an

ensemble approach. In order to improve secure

communication with a higher PDR and

minimal end-to-end delay is the major goal of

the suggested method. In order to investigate

node characteristics such as node history,

cooperative communication, and residual

energy, Tanimotokernelized SVR is first used

as a weak learner. Based on the analysis, the

nodes are classed as malicious or trusted.

Consequently, for secure data transfer, the

trusted MN is identified by the Linear

Program Boost ensemble classifier using ‘n’

number of weak learners (i.e., base classifier).

Accordingly, for secure routing in

MANET, P. Sathyarajet al [25] suggested a

real-time secure route analysis (RSRA)

technique. The technique takes into account

not only the strategy of intermediary nodes

along the detected route but also the presence

of IoT devices and their trustworthiness.

Subsequently, by generating a list of possible

paths between any two points, the method

starts. Subsequent, the trustworthiness of each

mobile node is confirmed by taking into

account its energy, the number of

transmissions involved, mobility speed, its

neighbour list, location, etc. Based on their

previous contributions to the network, the IoT

devices’ trustworthiness is determined. Due to

the mobile nodes, the approach measures

mobile node secure route support (MSRS),

whereas, for IoT devices, it measures device

support (DS). Accordingly, by taking into

account the number of IoT devices in the

route, the approach calculates the data

forwarding support (DFS) value. A single

route has been chosen based on the DFS

measure, which enhances MANET QoS.

Trust and trust computations were

discussed by Rakesh Kumar et al [26]. In

order to limit the effects of attacks, a trust-

based fuzzy bat (TBF) optimization model is

suggested and implemented in this paper.

Subsequently, carried out the sensitivity

analysis in different network scenarios. The

evaluation is based on performance indicators

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

78

Volume 22, 2023

such as PDR and normalized routing

overhead. The distrust threshold, trust update

interval, and trust component weight are all

changed. In an untrusted Manet’s

environment, a security solution is guaranteed

by employing a fuzzy bat with a trust

evaluation model. The TFB algorithm

outperforms existing approaches in terms of

network lifetime and throughput, as well as

end-to-end delay and energy consumption as

revealed in the results.

3. Research Problem Definition and

Motivation

MANET is a hotspot for study because of its

numerous drawbacks and benefits. Providing

secure communication between mobile nodes,

dealing with misbehaviour and location

updates, lowering overhead, and recognizing

node positions are all difficult problems in ad-

hoc networks, subsequently, in this network

trust methods are essential. Over self-

organized networks, MANET supports several

basic operations such as packet forwarding,

routing, communication, and network

administration. Since mobile nodes enter and

exit the network at unpredictable intervals,

MANET does not have a fixed topology.

Consequently, it affects the network’s memory

computations, energy, and bandwidth.

Identifying compromised, malevolent, and

selfish nodes that have been authorized

requires trust management. Sensor nodes, in

reality, have limited resources and other

unique characteristics, making WSN trust

management more important and difficult.

Until now, to improve robustness and security,

research on WSN trust management systems

has mostly concentrated on node trust

evaluation. Subsequently, MANET lacks a

centralized infrastructure, and establishing

trust is a critical task. The multi-hop module

and one-hop module are defined by the

distributed and adaptive trust metrics for

MANET. Recommendation and direct trust are

calculated in the one-hop module, whereas

indirect trust is calculated in the multi-hop

module. Energy trust and communication trust

are examples of direct trust. When assessing

communication trust, not only look at the

present value; predict it based on the

network’s status. The presence of misbehaving

nodes in the network’s numerous problems.

Owing to the waste of important resources can

cause the network to go down. Accordingly,

this encourages the search for a secure

communication environment.

Effective routing with enhanced QoS

become one of the major research challenges

in networks. The process of clustering nodes is

considered one of the major solutions for

scaling down ad-hoc networks and enhancing

effective routing. Hence, finding a suitable

procedure of cluster formation for efficient

routing in the network topology acts as a

primary concern for researchers. Thus, QoS-

aware security-based clustering in such

network act as a challenging and popular area

of research. MANETs suffer from security

issues due to the open and dynamic nature of

the network environment. In other words,

MANET is vulnerable to attacks caused by

malicious nodes but the dynamic nature of the

ad-hoc network plays a vital role in

developing secure and stable routing in

MANET. In addition, clustering can improve

routing performance and enhance secure route

connectivity between nodes. Consequently, the

lifetime of wireless nodes is considered very

important in MANETs, so optimization of

energy utilization plays a vital role in

MANETs. Transmission power, reception

power, and energy consumption by devices are

significant energy utilization constraints in a

MANET. This scheme is introduced to

increase the network lifetime, PDR and its

forwarding rates in the network to establish a

connection between two hops in a multi-hop

approach. When the communication nodes'

distance increases in Mobile Ad Hoc Network,

energy or power utilization also increases and

decreases the wireless node’s whole lifetime.

When the packets are forwarded to the

neighbour node in Mobile Ad Hoc Network,

the power utilization increases, and the node

consumes more power while it transmits or

receives the information. The power

management technique may be accustomed to

optimize the power within the Mobile Ad Hoc

Network. From this, a recent algorithm is

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

79

Volume 22, 2023

proposed for efficient routing and cluster head

selection, and network lifetime improvement,

respectively.

4. Proposed Research Methodology

MANET is a wireless network made up of

several mobile nodes that self-heal and self-

configure without the need for a fixed

infrastructure. Due to their self-configuring

nature, MANETs may be easily accepted in

different situations, including battlefield

communications, rescue, and emergency

operations. Due to the unreliability of

MANET wireless communications, nodes are

susceptible to a variety of security attacks,

which disrupt the network structure. Due to

the malicious node, a Mobile Node (MN) may

provide erroneous routing information to

subsequent nodes, and it drops the DP instead

of forwarding it. So, these challenges are

overcome and the network security is

improved by deploying trust-based node

evaluation techniques. The architecture of the

proposed work is portrayed in figure 1.

Figure 1: Flow Diagram of the Proposed

Work

In this MANET, the source node is

transferred to the destination node, initially, a

Hybrid Firefly Cyclic Rider Optimization

algorithm is proposed in this work for CH

selection. Subsequently, a Ridge Regression

Classification algorithm is presented to detect

the malicious node in this network.

Accordingly, Atom Emperor Penguin

Optimization (AEPO) algorithm is presented

to route the network and find the best

forwarding path, respectively.

4.1 Cluster Head Selection Using FCRO

Algorithm

Clustering is an important concept in MANET

where several nodes join to form a group

based on common features. The node is a

single system which is responsible to store and

process data. Whereas Cluster is formed based

on the collection of multiple nodes which

communicates with each other to perform a set

of operation. Furthermore, selecting an

efficient CH node is a difficult task due to the

limited battery power supply in MANET. In

terms of the power allocation of nodes to

clusters, cluster formation is a valuable

operation. Every cluster has one or more CH,

which are connected to build networks and

disseminate data. The CH node is a trusted

node that scans a node’s performance and

other security-related tasks are performed.

Malicious nodes can quickly drain energy and

degrade the overall longevity of networks;

hence a CH should be able to function in low-

energy and resource-constrained contexts.

Subsequently, this research proposes a Hybrid

Firefly Cyclic Rider Optimization (FCRO)

algorithm for optimal CH selection (CHS) to

improve the MANET network’s energy

efficiency and lifetime. Additionally, the

network’s dead nodes, overall throughput,

residual energy of nodes, convergence rate,

alive nodes, and network survivability index

are all improved.

Cluster Formation: Due to the energy

efficiency reflection, the logical identification

accounts for the cluster are almost often. The

use of a sensible cluster count can improve

network connection efficiency while also

balancing node energy loss and extending

network lifespan. Subsequently, in inter-

cluster communication, this framework uses

the multi-hop routing mechanism. The

distance between the BS and the extreme CH

is expressed as

Di

, which is classified as

many hops.

lekDi .

, here the equidistance

length is

le

and the number of clusters is

k

.

Under the multi-hop communication method,

the energy consumption is given as (1).

TxDARxmh EEEE 

(1)

Where,

mh

E

is represented as energy

consumption in multi-hop communication,

Rx

E

, and

Tx

E

are denoted as energy

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

80

Volume 22, 2023

consumption in receiver and transmitter data,

respectively.

4.1.1 CH Selection

Accordingly, in the establishment of clusters,

the best CH for data transmission must be

chosen. Consequently, this research aims to

choose the best CH while taking into account

key fitness criteria such as energy, distance,

and latency, which have already been

identified as issues when selecting the CH.

Additionally, the distance should be

minimised by using the shortest path selection,

and the presentation of data transfer is

improved. Another highly serious issue is the

energy usage of each node. Accordingly, the

information distribution is the largest problem

with the shortest length and the least amount

of energy. Furthermore, practically all

optimisation algorithms hold a high level of

responsibility for distance and energy

expenditure while choosing on CH selection.

Subsequently, to improve the node’s life

expectancy, multiple objectives are required.

Finally, the primary criteria to consider while

selecting the CH from a group of sensor nodes

are latency, energy, and distance.

Distance: Equation (2) depicts the distance

fitness function.

 

dis

b

dis

a

dis

if

f

f

(2)

 

 

 











 xy

N

x

N

yXXsx

dis

aDFBFf

1 1

(3)

 



 

 xy

N

x

N

yyx

dis

bDDf

1 1

(4)

Where

dis

i

f

illustrates the fitness function

for the distance,

 

dis

a

f

and

 

dis

b

f

are the distance

of two nodes. Subsequently,

x

F

is the distance

of CH,

s

B

the distance of BS,

x

D

and

y

D

the

distance of normal data and the count of

nodes.

In (2),

 

dis

a

f

the value must add a distance to

it so that the packets connected with it can be

transferred quickly from the common node to

the CH and the destination. The obtained

specific value must be low and lies between

[0, 1]. The value

dis

i

f

grows in proportion to

the distance between the common node and

CH. In (3),

sx BF 

which illustrates the

distance between the CH and BS,

xx DF 

defines the distance between the CH and

normal node, and

yx DD 

refers to the

distance between the two normal nodes, and

the count of nodes not provided in

x

th and

y

th CH is referred to as

x

N

and

y

N

the node

that is available within the CH is referred to as

x

R

for the

x

th CH.

Energy: The fitness function for energy is

shown in equation (5).

 

ene

b

ene

a

ene

if

f

f

(5)

Where

ene

i

f

represented as the fitness

function for the energy,

 

ene

a

f

and

 

ene

b

f

are

portrayed as the fitness function for the

cumulative clusters. Where the cumulative CH

 

ene

a

f

and

 

ene

b

f

is taken into account as a rise in

energy value, the

ene

i

f

value is presumed to be

greater than one and increases the number of

CHs.

Delay: The fitness function of the delay is

directly proportional to the member count

within the cluster. Accordingly, the CH owns

a certain number of members for delay

reduction. The fitness function for the delay is

represented by equation (6).

 

c

N

x

xx

del

iN

DF

f

c

1

max 





(6)

Where

del

i

f

portrays the fitness function of

delay, the maximum amount of CH is

contained in the numerator and the

denominator

c

N

contains every node in the

network.

del

i

f

value is relay within the interval

[0, 1] and it has to be lower for the filter CH

selection.

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

81

Volume 22, 2023

4.1.2 Firefly Algorithm (FA)

The firefly algorithm is a biologically inspired

algorithm that inspires the firefly's flashing

behaviour. Subsequently, this algorithm

employs the following three rules.

1. The fireflies are attracted to the

opposite mate.

2. The firefly's attraction is calculated

using the brightness value. The attraction of

the firefly reduces as the distance between the

two fireflies grows. The brighter firefly

attracts the less-brighter firefly, while the less-

brighter firefly attracts the brighter firefly.

3. The firefly’s brightness is calculated

using the objective function below.

The inverse square law governs the

variation in light intensity

 

rI

.

 

2

r

I

rI s



(7)

Where

s

I

is the source’s intensity and

r

is

the distance to the source. Lumens are the

metric for light intensity. The firefly’s

attractiveness is depicted as

2

0

r

e









(8)

Where the light absorption coefficient is

denoted as



and

0



is the fluctuation of

attractiveness at

0r

. Euclidean distance is

used to compute the distance between any two

fireflies at the coordinates

i

x

and

j

x

.

Consequently, it can be characterized as

follows:

 





 d

kkjkijiij xxxxr

1

2

,,

(9)

Where

ki

x,

is the

k

th component of the

i

th

firefly’s spatial coordinate

i

x

. The brighter

firefly

j

attracts the less bright firefly

i

. The

firefly’s movement is illustrated by

 

t

it

t

i

tj

r

t

i

t

ixxexx ij  



 

2

0

1

(10)

The current position is

t

i

x

. The second

word refers to the firefly’s mutual attraction.

Utilizing a Gaussian distribution at the time

t

,

the third term

t

i



denotes the production of

random vectors. When

0

0



; the second

term in equation 10 is omitted, the firefly takes

a random walk.

4.1.3 Hybrid Firefly Cyclic Rider

Optimization (FCRO) Approach

The brighter firefly attracts the less bright one

in every iteration of the firefly algorithm.

Consequently, in the position’s random

movement, this algorithm works well for

identifying the solution, but it ignores firefly’s

best position for CH selection, for finding the

best solution, it has an impact on the global

search behaviour. The ROA algorithm paired

with the firefly algorithm overcomes the

above-mentioned restrictions. Subsequently,

by finding better solutions in CH selection, the

ROA algorithm improves network

performance. The

best

p

and

best

g

values

derived by ROA are used by each firefly.

Furthermore, to attain greater results during

the riding process, this strategy strikes a

balance between exploitation and exploration.

The convergence speed is increased and the

precision of the solution is determined by the

movement of each firefly.

Optimal Selection of Cluster Count and

CH: This research aims to solve two major

optimization challenges optimal cluster counts

and Optimal CH selection. The best cluster

counts

k

objective function

1

OB

can be

established in (11).

 

tot

EgOB min

1

(11)

The optimal CH selection’s objective

function

2

OB

is defined in (12)

 

32 min FOB 

(12)

Where,

   

del

i

fFF *1* 23





,

   

QoS

FF 1

*1* 12





, and

 

ene

i

dis

if

fF 1

*1*

1





. Subsequently,



,



, and



are the constant values that are

fixed at 0.3, 0.3, and 0.9, respectively. Where

z

CH

;

CH

Nz ,,2,1 

is the number of CHs

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

82

Volume 22, 2023

and

i

k

;

k

Ni ,,1 

is the number of clusters

that are optimally selected.

4.1.4 Cyclic ROA Approach

ROA is a fictitious computing algorithm based

on the inspiration of a group of riders that

strive to reach a specific destination to win the

race. Accordingly, they are divided into four

groups, with the number of riders split evenly

among the four groups. Bypass rider, follower,

overtaker, and attacker are the four groups.

The following methods are used by each group

to reach the target:

 The leading path is avoided to get to

the destination is the main goal of bypass

rider.

 In the majority axis, the follower tends

to follow the leading rider.

 According to the leading rider’s

location to achieve the target, the overtaker is

the one that follows their path.

 To reach the destination, point as

quickly as possible, the attackers capture the

rider’s path.

The riders follow a pre-planned strategy,

with the appropriate use of the accelerator,

gear, steering, and brake being the most

important aspects to consider when achieving

the goal. Riders alter their position for each

instant of time while approaching the goal by

regulating these factors, and based on the

current success rate, it continues with the pre-

planned strategy, to the distance between the

rider's destination and the current location,

which is inversely related. The leading rider is

mentioned based on the current success rate.

Consequently, this process will continue until

the riders have been given the maximum

amount of time. The winner is then announced

with the leading driver. The Cyclic ROA

model algorithm is presented in table 1.

Table 1: Algorithm for Cyclic ROA Model

Input: Rider’s random positions,

t

R

Output: Leading rider,

I

R

Allot the population

Allot the rider components: Steering angle

A

,

Gear

G

,

Accelerator

a

and Break

Br

Determine the success rate

While

off

TT 

for

1x

to

R

Update bypass position rider as per equation

(13)

Update follower position rider as per equation

(14)

Update overtaker position rider as per equation

(15)

Update attacker position rider as per equation

(16)

Grade riders based on the success rate

Choose a rider with a higher success rate as

the leading one.

Update the rider constraints

Return

L

Z

1tt

end for

end while

End

Rider’s Position Update: By updating the

rider’s position in each set, the leading rider is

identified.

Update Procedure for Bypass Rider: The

bypass rider’s position update is offered on a

random basis because they bypass the normal

path without following the leading riders.

Consequently, this is depicted in (13). Where



and



are values assigned with a random

integer with ranges between 1, and RN



and



are random values between 0 and 1.

         

  

yyRyyRyxR tt

B

t





1*,,,

1

(13)

Update Procedure for Follower: Due to

the follower location being updated by

tracking the main rider’s position, these riders

achieve the destination successfully and

quickly. Based on the coordinate selector, the

location update of the follower is also

computed for the specified values

P

and is

reported in (14)

   

 

 

t

x

It cx

IF

tdiscIRAcIRcxR *,*cos,, ,1 



(14)

The distance that has to be travelled by

x

th

rider is shown

t

x

dis

and the leading rider’s

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

83

Volume 22, 2023

position is presented as

I

R

x

th rider’s steering

angle at

c

the coordinate is shown as

tcx

A,

.

Update Procedure for Overtaker: The

coordinate selector, relative success rate, and

direction indicator are all important variables

in the overtaker update process. The

overtaker’s position update equation is

represented in equation (15). The direction

indicator for the

x

th rider’s is given

 

xMt

at

the time

t

.

       

 

cIRxMcxRcxR I

tt

O

t,*,,

1



(15)

Update Procedure for the Attacker: As it

seeks to steal the leading rider position, the

attacker uses the same method as a follower to

update its position. The attacker’s position is

updated in the following manner (16)

   

 

 

t

x

It yx

IA

tdiscIRAcIRyxR 

,*cos,, ,1

(16)

Even if the ROA is relatively quick in

identifying the best solutions, it makes sense if

the algorithm is improved for a better

problem-solving scenario. Consequently, the

goal of this work is to offer a new enhanced

idea in ROA called FCRO, which is based on

the firefly method. The following is the

proposed algorithm: Each iteration

t

is

evaluated to see if the current

t

one achieves

the optimal answer compared to the prior

one

1t

. The procedure is carried out as usual,

if the evaluation is correct (finding the best

option over the prior

1t

). The parameter trail

is set to 0 and the iterations continue if it is not

improved. Consequently, the Hybrid - FCRO

algorithm chooses the best CH and improves

the network efficiency.

4.2 Malicious Node Detection

Identifying malicious, selfish, and

compromised nodes that have been

authenticated requires trust management.

Accordingly, it has been widely explored in a

variety of network contexts, including grid and

pervasive computing, peer-to-peer networks,

etc. Alternatively, for assessing a node’s trust,

trust management systems are tools as well as

detecting unexpected node behaviour (either

faulty or malicious) and hence selecting a

node for routing. To increase the QoS metric

performance, the research suggested a Ridge

Regression Classification technique that finds

the MANET’s trustworthy nodes and

transmission of data is carried on with the

trusted MN. While allowing trustworthy nodes

to route, effectively removing malicious and

selfish nodes, the suggested technique

effectively detects harmful and selfish nodes.

Clustering algorithms are employed in

various detection tactics to detect malicious

nodes, and they cluster the nodes into two

groups, such as malicious and benign groups.

Furthermore, the behaviour of other nodes

along a node’s linked multi-hop pathways

might affect its trust value, lowering the

performance of detection algorithms. Cluster

the nodes into three groups: low trust value

group (LTG), medium trust value group

(MTG), and high trust value group (HTG) to

improve detection accuracy. In order to

capture additional information about them,

inject the packets back into the network and

improve the routing of broadcast packets to

evaluate whether MTG nodes are benign or

malicious, which can help the regression learn

better. Subsequently, classify the nodes into

malicious or benign using a clustering method

based on the received trust levels.

4.2.1 Ridge Regression Algorithm

In a MANET, each normal node has a

classifier that can detect malicious nodes.

Additionally, this strategy, which employs the

observed behaviour of the nearby nodes by the

normal node and the classifiers, is used by

each normal node to regulate whether or not

the neighbouring nodes are malicious. To

make an accurate detection, there is adequate

information about the observed behaviour and

the detection operation can be carried out at

any moment. When there is insufficient data,

the Machine learning algorithms fail, therefore

frequent execution of malicious node detection

is unlikely. In these experiments, normal

nodes determine whether or not their

surrounding nodes are malicious after

observing their behaviour for a length of time

(e.g., 50 seconds or more). Consequently, the

technique for detecting malicious nodes does

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

84

Volume 22, 2023

not suffer the significant computation costs of

the normal nodes.

The ridge regression approach is used to

detect the malicious node in this study. When

the data is supplied as

 

ii yx ,

where

 

T

piiii xxxx ,,, 21 

is the input and

i

y

the

output in the ridge regression process, a linear

regression model can be stated as:

pipiii xxxy



 

2211

(17)

Where the regression coefficients for

various

i

x

are represented as

p



,,, 21 

.

The goal of linear regression is to select those

i



with the smallest residual sum of squares.

The optimization problem is therefore

formulated as follows:

























 



N

i j ijji xy

1

2

min



(18)

However, this model does not generalize

well to new data (i.e., data with a lot of

variation), resulting in overfitting. Ridge

regression is utilised to solve this problem

since it progressively narrows the regression

coefficients, resulting in a stable model.

Subsequently, by adding a constraint to the

optimization problem of (18), ridge regression

solves this problem. Consequently, the ridge

regression optimization problem can be

phrased as follows:

  

























jj

N

i j ijji tsxy 1..,min 2

1

2





(19)

Introduce the Lagrange multiplier



,

commonly known as the regularization

constant, to solve this optimization problem.

The Ridge estimate

 

T

p



ˆ

,,

ˆ

,

ˆˆ 21 

is

then supplied as follows:

2

1

2

minarg

ˆ  

























jj

N

i j jiji xy

 

(20)

Subjected to the condition that for each

ji

x

,

1

2

iji Nx

. Where



the constant that

determines the amount of regularization is

applied and

jj

2



is the regularization

term. Set

1



the maximum iterations to 10

in this experiment.

4.3 Trust-Based Routing Protocol

The routing of trustworthy nodes is permitted,

but malicious/selfish nodes are promptly

eliminated. To reduce routing overhead in

MANET, the multipath routing technique is

implemented. Consequently, by reducing

network traffic, it can be achieved.

Accordingly, the broadcasting message’s

effectiveness is improved is the goal of this

study and it is also used to reduce the routing

overhead. Routing is the method of

transferring data from one location to another.

In order to reach their destination, it also

allows messages to travel from one driving

node to another. Subsequently, to identify the

best-forwarded path in MANET to reduce

routing overhead, the Atom Emperor Penguin

Optimization (AEPO) method is used to

introduce a trust-based secure routing

protocol. In order to choose the best path for

trust evaluation criteria such as average

encounter rate (AER), forwarding rate,

Successful Cooperation Frequency (SCF), and

integrity factor, the developed algorithm is

used. Consequently, the sender nodes can

reduce delay and enhance their data

transmission rates as enabled by this

evaluation.

4.3.1 Atom Emperor Penguin Optimization

(AEPO) Algorithm

Atom search optimization (ASO) is a recently

proposed optimization algorithm based on

molecular dynamics. In order to discover the

optimal forwarding path, the ASO is paired

with the emperor penguin colony optimization

method. The location of atoms in ASO is

updated by:

     

11  tvtxtx iii

(21)

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

85

Volume 22, 2023

Where

 

1txi

is the

i

th atom’s position in

the

 

1t

th iteration,

 

txi

is the

i

th atom’s

location in the

t

th iteration, and

 

1tvi

is the

i

th atom's velocity in the

 

1t

th iteration,

and is computed as follows:

     

tatvrandtv iii 1

(22)

Where

 

tai

the acceleration of the

i

th

atom in the

t

th iteration and the rand is a

random number in [0, 1] is determined as

follows:

     

   

 

 

   

       

 

tm

txtx

e

txtx

tm

hthrand

tta

i

ibest

T

ji

bestKj i

jiji

i











 



20

2

713

,

2



(23)

Where



is the multiplier weight,

T

is the

maximum number of iterations,

best

x

is the best

atom in the current iteration, and

Kbest

is a

subset of the best atoms,

 

t



,

 

thij

, and

 

tmi

are determined using equations 4, 5,

and 6.

Where,

 

T

t

e

T

t

20

3

1

1



















,

 



















maxmax

maxmin

minmin

h

t

tr

h

t

tr

h

t

tr

h

t

tr

h

th

ij

ijij

ij







, and

   

 



N

jj

i

itM

tM

tm

1

. Accordingly,



the

depth weight

 

trij

is the distance between

i

th

and

j

th atoms at the

t

th iteration,

min

h

,

max

h

,

 

t



, and

 

tMi

are calculated

as:

 

tggh  0min

,

uh 

max

,

     

 

tK

tx

txt Kbestj ij

ij









,

, and

 

   

tFittFit

ibestworsrt

besti

etM 





.

Where

0

g

is equal to 1.1,

u

is equal to

1.24,

 

tg

and

 

tK

are estimated as:

 











 T

t

tg 2

sin1.0



(24)

   

T

t

NNtK  2

(25)

Where

N

is the number of atoms.

Subsequently, each penguin’s cost and

location are calculated. Penguins are priced

against one another. Penguins will always

choose a penguin with a low absorption cost

(high heat intensity). The cost is influenced by

the temperature and the length of the journey

involved. During the attraction process, the

heat strength will be adjusted as necessary.

The best answer is chosen after all others have

been sorted. Heat radiation, association, and

heat absorption are all subjected to a damping

ratio. In algorithm 1, pseudo-code for the EPC

algorithm is described. The following are the

rules that apply to this algorithm:

 Every penguin in the original

population radiates heat and is drawn to others

with a similar thermal absorptivity.

 All penguins are thought to have the

same body surface area.

 The influence of the earth's surface and

atmosphere are not taken into account when

the penguin absorbs all of the thermal

radiation.

 Penguin heat radiation is a straight

line.

 The attraction between two penguins is

determined by the quantity of heat that

separates them. Longer distances receive less

heat, while shorter distances receive more

heat.

 Accordingly, there is a variation with a

consistent distribution in the spiral movement

of the penguins during the absorption process.

Heat Radiation: The heat radiation

transfer must be computed to determine the

intensity and attractiveness of the heat. Each

penguin’s body surface area must be

calculated to determine how much heat it

radiates.

4

spenquin TAQ





(26)

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

86

Volume 22, 2023

Where heat transfer is defined as

penquin

Q

that will be calculated for unit time

 

W

, the total surface area

2

56.0 m

is denoted

as

A

.

4

s

T

is the absolute temperature in Kelvin

(K) and 35 degrees Celsius

 

C



is equal to

308.15 K, the emissivity of bird plumage is



.

The Stefan–Boltzmann constant is also known

as

 

42

1086703.5 kmW



.

Attractiveness: Finally, the attractiveness

Q is defined as,

s

seTAQ







4

(27)

Spiral Movement: In this scenario, the

system’s structure features have uncertain

borders and a spiral pattern around the centre.

The centre of the huddle is the warmest

temperature, whereas the outside is

significantly colder. Penguins do not compete

for personal advantage. Consequently, the

entire huddle moves in a leisurely spiral

motion, with each penguin taking turns at all

spots in the formation.

 































 





















xi

yj

ib

xi

yj

ib

eQ

b

keQeQ

b

aex

xi

yj

lb

tantan

1ln

1

1ln

1

cos

tan

(28)

 































 





















xi

yj

ib

xi

yj

ib

eQ

b

keQeQ

b

aey

xi

yJ

lb

tantan

1ln

1

1ln

1

sin

tan

(29)

New Position: Equation (15) is used to

calculate the new position, and it is multiplied

by the mutation factor and by a random vector,

respectively, to arrive at the new position. The

random vector's coefficient is tacked on in this

fashion.

The data window is moved one step

forward when the signals are reconstructed

and all parameters with variables are obtained

after a certain number of iterations, then the

recovery algorithm is reworked again.

Subsequently, for each initialization, global

convergence can be provided by EPC and

solves the problem of convex optimization.

Consequently, this Atom Emperor Penguin

Optimization (AEPO) algorithm finds the best

forwarding paths to route the nodes in the

MANET.

5. Experimentation and Result

Discussion

The proposed method’s performance is

evaluated using the Matlab software with the

version R2021a, their operating system is

Windows 10 Home. The memory capacity of

the proposed method is 6 GB DDR3, with the

processor of Intel Core i5 @ 3.5 GHz.

MATLAB simulation software is utilized for

verifying the validity of the proposed AEPO

algorithm. Subsequently, the detailed settings

of simulation parameters are listed in Table 2.

The simulation parameters are described, that

the number of nodes utilized in this work is

100, the initial energy of each node is 0.5 J,

transmitter and receiver required to run

circuity is 50*10-9J/bit. Accordingly, the

number of decision variables is 200, the

number of initial dead nodes is 0, the size of

the message is 128 bytes, and the number of

iterations is 100, respectively.

Table 2: Simulation Parameters

Parameters

Values

Number of Nodes

100

Initial Energy of Each

Node

0.5 (Joule)

Transmitter Energy

Required to Run Circuity

50*10-9(Joules

/bit)

Receiver Energy Required

to Run Circuity

50*10-9J/bit

Number of Decision

Variables

200

Data Aggregation Energy

50*10-9(Joules

/bit)

Number of Initial

Operating Nodes

100

Number of Initial Dead

nodes

0

Packet Size

128 bytes

Number of Iterations

100

Subsequently, the proposed technique’s

performance is evaluated based on various

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

87

Volume 22, 2023

parameters including PDR, packet loss ratio

(PLR), end-to-end delay, throughput,

normalized energy analysis, and normalized

routing overhead. Accordingly, the proposed

AEPO algorithm is compared with the existing

Hybrid Particle Swarm Optimization-Genetic

Algorithm (PSO-GA)Hamza, F, et al

(2021)[27], Ticket ID Cluster Manager (TID-

CMGR) Venkatasubramanian, et al

(2021)[28], and Modified Firefly Algorithm

(MFFA)Kumar, (2021)[29].

Packet Delivery Ratio: PDR is calculated

as the ratio of the number of data packets

received to the total number of data packets

sent via trusted nodes. PDR is formalized as

below,

100















xDp

Tn

Rn

PDR

i

(30)

The above equation (30),

i

Dp

n

indicates

several data packets,

x

R

symbolizes received

x

T

and is then transmitted. PDR is calculated in

percentage (%).

Figure 2: Performance Graph for Packet

Delivery Ratio

Figure 2 represents the PDR graph and the

values of the proposed AEPO algorithm

compared with the existing PSO-GA, TID-

CMGR, and MFFA methods. The PDR

represents the number of packets successfully

received at the destination end from the

packets to be sent. The PDR of the proposed

AEPO algorithm is approximately 2% higher

than the other existing methods.

Packet Loss Rate: The ratio of the number

of data packets lost to the total number of data

packets send is used to calculate PLR. The

formalized PLR is as follows:

100















xDp

Dp

Tn

lostn

PLR

i

(31)

From (31),

i

Dp

n

indicates several data

packets,

x

T

are then transmitted. PLR is

calculated in percentage (%).

Figure 3: Performance Graph for Packet Loss

Ratio

Figure 3 illustrates the graph of PLR

compared with different existing methods like

PSO-GA, TID-CMGR, and MFFA methods.

The PLR is measured based on the node

density subsequently, the performance of the

proposed AEPO algorithm is 5% higher than

the existing PSO-GA, 3% higher than the

TID-CMGR, and 1.5% higher than the MFFA

methods.

End to End Delay: E2E delay is calculated

as the time difference between the data packet

arriving at the destination and the data packet

sent from the source node. E2E delay is

formalized as below,

 

sdal TTdelayEE 2

(32)

From (32), the data packet arrival time is

indicated as

al

T

, and the data packet sending

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

88

Volume 22, 2023

time is indicated as

sd

T

. In milliseconds (ms),

the

delayEE2

is determined.

Figure 4: Result for End-to-End Delay

The end-to-end delay graph is portrayed in

figure 4. The end-to-end delay for the

proposed AEPO algorithm is compared with

the existing PSO-GA, TID-CMGR, and

MFFA. Subsequently, it depicted that the

AEPO algorithm performs the best

performance than the other existing methods.

Accordingly, it portrays that the AEPO

method is approximately 2% higher than the

other existing methods, respectively.

Throughput: The throughput measure

defines the total packets sent by the transmitter

node to the total packets received at the

receiver end.

t

SizePacketceivedPackets

Throughput 

Re

(33)

Figure 5: Performance Graph for Throughput

Figure 5 depicts the throughput of the

proposed AEPO algorithm. Consequently, the

suggested method is compared to the existing

PSO-GA, TID-CMGR, and MFFA. The

throughput is evaluated based on the rounds

from 0 to 2000, and the initial throughput

value is 1. When the round is 1000 to 1500,

the throughput is gradually decreased to 0.8

and when it reaches 1500 to 2000, the

throughput is suddenly reduced to 0.17.

Consequently, the AEPO algorithm performs

better than the other methods, and it is roughly

3% higher than the existing methods.

Routing Overhead (RO): Network

overhead is the number of control (hello

packets) and routing packets required for

overall network communication.

 

ceivedPacketsDataofNumber

PacketRoutingandContolTotal

ratioinOverhead Re



(34)

Figure 6: Graph for Routing Overhead

Figure 6 elucidates the performance and the

comparison graph for routing overhead. The

AEPO algorithm is compared with the various

algorithm including PSO-GA, TID-CMGR,

and MFFA. The experimental results proved

that the proposed algorithm has taken very less

RO when compared with other existing

algorithms. Subsequently, their performance

of the AEPO algorithm is approximately 4%

higher, respectively.

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

89

Volume 22, 2023

Figure 7: Graph for Number of Alive Nodes

Figure 7 delineates the analysis of alive

nodes of the proposed AEPO model and the

conventional models. The existing methods

like PSO-GA, and TID-CMGR. The alive

nodes have to be at maximum for the AEPO

algorithm. From the starting itself, the alive

nodes are at maximum and in the subsequent

rounds, it gets decreases as the rounds

increase. The alive nodes are initially at

maximum with a count of 500. Subsequently,

it gradually decreases and reaches the count

within the range of 150–200.

Figure 8: Normalized Network Energy

Analysis

The analysis of the normalized energy of

the proposed over the conventional models is

described in figure 8. The existing methods

like PSO-GA, and TID-CMGR.Initially, the

normalized network energy is fixed within the

interval of 0.05 to 0.06. Consequently, during

the 2000th round, the energy would steadily

decline to the bottom, until the normalized

energy of the proposed AEPO algorithm

reaches a value in the range of 0 to 0.01 and is

maximal when compared to other traditional

models.

Figure 9: Cost Function of the AEPO

Algorithm

The analysis of the cost function of the

proposed and classical models is depicted in

figure 9. The graph explains the AEPO

algorithm with a lower cost function than any

of the other models. Continuously, at the

eighth iteration, the performance of the AEPO

algorithm achieves the lowest cost function

which is 3%, 3.5%, and 3.1% better than PSO-

GA, TID-CMGR, and MFFA, respectively.

The analysis has made clear that the adopted

model achieves a lower cost function than the

other related conventional models.

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

90

Volume 22, 2023

Figure 10: Performance Graph for Average

Life Time of CH

The average lifetime of CH is portrayed in

figure 10; it compares the AEPO method with

the existing methods. The AEPO algorithm

has a higher value than the other existing

methods. The existing methods like PSO-GA,

TID-CMGR, and MFFA. The CH has a

maximum lifetime than the other existing

methods and the AEPO algorithm is

approximately 5.1%, 4.2%, and 3.5% higher

than the existing methods.

Figure 11: Performance Graph for Average

Transmission Delay

The average transmission delay graph is

demonstrated in above figure 11. The

transmission delay for the AEPO algorithm is

less than the other, subsequently, the AEPO

algorithm is higher than the existing PSO-GA,

TID-CMGR, and MFFA. The average

transmission delay is measured with the

simulation time as 200 minutes to 1600 min,

respectively.

Figure 12: Performance Graph for Energy

Consumption

The energy consumption of the AEPO

algorithm is illustrated in figure 12; it

demonstrates that the AEPO algorithm has

less energy consumption than the other

existing methods. The energy consumption is

evaluated based on the node density, in this

work, the node density is taken between 50 to

500. Subsequently, the AEPO algorithm is

approximately 4.8%, 2.3%, and 0.8% higher

than the PSO-GA, TID-CMGR, and MFFA.

Figure 13: Performance Graph for the

Network Lifetime

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

91

Volume 22, 2023

The network lifetime graph for the AEPO

algorithm is portrayed in figure 13.

Additionally, it depicted that the network

lifetime of the AEPO algorithm produces

higher values than the other methods. The

network lifetime is evaluated for the number

of nodes from 20 to 100. The AEPO algorithm

is 2%, 3.5%, and 4% higher than the existing

PSO-GA, TID-CMGR, and MFFA.

Table 3: Comparison Table of the Research Work

PSO-

GA

Hamza

et al.,

2021

TID-CMGR

Venkatasubramanian

et al., 2021

MFFA

Kumar

et al.,

2021

AEPO

Proposed

PDR (%)

No.of

Packets

30

83.1

80.8

78

84.8

300

88

85.3

83

91.1

PLR (%)

No.of

Nodes

100

12

7.3

2.2

0.53

500

18.8

14.5

10

3.94

End to End

Delay (ms)

No.of

Packets

30

13.5

15

17.5

12

300

30

35

30

Throughput

Round

500

1

2000

0.34

0.17

0.497

0.5

Average

Transmission

Delay (s)

Simulator

Time

(min)

300

3500

2990

1900

1500

6980

6000

6150

5800

Energy

Consumption

(mJ)

Node

Density

100

1.08

0.77

0.7

0.62

500

1.33

1.05

1.03

1.02

Network

Lifetime

No.of

Nodes

20

340

200

180

430

180

1200

750

690

1310

Normalized

Network

Energy

No.of

Rounds

500

0.017

0.018

-

0.0165

2000

0.002

0.004

-

0.001

Transmission

Success Rate

Simulator

Time

(min)

200

0.78

0.69

0.502

0.87

1200

0.5

0.49

0.54

0.65

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

92

Volume 22, 2023

Table 3 reveals the comparison table of the

work, it portrays the performance values for

PDR, PLR, end-to-end delay, throughput,

average transmission delay, energy

consumption, network lifetime, normalized

network energy, and transmission success rate.

The table portrays both the existing and

proposed methods, the existing methods like

PSO-GAHamza, F, et al (2021), TID-CMGR

Venkatasubramanian, et al (2021), and

MFFAKumar, (2021).

Figure 14: Graph for Transmission Success

Rate

The transmission success rate graph of the

AEPO algorithm is portrayed in figure 14. The

figure demonstrates that the transmission

success rate of the proposed method is higher

than the other existing methods. The

transmission success rate of the AEPO

algorithm ranges from 0.8 to 0.65 for the

simulation time of 0 to 1200 min. The AEPO

algorithm is approximately 4%, 4.1%, and

5.4% higher than PSO-GA, TID-CMGR, and

MFFA, respectively.

Figure 15: Misclassification Node vs

Terminal Node

Figure 15 portrays the graph for the number

of misclassification nodes vs the number of

terminal nodes. The proposed method is

compared with the existing PSO-GA, TID-

CMGR, and MFFA methods. While compare

to these existing methods, the proposed

method produces low misclassification nodes.

6. Conclusion

MANETs are noticeable by an individual

feature which contains the nonappearance of

any essential organization else any necessitate

for infrastructure unit. Many hop topologies

are outlined in the network. Subsequently,

with the help of a comprehensive source

known as energy, extreme mobile nodes form

and an ad hoc network is powered.

Accordingly, there has been no major

improvement in the domain of energy

characteristics to reduce energy usage, the

lifespan of a network is heavily reliant on the

technology incorporated within the rules’

sequence. In this research, a Hybrid Firefly

Cyclic Rider Optimization (FCRO) algorithm

is presented to select the CH. Accordingly, it

improves the MANET network efficiency.

Subsequently, malicious node detection is an

important process, in this malicious node is

detected using the Ridge Regression

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

93

Volume 22, 2023

Classification algorithm, which classifies

malicious nodes and trusted nodes.

Accordingly, Hybrid Firefly Cyclic Rider

Optimization (FCRO) algorithm is presented

to route the nodes to the destination node.

 Subsequently, the proposed method is

implemented using MATLAB software.

 The performance metrics are packet

deliver ratio, packet loss ratio, Routing

overhead, throughput, end-to-end delay,

transmission delay, network lifetime, and

energy consumption.

 The AEPO algorithm is compared with

the existing PSO-GA, TID-CMGR, and

MFFA.

 The performance of the proposed

AEPO algorithm is approximately 1.5%,

3.2%, 2%, 3%, and 4% higher than the

existing methods for PDR, PLR, end-to-end

delay, throughput, and network lifetime.

Accordingly, this analysis allows sender

nodes to reduce the delay and improve their

data transmission speeds. In terms of

demonstrating the true contribution of

different nodes to trust evaluation, the

proposed method has an obvious advantage.

Consequently, the routing technique will be

used in future investigations to achieve a

secure and effective transmission. Future work

should focus on improving data transmission

security by utilising authentication approaches

to manage the increased demand and higher

security. Table 4 shows the abbreviations used

in the article.

Table 4: Abbreviations

PSO-GA

Hybrid Particle Swarm

Optimization-Genetic Algorithm

TID-

CMGR

Ticket ID Cluster Manager

MFFA

Modified Firefly Algorithm

FCRO

Hybrid Firefly Cyclic Rider

Optimization

AEPO

Atom Emperor Penguin

Optimization

PDR

Packet Delivery Ratio

PLR

Packet Loss Rate

CH

Cluster Heads

QoS

Quality of Service

MN

Mobile Node

RO

Routing Overhead

References

[1] Xu, J., Feng, S., Liang, W., Ke, J., Meng,

X., Zhang, R. and Shou, D., 2020. An

algorithm for determining data forwarding

strategy based on recommended trust value

in MANET. International Journal of

Embedded Systems, 12(4), pp.544-553.

[2] Alkhamisi, A.O., Buhari, S.M.,

Tsaramirsis, G. and Basheri, M., 2020. An

integrated incentive and trust-based

optimal path identification in ad hoc on-

demand multipath distance vector routing

for MANET. International Journal of Grid

and Utility Computing, 11(2), pp.169-184.

[3] Yang, H., 2020. A study on improving

secure routing performance using trust

model in MANET. Mobile Information

Systems, 2020.

[4] Nandgave-Usturge, S., 2020. Water spider

monkey optimization algorithm for trust-

based MANET secure routing in IoT. Int J

Scientific Res Eng Trends, 6(2), pp.980-

984.

[5] Abhilash, K.J. and Shivaprakasha, K.S.,

2020. Secure routing protocol for

MANET: A survey. In Advances in

communication, signal processing, VLSI,

and embedded systems (pp. 263-277).

Springer, Singapore.

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

94

Volume 22, 2023

[6] Malik, N.A. and Rai, M., 2020, April.

Enhanced secure and efficient key

management algorithm and fuzzy with

trust management for MANETs.

In Proceedings of the International

Conference on Innovative Computing &

Communications (ICICC).

[7] Panwar, A., Panwar, B., Rao, D.S. and

Sriram, G., 2020. A trust based approach

for avoidance of wormhole attack in

Manet. International Journal of Computer

Science and Mobile Computing, 9, pp.47-

57.

[8] Thapar, S. and Sharma, S.K., 2020,

November. Direct Trust-based Detection

Algorithm for Preventing Jellyfish Attack

in MANET. In 2020 4th International

Conference on Electronics,

Communication and Aerospace

Technology (ICECA) (pp. 749-753).

IEEE.

[9] Gomathy, V., Padhy, N., Samanta, D.,

Sivaram, M., Jain, V. and Amiri, I.S.,

2020. Malicious node detection using

heterogeneous cluster based secure routing

protocol (HCBS) in wireless adhoc sensor

networks. Journal of Ambient Intelligence

and Humanized Computing, 11(11),

pp.4995-5001.

[10] Gupta, M., Garg, P., Gupta, S. and Joon,

R., 2020. A Novel Approach for Malicious

Node Detection in Cluster-Head Gateway

Switching Routing in Mobile Ad Hoc

Networks. International Journal of Future

Generation Communication and

Networking, 13(4), pp.99-111.

[11] Hassan, K.L., Mandal, J.K. and Mondal,

S., 2020. A Dynamic Threshold-Based

Trust-Oriented Intrusion Detection System

in MANET. In Proceedings of the Global

AI Congress 2019 (pp. 699-711). Springer,

Singapore.

[12] Fatemidokht, H. and Rafsanjani, M.K.,

2020. QMM-VANET: An efficient

clustering algorithm based on QoS and

monitoring of malicious vehicles in

vehicular ad hoc networks. Journal of

Systems and Software, 165, p.110561.

[13]Gu, K., Dong, X. and Jia, W., 2020.

Malicious node detection scheme based on

correlation of data and network topology

in fog computing-based VANETs. IEEE

Transactions on Cloud Computing.

[14] Ghaleb, S.A.M. and Vasanthi, V., 2020.

Energy Efficient Multipath Routing Using

Multi-Objective Grey Wolf Optimizer

based Dynamic Source Routing Algorithm

for MANET. International Journal of

Advanced Science and Technology, 29(3),

pp.6096-6117.

[15] Chen, Z., Zhou, W., Wu, S. and Cheng,

L., 2020. An adaptive on-demand

multipath routing protocol with QoS

support for high-speed MANET. IEEE

Access, 8, pp.44760-44773.

[16] Balshetwar, S.V., Tugnayat, R.M. 2015.

Techniques for Analyzing Framed Data,

Global Journal of Engineering Science and

Researches, 2(8), 80-83.

[17] Aruna, R., Subramanian, R.,

Sengottuvelan, P. and Shanthini, J., 2019.

Optimized energy efficient route assigning

method using related node discovery

algorithm in MANET. Cluster

Computing, 22(1), pp.469-479.

[18] Anugraha, M. and Krishnaveni, S.H.,

2022. SRTE: Security Resource Allocation

for Trust Model in Evaluate the Strong

Node. Webology, 19(1).

[19] Das, M.V., Premchand, P. and Raju, L.R.,

2021. Security Enhancing based on Node

Authentication and Trusted Routing in

Mobile Ad Hoc Network

(MANET). Turkish Journal of Computer

and Mathematics Education

(TURCOMAT), 12(14), pp.5199-5211.

[20] Sirajuddin, M., Rupa, C., Iwendi, C. and

Biamba, C., 2021. TBSMR: a trust-based

secure multipath routing protocol for

enhancing the QoS of the mobile ad hoc

network. Security and Communication

Networks, 2021..

[21] Jari, H., Alzahrani, A. and Thomas, N.,

2021, November. A Novel Indirect Trust

Mechanism for Addressing Black hole

Attacks in MANET. In Proceedings of the

11th ACM Symposium on Design and

Analysis of Intelligent Vehicular Networks

and Applications (pp. 27-34).

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

95

Volume 22, 2023

[22] Gopala Krishnan, C., Nishan, A.H.,

Gomathi, S. and AravindSwaminathan, G.,

2022. Energy and Trust Management

Framework for MANET using Clustering

Algorithm. Wireless Personal

Communications, 122(2), pp.1267-1281.

[23] Kaur, J. and Kaur, S., 2021. Novel trust

evaluation using NSGA-III based adaptive

neuro-fuzzy inference system. Cluster

Computing, 24(3), pp.1781-1792.

[24] Anitha Josephine, J. and Senthilkumar,

S., 2021. Tanimoto support vector

regressive linear program boost based

node trust evaluation for secure

communication in MANET. Wireless

Personal Communications, 117(4),

pp.2973-2993.

[25] Sathyaraj, P. and Rukmani Devi, D.,

2021. Designing the routing protocol with

secured IoT devices and QoS over Manet

using trust-based performance evaluation

method. Journal of Ambient Intelligence

and Humanized Computing, 12(7),

pp.6987-6995.

[26] Kumar, R. and Shekhar, S., 2021. Trust-

Based Fuzzy Bat Optimization Algorithm

for Attack Detection in Manet. In Smart

Innovations in Communication and

Computational Sciences (pp. 3-12).

Springer, Singapore.

[27]Hamza, F. and Vigila, S.M.C., 2021.

Cluster Head Selection Algorithm for

MANETs Using Hybrid Particle Swarm

Optimization-Genetic Algorithm. Int. J.

Comput. Netw. Appl, 8(2), pp.119-129.

[28]Venkatasubramanian, S., Suhasini, A. and

Vennila, C., 2021. An Energy Efficient

Clustering Algorithm in Mobile Adhoc

Network Using Ticket Id Based Clustering

Manager. International Journal of

Computer Science & Network Security,

21(7), pp.341-349.

[29]Kumar, M., 2021. An Optimized

Utilization of Battery Backup in MANET

Using Modified Firefly Algorithm.

Turkish Journal of Computer and

Mathematics Education (TURCOMAT),

12(2), pp.2086-2094.

Contribution of Individual Authors to the

Creation of a Scientific Article (Ghostwriting

Policy)

The authors equally contributed in the present

research, at all stages from the formulation of the

problem to the final findings and solution.

Sources of Funding for Research Presented in a

Scientific Article or Scientific Article Itself

No funding was received for conducting this study.

Conflict of Interest

The authors have no conflicts of interest to declare

that are relevant to the content of this article.

Creative Commons Attribution License 4.0

(Attribution 4.0 International, CC BY 4.0)

This article is published under the terms of the

Creative Commons Attribution License 4.0

https://creativecommons.org/licenses/by/4.0/deed.en

_US

WSEAS TRANSACTIONS on COMMUNICATIONS

DOI: 10.37394/23204.2023.22.7

S. Mohan, P. Vimala

E-ISSN: 2224-2864

96

Volume 22, 2023