Received: January 3, 2023. Revised: April 21, 2023.

Metaheuristic Based Energy Efficient Routing Protocol in MANET Using Battle

Royale Optimization

Vasantha Kumara Mahadevachar1* Naveen Thimmahanumaiah Hosur1

1
Department of Computer Science & Engineering, Government Engineering College, Hassan, India
* Corresponding author’s Email: [email protected]

Abstract: Trust based routing with energy efficiency in mobile ad hoc networks (MANET) is a challenging task due
to the complex nature of the network. Based on the MANET infrastructure, the interoperability of temporary
communication is assured; however, there is no supervisory control process for MANET routing in terms of trust and
security. As a result, stability routing is avoided in MANETs which can damage the network’s dynamic nature. In this
paper, a trust based multi-objective battle royale optimization (T-MOBRO) algorithm is proposed based on the
conventional battle royale optimization (BRO) algorithm. To address the limitation of trust based energy efficiency
routing in conventional BRO, a novel method T-MOBRO is suggested. The T-MOBRO algorithm establishes trust
based routing along with energy efficiency by considering trust, energy, distance as its fitness function. The multi-
objectives of the proposed model are said to be distance, energy consumption, packet forwarding rate (PFR), end to
end delay (EED), direct trust (TrustDT), and indirect trust (TrustIDT). The BRO algorithm is based on gameplay that is
inspired by battle royale video games. The proposed approach is evaluated using throughput, average energy
consumption, packet delivery ratio, end to end delay (EED). The PDR for the developed T-MOBRO method for 10
nodes is 99.58%, throughput is 4032.29 bits/second, end to end delay is less with 0.0070 seconds, and energy
consumption of the MANET is less with 1.7Joules for the T-MOBRO method.
Keywords: Battle royale optimization algorithm, Mobile ad hoc networks, Packet forwarding rate, Trust based routing
protocol, Wireless sensor networks.

vulnerable to several attacks, particularly packet drop

1. Introduction attacks. Delivering long-standing routes may help to
reduce communication delays between network
Infrastructure networks and infrastructure-less
nodes [4]. Hence, secure and energy-efficient
networks are two subclasses of wireless networks. A
protocols are required. The routing protocols should
mobile ad hoc network is a wireless network topology
be designed to reject commonly used attacks such as
that is a part of infrastructure networks with movable
network data intercepting, hijacking, and jamming.
communication routes and mobile network nodes.
Some of the MANET security services include
MANETs are a developing area in wireless
authentication, confidentiality, integrity, anonymity,
networking in which mobile nodes communicate in
and availability which are effective for mobile users.
impromptu or ad hoc networks. The node topology of
Another significant security attribute that needs to be
MANET networks is always in an active state due to
addressed is QoS (quality of service). The QoS
its node mobility [1, 2]. MANET mobile nodes move
constraints are violated when the network defense
randomly and arbitrarily to form a non-permanent
mechanism is poor, as it allows unauthorized access
and effective topology suitable for sensitive and
to the network [5, 6]. Different applications have
crucial applications such as military and emergency
different QoS requirements in terms of data loss,
operations [3]. Routing is regarded as one of the most
postponement, and throughput. QoS management is
significant issues in MANETs due to its active and
critical for providing beginning-to-finish QoS
disseminated nature. MANETs routing protocols are
guarantees. Routing protocols that are already existed
International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 2

such as ad hoc on-demand distance vector popularly routing protocols in MANET. Detailed information
known as AODV, temporally ordered routing about the T-MOBRO method along with its trust
algorithm, and dynamic source routing are factors is given in section 3. The outcomes of the T-
inadequate for real-time applications. A MOBRO method are provided in section 4. whereas
communication network is required by the section 5 provides the conclusion.
application to ensure QoS parameters. [7]. To
improve routing efficiency, the organization of 2. Literature review
MANET into a hierarchy should be improved, which
Sathyaraj [11] developed a secure route analysis
can be achieved through the clustering schemes. The
method for real-time applications that consider the
technique of clustering is performed for re-
trust of IoT devices and their discovered routes of
organizing all nodes into small-sized clusters based
intermediate nodes for secure routing in MANET.
on their regional proximity and the cluster head, with
The technique used two approaches in which the first
each cluster determined by the same rule. The task of
approach involved finding a path list from the source
the cluster head is to supervise cluster operations like
to the destination node. The second approach was
cluster process management, routing table updates,
implemented to evaluate mobile node secure route
and route discovery. Two types of nodes can be
support by evaluating IoT device support. The
observed in the clustering process namely: ordinary
method calculates the data forwarding support value
nodes and gateway nodes. Ordinary nodes are those
(DFS) by taking the no. of IoT devices in the route
in the cluster that are not the cluster head. Gateway
for contemplation. Based on the value of DFS, an
nodes are nodes that have inter-cluster links and can
individual path has been selected to improve
communicate with one or more clusters. If the
MANET's QoS. The proposed method achieved the
destination is within the cluster, packets are sent to
highest efficiency and increased throughput in secure
the cluster head; otherwise, they are sent to the
routing. Alagan Ramasamy Rajeswari [12]
gateway node [8, 9]. Numerous routing protocols
developed a new trust-based secure energy aware
have been proposed based on different network
clustering (TSEAC) model along with two novel
structures, mobility scenarios, and different programs
algorithms namely: an energy-efficient trust-aware
for ad hoc networks. The mobility of nodes is a
secure clustering algorithm and a filter untrustworthy
difficult task for routing protocol designers [10].
recommender (FUR) algorithm. The TSEAC method
The major contribution of the work is discussed
was implemented to build a stable and secure
below:
trustworthy cluster head. The proposed approach
built the cluster head by reducing malicious nodes.
• The trust routing in MANET to a respected
The FUR algorithm was responsible for improving
destination is generated by using the trust the process of clustering by minimizing distortion
based multi-objective battle royale
attacks. A beta distribution technique was also
optimization (T-MOBRO) algorithm. The
implemented to predict the value of node trust in both
algorithm considers trust as the primary cost a direct and indirect manner. The proposed TSEAC
value for avoiding malicious attackers in the
method outperforms the network lifetime compared
route that helps to improve the data delivery
to the CBTRP, AOTDV, and CBRP existing methods.
over the network.
Ankita A. Mahamune [13] implemented an Efficient
• The primary objective of this work is to Trust based Routing Scheme (ETRS) to provide
provide trust measurements in the MANET secure communication in MANET by detecting,
based on trust parameters using a conventional preventing and mitigating malicious nodes. Ad hoc
battle royale optimization algorithm. Further, on-demand distance vector was designed and
the T-MOBRO algorithm is optimized using a implemented within the context of an inherent
trust, energy, and distance. dynamic routing protocol (AODV). The proposed
• The EED and distance used in the T-MOBRO scheme's performance was validated by comparing it
method are used to minimize the delay during to the cutting-edge evolutionary self-cooperative
the transmission by selecting the trusted route trust evolutionary self-cooperative trust (ESCT)
with less traffic and less distance. The energy scheme and the traditional AODV. The proposed
consumed in the network is achieved by framework is validated using performance metrics
mitigating malicious nodes. such as PDR, EED, throughput, and jitter and other
metrics were used to demonstrate the supremacy.
The rest of the paper is arranged as follows: M. S. Usha [14] developed a novel energy
Section 2 offers the related work accomplished on the efficient trust aware routing protocol novel also

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 3

known as NETAR for the conventional AODV to routing protocols in MANETS. Analysing the effect
improve the three degrees of trust between the nodes of various parameters on the performance of the
in the MANET. The nodes' trust was built by network and its variations were researched in this
combining the estimation of trust rate of neighbor- paper. The reactive classes compared were dynamic
node, calculation of bandwidth, energy and source routing (DSR) and ad hoc on-demand distance
prediction of malicious behavior, and performance vector (AODV) and proactive classes compared were
metrics such as PDR, delay, false positive, and destination sequenced distance vector (DSDV) and
throughput to improve network link life time (NLT). optimized link state routing (OSLR). The results have
The NS2 simulator was employed where the shown that OSLR has outperformed in terms of
simulation results outperformed the existing routing throughput, PDR, EED, and consumed energy.
protocols with the proposed NETAR framework. However, OSLR is not suitable for sensor network
Ganesh Kumar Wadhwani [15] developed a trust- which is a major drawback of using this algorithm.
based framework to build a flexible network that was The future plans of this research were to propose a
adaptable to malicious nodes using the ad hoc on- hybrid protocol with high lifetime of the network.
demand distance vector (AODV) routing protocol as Sreenivasulu Ummadisetty [19] proposed a modified
a basic routing protocol. The performance metrics of moth flame optimization algorithm which was
the proposed framework were measured using packet employed in automatic identification of atrial
loss, average packet delay, and varying densities of fibrillation (AF). The proposed approach was
nodes that are malicious. In presence of malicious deployed for detecting AF in short electrocardiogram
nodes, the proposed framework resulted in better (ECG) recordings. Other techniques such as heart rate
performance compared to the basic routing protocol. variability (HRV) and frequency analysis were
Mahaboob John [16] proposed a regional mobility acquired for extracting features. The obtained results
energy approximation-based secured routing have shown that the HRV technique is efficient for
algorithm to address the inadequacy of QoS detecting AF from short ECG recordings. The
performance in MANET for real-time applications. proposed MMFO has a drawback of local optima
The RRME (real-time regional mobility and energy) entrapping, low population diversity, and imbalance
estimation routing method divides the complete between exploitation and exploration phases. The
network into regions, with each node performing future work of this research is to perform cascading
approximations in a neighbor selection based on classification to check the performance of the system.
various features such as mobility and energy. As a N. D. S. S. Kiran Relang [20] proposed an enhanced
result, the method computed the QoS, which was a whale optimization (EWO) algorithm for the
support factor for any route by calculating the data classification of ground water quality by considering
forwarding support, throughput support, and lifetime the real time Indian water quality database. An
maximization support (LMS) (QoSSF). Based on the AlexNet model and KNN were used in classification
QoSSF value, an individual path has been chosen to of optimal features. The proposed optimization
conduct routing that maximizes QoS in MANET. algorithm has achieved greater classification results
Mariappan Rajashanthi [17] developed a secure compared to the existing systems. Other databases
multipath routing protocol with an encryption with large amount of data samples would check the
technique and optimal fuzzy logic. The proposed robustness of the classification model, which is a
framework was dependent on the quality of service drawback for this work. To further improve the
which provided a reliable communication of data. groundwater quality classification, a deep learning
grey wolf optimization's adaptive formation approach based ensemble models can be implemented.
envisioned the optimal path. Hence, an optimal path The recent optimization techniques such as grey
was considered from the known routes to protect the wolf optimization's (GWO) [17], modified moth
techniques of data key management; in this case, the flame optimization algorithm (MMFO) [19], and
homomorphic encryption method was used. The enhanced whale optimization (EWO) [20]. GWO
performance metrics of the proposed framework reduces time consuming tasks and reduces
were EED, PDR, etc. The proposed energy efficient operational time for high dimensional data. MMFO is
framework resulted in better simulation results along used to solve local optima problem. EWO is used in
with the network lifetime compared to the existing solving constrained and unconstrained optimizations
state-of-the-art methods. Future work can be focused problem. However, the discussed algorithms such as
on the communication delay and improve the QoS GWO, MMFO, and EWO was not suited to solve the
which is a limitation of this work. Nahid Ebrahimi problems related to trust and security issues in
Majd [18] developed and approach of analysing MANETs.
reactive and pro-active classes which were optimized From the observations, the limitations of existing
International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 4

optimized power even if the nodes' mobility is

random. Let 𝐺 = (𝑈, 𝑉) be the graph in MANET
with a set of nodes 𝑈 = {𝑢1 , 𝑢2 , … , 𝑢𝑗 , … , 𝑢𝑚 } where
m is the total no. of nodes in MANET. 𝑉 represents
the link set that connects two mobile nodes and is
given by {𝑣1 , 𝑣2 , … , 𝑣𝑗 , … , 𝑣𝑚 }, where 1 ≤ 𝑗 ≤ 𝑟 ≤
𝑚 , and 1 ≤ 𝑗 ≤ 𝑚 respectively. Excess energy is
required for each node in the model to transmit data
from nodes. The distance is considered to be the
primary factor in data transmission from the source
to the destination node.

3.2 Battle royale optimization (BRO) algorithm

Some video game adaptations of the battle royale

genre have players jump out of an aeroplane and land
on the map via parachute. Additionally, BRO begins
with a random population that is evenly distributed
across the problem space, just like other swarm-based
algorithms. Next, each person (soldier/player) fires a
weapon at the soldier who is closest to them in an
attempt to injure them. Therefore, soldiers in better
positions harm their closest neighbors. Its damage
level rises by one when a soldier is injured by another
[21]. The proposed algorithm is diagrammatically
Figure. 1 Flowchart of the proposed MANET model represented in Fig. 2.
These interactions are mathematically calculated
researches has shown that the lack of finding optimal as 𝑥𝑖 . 𝑑𝑎𝑚𝑎𝑔𝑒 = 𝑥𝑖 . 𝑑𝑎𝑚𝑎𝑔𝑒 + 1 where 𝑥𝑖 is the
route, high energy consumption, has decreased the damage level of ith soldiers in the overall population.
robustness of the network. To overcome this a T- The 𝑥𝑖 . 𝑑𝑎𝑚𝑎𝑔𝑒 can be set to zero if the damaged
MOBRO method is introduced which enhances and soldiers can cause harm to their opponents in the next
improves robustness of the network by considering iteration. Additionally, soldiers want to switch
multi-objectives. positions as soon as they take damage to attack
enemies from a different angle. As a result, to
3. Methodology concentrate on exploitation, the damaged solder
moves in the direction of a position midway between
In this research, the discovery of routing protocol its previous location and the best position (elite) thus
using the trust based battle royale optimization(T- far. These interactions are calculated as shown in Eq.
MOBRO) algorithm is developed for achieving (1):
trusted reliable transmission over the MANET, which
is shown as a flowchart in Fig. 1. The generation of 𝑥𝑑𝑎𝑚,𝑑 = 𝑥𝑑𝑎𝑚,𝑑 + 𝑟(𝑥𝑏𝑒𝑠𝑡,𝑑 − 𝑥𝑑𝑎𝑚,𝑑 ) (1)
an energy-efficient trusted route is utilized in
mitigating malicious nodes over the network which Where 𝑟 is a random number from the range [0,1]
can also be used to enhance the packet delivery of the and 𝑥𝑑𝑎𝑚,𝑑 is the damaged soldier’s position in
network. The BRO algorithm is a last-man-standing
d(dimension). If a soldier's damage level rises above
gameplay that is inspired by the Japanese film battle
a predetermined threshold, to concentrate on
royale and the subgenre of the battle royale video
exploration, the soldier dies, respawns at random
game.
from the feasible problem space, and 𝑥𝑖 . 𝑑𝑎𝑚𝑎𝑔𝑒 is
3.1 Mobile ad hoc network model again set to value zero. We found that a threshold
value of 3 was suitable after much trial and error. This
A secure data transmission route path is chosen action prevents hasty convergence and offers better
from the source node (transmitter) to the destination exploration. The soldier returning to the trouble spot
node (receiver) in MANET. To maintain the after being killed looks like this:
connectivity of the network, the model measures

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 5

Figure. 2 Flowchart of battle royale optimization

𝑥𝑑𝑎𝑚,𝑑 = 𝑟(𝑢𝑏𝑑 − 𝑙𝑏𝑑 ) + 𝑙𝑏𝑑 (2) The standard deviation (SD) of the overall
population in the d dimension is represented in this
Where, 𝑢𝑏𝑑 𝑎𝑛𝑑 𝑙𝑏𝑑 are upper and lower bounds equation by 𝑆𝐷(𝑥 ̅̅̅)
𝑑 and the best solution so far is
of d dimensions in the problem spot. Additionally, in represented by 𝑥𝑏𝑒𝑠𝑡,𝑑 . As a result, the 𝑙𝑏𝑑 /𝑢𝑏𝑑 sets
each ∆ iteration, the problems attainable search space to the original 𝑙𝑏𝑑 /𝑢𝑏𝑑 if it surpasses the original
starts to contract in the direction of the ideal lower and upper bound. Additionally, to emphasize
resolution. The initial value was ∆ = elitism, the top player or soldier from every iteration
log10 (𝑀𝑎𝑥𝐶𝑖𝑟𝑐𝑙𝑒) but it later changed to ∆ = ∆ + is retained and regarded as elite. Aside from the
∆ problem's dimensional area, the proposed approach's
𝑟𝑜𝑢𝑛𝑑( ). The maximum number of generations, in
2 computational complexity is also influenced by the
this case, is MaxCircle. Exploration and exploitation
overall population and maximum allowed iterations.
are both aided by this interaction. The lower/upper
Considering the total size of the overall population as
boundary will therefore be changed as shown in Eqs.
n, the computing complexity of overall solutions is
(3) and (4)
𝑂(𝑛2 ) because every solution should be compared to
all others to calculate its distance from all other
𝑙𝑏𝑑 = 𝑥𝑏𝑒𝑠𝑡,𝑑 − 𝑆𝐷(𝑥
̅̅̅)
𝑑 (3)
solutions. Therefore, the computational complexity
of BRO is 𝑂(𝑛3 ) given the number of iterations.
𝑢𝑏𝑑 = 𝑥𝑏𝑒𝑠𝑡,𝑑 + 𝑆𝐷(𝑥
̅̅̅)
𝑑 (4)

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 6

3.3 Energy consumption are present and determined at time 𝜔, 𝑖 and 𝛾 with the
fitness factor falling between 0 and 1. so that the 𝜔 +
One of the most crucial trust metrics is energy 𝛾 = 1. Some of the direct trust must be considered
because it is the main source of life for the sensor while calculating a node's degree of trust based on
node. Mobile sensor nodes face energy efficiency packets received as shown in Eq. 8
issues in the network as sensor nodes move at
different speeds. A node with insufficient energy 𝐷𝑇 ( )
𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝑅𝑋 ( )
𝑖 = 𝑇𝑥𝑦 𝑆𝐷
𝑖 + 𝑇𝑥𝑦 (𝑖) (8)
backup cannot function effectively in the network,
and the network's performance may suffer as a result. 𝑅𝑋 ( )
Here, 𝑇𝑥𝑦 𝑖 be the no. of received packets with
A node with no energy backup is denoted by value 0,
and a node with full energy backup is denoted by the use of node 𝑥 at the time 𝑖 . The total no. of
value 1. Consequently, the range of values intended packets sent by the nodes with the use of 𝑦 at the time
𝑆𝐷
for energy backup is 0 to 1. Clustering manages 𝑖 is represented by 𝑇𝑥𝑦 (𝑖) . The transmission of
energy consumption in the network effectively. packets in the nearby nodes serves as the basis for the
Mathematically, the energy consumption of a node indirect trust degree, which can be defined as shown
𝑁𝑖 at ∆𝑡 time interval is given by: in Eq. (9):

𝐼𝐷𝑇 ( ) 1 𝑁𝑑 𝐷𝑇
𝐸𝑛𝑒𝑟𝑔𝑦𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 (𝑁𝑖 , ∆𝑡) = 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝑖 = (𝑁 ) ∗ ∑𝑑=1 𝑇𝑟𝑢𝑠𝑡𝑑𝑦 (𝑖) (9)
𝑑
𝐸𝑛𝑒𝑟𝑔𝑦𝑟𝑒𝑠 (𝑁𝑖 , 𝑡0 ) − 𝐸𝑛𝑒𝑟𝑔𝑦𝑟𝑒𝑠 (𝑁𝑖 , 𝑡1 ) (5)
𝐷𝑇
Where 𝑇𝑟𝑢𝑠𝑡𝑑𝑦 (𝑖) be the degree of trust that
Where 𝐸𝑛𝑒𝑟𝑔𝑦𝑟𝑒𝑠 is the residual energy of node
exists for existing works in the nearby neighborhoods
𝑁𝑖 at time intervals 𝑡0 and 𝑡1
at the time 𝑖. The 𝑁𝑑 represent the neighboring nodes'
3.4 Trust model of BRO (T-MOBRO) representational numbers. These factors can raise the
level of trust for the detection of malicious nodes that
Trust is a factor that promotes secure and healthy have higher trust values for sending packets with full
communication in MANET by preventing malicious trust. The detection of malicious nodes determines
nodes to communicate in the network to avoid the trust level in some fundamental ways. With the
miscommunication. As a result, the network's trust values in the packet transmission, the nodes may
transmission and reception are started based on the have a wider choice. As a result, each iteration is the
trust level of each node after the trust of all nodes is basis for a model's update.
measured and analysed. In the developed routing
∗ (
model, the trust model offers security. To assess the 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝑖 + 1) = 𝛼 ∗ 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 (𝑖 ) + (1 − 𝛼) ∗
collection of trusted nodes, each node trust factor is 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 (𝑖 + 1) (10)
estimated. The nodes trust value is determined when
the transmission range is less distant from the where 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 ∗ (
𝑖 + 1) be the degree of trust for
neighbour node. A distance-based RSSI is measured the standard measure for a time 𝑖 + 1 and α be the
as shown in Eq. (6). The trust value can be fitness factor that denotes the balanced in the standard
determined using Eq. (7) which evaluates direct and measure for the present and previous iterations, with
indirect trust. the value the specifically in the range of 0< 𝛼 < 1.
𝐷𝑖𝑠𝑡𝑠,𝑖 = 𝑅𝑆𝑆𝐼(𝑁, 𝐺𝑖 ) (6) 3.5 Formulation of the fitness function

Where, N= number of nodes, The main goal is to create a set of multiple-layer

𝐷𝑖𝑠𝑡𝑠,𝑖 is the received signal strength between the energy metrics by creating a novel combined energy
nodes of sender node 𝑆 along with the present metric based on the metrics described above and
neighbor node 𝑖. incorporating it into our fitness function. For each
𝐺𝑖 is the ith node in the graph metric from the source node 𝑁0 to destination node
𝑁𝐾 , a k-hop Path 𝑃 = 𝑁0 , 𝑁1 , … , 𝑁𝐾 is considered.
𝐷𝑇 ( )
𝑇𝑟𝑢𝑠𝑡𝑥𝑦 (𝑖 ) = 𝜔𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝐼𝐷𝑇
𝑖 + 𝛾𝑇𝑟𝑢𝑠𝑡𝑥𝑦 (𝑖) (7) The above metrics are used to calculate the energy
and cost of a node, where each metric is provided
Where, 𝑇𝑟𝑢𝑠𝑡𝑥𝑦𝐷𝑇 ( )
𝑖 i is the fundamental trust with a weight value to adjust the impact of each
routing metric, as shown in the Eq. (11). A node's
among the degree for the currently present node and
𝐼𝐷𝑇 energy fitness function is a single function formed by
a time 𝑖. 𝑇𝑟𝑢𝑠𝑡𝑥𝑦 (𝑖), i is the indirect trust that is combining the weighted metrics, and it aims to
present to some extent for the neighboring nodes that
International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 7

minimize the weighted sum: Table 1. Simulation parameters

Parameters Values
𝑃𝑇(𝑖) 𝑃𝑅(𝑖) 𝑁 No. of nodes 10 - 50
𝐶𝑛 (𝑖 ) = 𝛼1 ( + ) + 𝛼2 +
𝑃𝑇𝑚𝑎𝑥 𝑃𝑅𝑚𝑎𝑥 𝑁𝑚𝑎𝑥 Initial energy 0.5J
𝐸𝑅(𝑖)
𝛼3 (1 − ), (11) Area size 1000𝑚 × 1000𝑚
𝐸0 Range of transmission 50𝑚
Packet size 512 bytes
∑3 α = 1, Mobility model Random waypoint movement
{ i=0 i (12)
0 ≤ αi ≤ 1.

Where, 𝑃 𝑇(𝑖) =Tx power of node 𝑁𝑖 , can be regarded as reliable. Therefore, it is always
𝑃𝑇𝑚𝑎𝑥 = maximum Tx power, best to pick a route with lots of reliable nodes. The
𝑃𝑅(𝑖) =Rx power by node 𝑁𝑖 , packet inflow and outflow are considered when
𝑃𝑅𝑚𝑎𝑥 =Maximum Rx power calculating the PDR. Let's say that 𝛼 and, 𝛽
𝑁=Node Connectivity Index of node 𝑁𝑖 , respectively, represent the inflow and the outflow.
𝑁𝑚𝑎𝑥 =Number of nodes in the The following condition from Eq. (15) is met by a
network minus one normal node:
𝐸𝑅(𝑖) =Remaining energy capacity of node 𝑁𝑖 ,
𝐸0= energy capacity of node 𝑁𝑖 initially 𝛼=𝛽 (15)

This method's fundamental goal is to maximize Assume that the node doesn't show any interest in
energy efficiency by choosing the smallest Tx and Rx forwarding packets when its packet outflow is less
power, node connectivity index, and consumed than half of its packet inflow, or a little plus or minus.
energy, which is nothing more than the maximum The following condition can be used to indicate this:
amount of available energy:
𝛽
𝛼= 2
(16)
𝐸𝑐𝑜𝑛𝑠 (𝑖) 𝐸0 −𝐸𝑟 (𝑖) 𝐸𝑟 (𝑖)
𝐸0
= 𝐸0
= 1− 𝐸0
(13)
The node is completely unreliable and unfit to be
a part of the network when the 𝛽 rate is less than half
Where, 𝐸𝑐𝑜𝑛𝑠 (𝑖) =Consumed energy by node 𝑁𝑖
𝐸0=energy capacity initially for node 𝑁𝑖 , 𝐸𝑟 (𝑖)= of 𝛼. These types of nodes are taken into account
when choosing the route, making data routing risk-
Remaining energy capacity of the node 𝑁𝑖
free and secure. As a result, PFR is yet another crucial
The optimization metric is the path cost from
node trust metric that can help to improve route
source to destination nodes. The objective is to find
reliability.
the path that has the lowest total cost. The energy cost
function which is represented in Eq. (11) for the new
node is used to calculate the path costs between each 4. Experimental results
source to destination nodes. Eq. (14) is the fitness The outcomes of the proposed T-MOBRO
function of this work. method are shown and explained in this section. The
implementation and simulation of T-MOBRO
𝐶𝑝(𝑃𝑖 ) = ∑𝑛𝑖=0 𝐶𝑛 (𝑖) (14) method are done in MATLAB R2020a. The T-
MOBRO method is used to improve the efficient
Where 𝐶𝑝(𝑃𝑖 ) is the cost of the path (𝑃𝑖 ), 𝐶𝑛 (𝑖) energy in MANET. The network is initialized with 50
is the energy cost of the node 𝑁𝑖 , 𝑛 represents a nodes and initial energy of 0.5J are deployed in the
number of nodes in the path. area of 1000𝑚 × 1000𝑚. The range of transmission
is 50𝑚. Table 1 provides the simulation parameters
3.6 Packet forwarding rate (PFR) considered while analyzing the T-MOBRO method.
By taking into account the packet flow, the PFR 4.1 Performance analysis
is designed to analyze the node's nature. All packets
that are meant to be forwarded must be forwarded by The performance of the trust battle royale
a normal node. However, there are instances where optimization (T-MOBRO) method is evaluated by
the malicious nodes fail to perform the routing task, using PDR, throughput, EED and energy
which causes a serious problem. When a node consumption. The performances are analyzed with
expresses an interest in forwarding packets, that node conventional optimization namely trust-based secure

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 8

Table 2. Comparison of PDR improved along with secure routing with the
PDR (%) proposed optimization technique by implementing
Numb T- the T-MOBRO algorithm within the MANET which
TSEA ETRS MMF EWO
er of MO is the main issue addressed in the literature review
C [12] [13] O [19] [20]
nodes BRO section.
10 90.24 50.00 85.33 82.34 48.49
20 85.45 55.00 80.73 80.23 46.78 4.1.1. Packet delivery ratio
30 72.63 65.00 75.52 76.46 44.26
40 65.75 75.00 68.74 74.34 41.63 The packet delivery ratio is calculated by dividing
50 60.93 75.00 64.32 72.55 38.29 the number of packets received at the destination by
the number of packets sent by the transmitter node.
The comparison of PDR for T-MOBRO with TSEAC
[12], ETRS [13], MMFO [19], EWO [20], and T-
MOBRO is shown in Table 2 and represented in Fig
3. Fig. 3 shows that the T-MOBRO achieves better
PDR than both the methods. For example, the PDR
of the T-MOBRO with 10 nodes is 48.49% which is
high when compared to the other models. The
mitigation of malicious attacks using T-MOBRO is
used to avoid data loss while broadcasting the data
packets.
4.1.2. Throughput
Figure. 3 Analysis of packet delivery ratio
Throughput is the no. of packets that are
successfully received at the destination over the
MANET. The comparison of throughput for T-
MOBRO with TSEAC [12], ETRS [13], MMFO [19],
EWO [20], and T-MOBRO is shown in table 3 and
represented in Fig 4. Fig. 4 shows that the T-MOBRO
has higher throughput when compared to the all
methods. For example, the throughput of the T-
MOBRO for 10 nodes is 4032.29 bits/sec which is
high when compared to the both the methods. The
throughput of T-MOBRO is increased because of
Figure. 4 Analysis of throughput avoiding malicious attacks on the route. The
development of trust based energy-efficient route
energy-aware clustering (TSEAC) and efficient trust over the MANET increases the robustness against
based routing scheme (ETRS) which provides its malicious attacks that helps to improve the data
energy-efficient routing. The results are analyzed by delivery.
varying the no. of nodes that exist in the network. The 4.1.3. End to end delay
comparative analysis of the T-MOBRO is provided
in this section where existing researches such as EED is defined as the time taken for packet
TSEAC [12] and ETRS [13] is used to perform the transmission from the source to the destination. The
comparison. Further, the recent optimization comparison of EED for T-MOBRO with TSEAC [12],
techniques such as MMFO [19], and EWO [20] are ETRS [13], MMFO [19], EWO [20], and T-MOBRO
developed for performing trust based routing in is shown in Table 3 and represented in Fig. 4 is shown
MANETs to evaluate the performance of T-MOBRO. in Table 4 and represented in Fig 5. Fig. 5 shows that
The aforementioned approaches are implemented the T-MOBRO achieves less EED than all the
with the same specifications as shown in Table 2. methods. For example, the EED of the T-MOBRO
The graphical comparison of PDR, Throughput, with 10 nodes is 0.0070S which is less when
EED, and Energy consumption is shown in Figs. 3, 4, compared to the other methods. The EED of T-
5, and 6. The trust-based energy efficient secure MOBRO is minimized by identifying the path with
optimal route is implemented to increase the lesser delay and less transmission distance. Further,
robustness of the network. The energy efficiency is the formulated cost function of T-MOBRO is used to

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
 Received: January 3, 2023. Revised: April 21, 2023. 9

Table 3. Comparison of throughput(bits/sec)

Number of Throughput (bits/sec)
nodes TSEAC [12] ETRS [13] MMFO [19] EWO [20] T-MOBRO
10 6560.00 4553.18 5325.26 5034.24 4032.29
20 6800.00 3766.93 4847.34 4839.57 3423.45
30 5160.00 4097.06 4432.36 4677.65 3346.23
40 5520.00 4389.55 4032.17 4338.54 3273.06
50 4860.00 4388.43 3924.32 4033.34 3045.32

Table 4. Comparison of EED minimize the routing overhead that is further used to
EED(Sec) minimize the EED.
Numb
TSE T-
er of ETRS MMF EWO
AC MOB 4.1.4. Energy consumption
nodes [13] O [19] [20]
[12] RO
10 0.29 0.0073 0.0245 0.0372 0.0070 Table 5 depict the analysis of the consumption of
20 0.21 0.0120 0.0226 0.0350 0.0116 energy for the proposed T-MOBRO algorithm with
30 0.25 0.0115 0.0189 0.0329 0.0102 existing methods such as TSEAC [12], ETR [13],
40 0.22 0.0097 0.0148 0.0298 0.0081 MMFO [19], EWO [20] for various numbers of nodes
50 0.21 0.0076 0.0125 0.0240 0.0065 considered. Fig. 6 shows that the proposed T-
MOBRO consumed less energy compared to all other
methods. For example, the energy consumption by T-
MOBRO method with 10 nodes is 1.7J which is less
when compared to the TSEAC.

5. Conclusion
In this research, the trust based route discovery is
performed using the T-MOBRO method. The trust
based battle royale optimization algorithm is
optimized by using a trust, EED, energy, and distance.
The trust value incorporated in the T-MOBRO is used
Figure. 5 Analysis of EED to avoid malicious attacks during route generation. In
route discovery, the node failure is avoided using the
Table 5. Comparison of energy consumption residual energy whereas the distance and EED are
Energy consumption(Joules) used to identify the route with less distance. The
Numb MMF T- shortest path generation and lesser routing overhead
TSEAC ETR EWO
er of O MOB
[12] S [13] [20] are used to minimize the delay while transmitting the
nodes [19] RO
data in MANET. From the results, it is concluded that
10 2.1 3.2 3.8 3.4 1.7
20 2.9 3.6 4.2 3.7 2.5
the T-MOBRO method achieves better performance
30 4 3.9 4.6 3.9 3 than the TSEAC, ETRS, MMFO, and EWO. The
40 4.5 4.3 5.1 4.1 4 PDR of the T-MOBRO method for 10 nodes is
50 5 4.8 5.4 4.3 4.5 48.49% which is high when compared to the TSEAC.
The throughput of T-MOBRO method is 4032.29
which is less compared to other methods. End to end
delay of the proposed T-MOBRO method is
0.0070seconds which is less compared to the other
methods at node 10. Energy consumption achieved at
node 10 with the T-MOBRO technique is 1.7J which
is less compared to the existing technique. The main
aim of this work is to establish a trust based routing
protocol. Future research can be focused on
establishing a secure routing protocol in MANET by
using the proposed optimization method by detecting
and mitigating malicious attacks.
Figure. 6 Analysis of energy consumption

International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
Received: January 3, 2023. Revised: April 21, 2023. 10

Notation table [2] G. C. Krishnan, A. H. Nishan, S. Gomathi, and

Parameters Representation G. A. Swaminathan, “Energy and trust
𝑈 Set of nodes management framework for MANET using
𝑉 Link set of nodes clustering algorithm”, Wireless Personal
𝑚 Total number of nodes Communications, Vol. 122, No. 2, pp. 1267-
Damage level of 𝑖𝑡ℎ 1281, 2022.
𝑥𝑖 . 𝑑𝑎𝑚𝑎𝑔𝑒
soldier [3] T. A. N. Abdali, R. Hassan, R. C. Muniyandi, A.
damaged soldier’s position H. M. Aman, Q. N. Nguyen, and A. S. A.
𝑥𝑑𝑎𝑚,𝑑
in dimension d
Khaleefa, “Optimized particle swarm
𝑟 random number
Upper bounds of
optimization algorithm for the realization of an
𝑢𝑏𝑑 enhanced energy-aware location-aided routing
dimension d
lower bounds of dimension protocol in MANET”, Information, Vol. 11, No.
𝑙𝑏𝑑 11, p. 529, 2020.
d
best solution in dimension [4] K. Sudhakar, N. Sengottaiyan, and S.
𝑥𝑏𝑒𝑠𝑡,𝑑
d Anbukaruppusamy, “A hybrid swarm intelligent
Standard Deviation of the framework to support efficient military
𝑆𝐷(𝑥 ̅̅̅)
𝑑 overall population in the d communication in MANET”, Journal of
dimension Ambient Intelligence and Humanized
computational complexity Computing, Vol. 12, No. 5, pp. 5215-5223, 2021.
𝑂(𝑛3 )
overall population [5] K. Ourouss, N. Naja, and A. Jamali, “Defending
residual energy of node 𝑁𝑖 against smart grayhole attack within MANETs:
𝐸𝑛𝑒𝑟𝑔𝑦𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 (𝑁𝑖 , ∆𝑡)
at time intervals ∆𝑡
A reputation-based ant colony optimization
Direct trust of ith node
𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝐷𝑇 (𝑖) approach for secure route discovery in DSR
with the use of x and y
nodes protocol”, Wireless Personal Communications,
Indirect trust of ith node Vol. 116, No. 1, pp. 207-226, 2021.
𝑇𝑟𝑢𝑠𝑡𝑥𝑦 𝐼𝐷𝑇
(𝑖) with the use of x and y [6] M. Thebiga and S. R. Pramila, “Adaptable and
nodes energy efficacious routing using modified
𝜔 Time constant emperor penguin colony optimization multi-
𝛾 fitness factor faceted metaheuristics algorithm for MANETS”,
Degree of trust from Wireless Personal Communications, Vol. 118,
𝑅𝑋 (𝑖)
𝑇𝑥𝑦 receiver side at time 𝑖 with No. 2, pp. 1245-1270, 2021.
the use of x and y nodes [7] N. Veeraiah and B. T. Krishna, “An approach for
Degree of trust from optimal-secure multi-path routing and intrusion
𝑆𝐷
𝑇𝑥𝑦 (𝑖) sender side at time 𝑖 with detection in MANET”, Evolutionary
the use of x and y nodes Intelligence, Vol. 15, No. 2, pp. 1313-1327.
𝐶𝑝(𝑃𝑖 ) cost of the path (𝑃𝑖 )
[8] B. K. Tripathy, S. K. Jena, P. Bera, and S. Das,
S, “An adaptive secure and efficient routing
Conflicts of interest protocol for mobile ad hoc networks”, Wireless
The authors declare no conflict of interest. Personal Communications, Vol. 114, No. 2, pp.
1339-1370, 2020.
Author contributions [9] K. Karthick and R. Asokan, “Mobility aware
quality enhanced cluster based routing protocol
The paper background work, conceptualization, for mobile ad-hoc networks using hybrid
methodology, dataset collection, implementation, optimization algorithm”, Wireless Personal
result analysis and comparison, preparing and editing Communications, Vol. 119, No. 4, pp. 3063-
draft, visualization have been done by first author. 3087, 2021.
The supervision, review of work and project [10] B. Devika and P. N. Sudha, “Power optimization
administration, has been done by second author. in MANET using topology management”,
Engineering Science and Technology, an
References International Journal, Vol. 23, No. 3, pp. 565-
[1] V. V. Sarbhukan and L. Ragha, “Establishing 575, 2020.
secure routing path using trust to enhance [11] P. Sathyaraj and R. D. Devi, “Designing the
security in MANET”, Wireless Personal routing protocol with secured IoT devices and
Communications, Vol. 110, No. 1, pp. 245-255, QoS over Manet using trust-based performance
2020. evaluation method”, Journal of Ambient
International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01
Received: January 3, 2023. Revised: April 21, 2023. 11

Intelligence and Humanized Computing, Vol. 12, 214-223, 2023, doi: 10.22266/ijies2023.0228.19.
No. 7, pp. 6987-6995, 2021. [21] T. R. Farshi, “Battle royale optimization
[12] A. R. Rajeswari, S. Ganapathy, K. Kulothungan, algorithm”, Neural Computing and Applications,
and A. Kannan, “An efficient trust‐based secure Vol. 33, No. 4, pp. 1139-1157, 2021.
energy‐aware clustering to mitigate trust [22] K. S. Shivakumar and V. C. Patil, “An optimal
distortion attack in mobile ad‐hoc network”, energy efficient cross-layer routing in
Concurrency and Computation: Practice and MANETs”, Sustainable Computing: Informatics
Experience, Vol. 33, No. 13, p. e6223, 2021. and Systems, Vol. 28, p. 100458, 2020.
[13] A. A. Mahamune and M. M. Chandane, “An
Efficient Trust-Based Routing Scheme Against
Malicious Communication in MANET”,
International Journal of Wireless Information
Networks, Vol. 28, No. 3, pp. 344-361, 2021.
[14] M. S. Usha and K. C. Ravishankar,
“Implementation of trust-based novel approach
for security enhancements in MANETs”, SN
Computer Science, Vol. 2, No. 4, p. 257, 2021.
[15] G. K. Wadhwani, S. K. Khatri, and S. K. Mutto,
“Trust framework for attack resilience in
MANET using AODV”, Journal of Discrete
Mathematical Sciences and Cryptography, Vol.
23, No. 1, pp. 209-220, 2020.
[16] Y. M. M. John and G. Ravi, “Retracted: Real
time regional mobility energy feature
approximation‐based secure routing for
improved quality of service in MANET”,
International Journal of Communication
Systems, Vol. 35, No. 16, p. e4713, 2022.
[17] M. Rajashanthi and K. Valarmathi, “A secure
trusted multipath routing and optimal fuzzy
logic for enhancing QoS in MANETs”, Wireless
Personal Communications, Vol. 112, No. 1, pp.
75-90, 2020.
[18] N. E. Majd, N. Ho, T. Nguyen, and J. Stolmeier,
“Evaluation of parameters affecting the
performance of routing protocols in mobile ad
hoc networks (MANETs) with a focus on energy
efficiency”, In: K. Arai, and R. Bhatia (eds),
Advances in Information and Communication:
Proceedings of the 2019 Future of Information
and Communication Conference (FICC), San
Francisco, USA, Vol. 2, pp. 1210-1219, 2020,
Springer Cham.
[19] S. Ummadisetty and M. Tatineni, “Automatic
Atrial Fibrillation Detection Using Modified
Moth Flame Optimization Algorithm”,
International Journal of Intelligent Engineering
and Systems, Vol. 16, No. 1, pp. 435-445, 2023,
doi: 10.22266/ijies2023.0228.38.
[20] N. D. S. S. K. Relangi, A. Chaparala, and R.
Sajja, “Effective Groundwater Quality
Classification Using Enhanced Whale
Optimization Algorithm with Ensemble
Classifier”, International Journal of Intelligent
Engineering and Systems, Vol. 16, No. 1, pp.
International Journal of Intelligent Engineering and Systems, Vol.16, No.4, 2023 DOI: 10.22266/ijies2023.0831.01