Multimedia Tools and Applications

https://doi.org/10.1007/s11042-024-19542-9

Secure and efficient message transmission in MANET using

hybrid cryptography and multipath routing technique

Jayantkumar A. Rathod1 · Manjunath Kotari2

Received: 28 July 2023 / Revised: 9 April 2024 / Accepted: 28 May 2024

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024

Abstract
MANETs, or Mobile Ad hoc Networks, are a group of self-organized mobile devices that
can connect with one another without relying on a centralized infrastructure or author-
ity. The key benefit of MANETs lies in their ability to handle mobility while facilitating
data communication among users within the network. However, ensuring the security of
MANETs is essential for their proper functioning, particularly in terms of maintaining data
confidentiality and integrity during transmission. MANETs are highly vulnerable to attacks
due to their inherent properties.Clustering is an energetic technique employed in MANETs
to handle mobile nodes effectively. It involves dividing the network into sub-networks
known as clusters, which can consist of overlapping or disjointed nodes. For each cluster,
an enhanced node termed the Cluster Head (CH) is chosen to oversee routing tasks. This
approach reduces the operating expense on member nodes and improves the overall system
efficiency. However, the relationship between nodes and CHs can change arbitrarily, result-
ing in re-union and re-grouping in the clustered MANET.
In a MANET, which is characterized by its dynamic and unpredictable nature, multiple
path discovery plays a vital role in maintaining connectivity, improving reliability, optimiz-
ing resource utilization, and ensuring resilience to network conditions and node failures.
This study aims to create multiple paths between a source and destination using a hybrid
protocol that combines AODV (Ad hoc On-Demand Distance Vector) and MBOMRP
(Multi-Path Byzantine OLSR). Additionally, a hybrid cryptographic technique is applied
to securely transmit fragmented data. The primary objective of this proposed solution is to
guarantee the appropriate, secure, and accurate delivery of packets. The results of our study
demonstrate that the transmission of encrypted messages through multiple paths enhances
network performance, efficiency, and the secure transmission of data within these self-con-
figuring and infrastructure-less networks. It achieves these benefits while minimizing band-
width consumption during routing operations.

Keywords Mobile Ad-Hoc Network (MANET) · Clustering · Energy efficiency ·

Cryptographic techniques · Multipath route discovery · AODV · MBOMRP

Extended author information available on the last page of the article

13
Vol.:(0123456789)
 Multimedia Tools and Applications

1 Introduction

The speedy progress of the Internet, alongside diverse wireless networking technologies,
has resulted in a broad applicability. Among these technologies, Mobile Ad hoc Networks
(MANETs) show immense potential for wireless network research and application devel-
opment [1]. The field of wireless communication has witnessed significant growth, with ad
hoc networks emerging as a dynamic and promising area of study. These networks have the
ability to operate independently or connect with different networks and the Internet invari-
ous ways, presenting opportunities for innovative applications. Examples of such applica-
tions include managing road safety, monitoring households, implementing health monitors,
conducting tragedy and recovery actions, supporting security activities, supervising weap-
ons, and facilitating robotics, among others [2].Mobile Ad hoc Networks (MANETs) serve
as a crucial communication infrastructure in various domains, enabling flexible connec-
tivity in different scenarios. They find extensive applications in emergency and recovery
activities, army and war zone, AI drive models, seminars, health tracking, home automa-
tion, and safety-related applications. Furthermore, the application of MANETs in drone
technology has gained significant attention, particularly in areas such as environmental
control, including forest fire management, pollution monitoring, and surveillance [3].
Mobile ad hoc networks face different challenges in terms of communication, particularly
in adverse environments. The communication process involves two crucial phases: route
discovery and data transmission. Unfortunately, both of these phases are prone to various
forms of attacks. In the route discovery phase, adversaries can undermine the process by
assuming the identity of the destination, providing outdated or corrupted routing informa-
tion, or spreading falsified control traffic [4]. These malicious actions hinder the dissemi-
nation of genuine route control traffic and negatively impact the overall understanding of
network topology among legitimate nodes. Furthermore, adversaries can also disrupt the
data transmission phase, leading to significant data loss. They can achieve this by tamper-
ing with or fraudulently redirecting data traffic, dropping data packets, or even inserting
fabricated data packets into the network [5].
To ensure comprehensive security in MANET communication, it is essential to pro-
tect both the route discovery and data transmission phases. While secure routing protocols
play a crucial role in guaranteeing the accuracy of the discovered topology information,
they alone cannot ensure the secure and uninterrupted delivery of transmitted data [6]. It is
important to recognize that safeguarding the integrity and availability of data requires addi-
tional measures beyond secure routing protocols.The objective of this study is to develop
an energy-efficient Multipath routing protocol that enables reliable and secure informa-
tion exchange in MANETs. The protocol aims to increase network lifetime by optimiz-
ing energy usage by adopting the clustering technique. Also, for ensuring secured message
transmission a combination of cryptographic techniques is applied for each fragmented
message with minimum delay.
With the dynamic and unstable nature of the topology in MANETs, these networks
encounter higher levels of delays and loss of packets compared to stable networks. The
mobility of nodes in MANETs introduces the challenge of limited power due to battery
dependency. Hence, it becomes imperative to explore power-saving approaches, enhance
energy efficiency, and extend the lifespan of the network. MANET nodes have uncondi-
tional flexibility and can establish connections with one another without centralized track-
ing. Consequently, managing network and routing are distributed among all nodes in the
system. These features render MANETs more vulnerable to routing and security concerns

13
Multimedia Tools and Applications

compared to traditional networks. Therefore, it is crucial to develop effective mechanisms

that address the challenges arising from node mobility when making routing decisions.
Clustering is an effective technique employed in managing mobile nodes in larger
dimensions and dense MANETs. It involves dividing the network into groups, known as
clusters [7], which can either have overlaying nodes or be separate. Each cluster is led by
a special node known as the Cluster Head (CH), which handles routing-related tasks. The
use of clustering reduces the overhead on member nodes and improves system efficiency.
In clustered MANETs, the connection among nodes and CHs can change arbitrarily,
encouraging to re-associations and re-clustering processes. Clustering proves particularly
valuable in large dimensions and dense MANETs for the efficient management of mobile
nodes.
To enable effective networking and find optimal paths between nodes, a routing protocol
must be efficient and reliable [8]. This research focuses on developing a routing protocol
that creates the longest possible route between source and destination nodes while ensur-
ing the secure and perfect delivery of fragmented messages. The objective is to minimize
usage of bandwidth, and loss of packet, and control overhead, thus improving network per-
formance. Multipath routing plays a crucial role in achieving these goals by discovering
multiple routes between the source and destination nodes. By utilizing multiple paths, mul-
tipath routing optimizes bandwidth usage and conserves energy in member nodes. It offers
various benefits such as designing resilient networks and evenly distributing network load.
Multipath routing is a valuable approach for enhancing network performance and address-
ing the challenges of efficient and reliable message delivery [9, 10]. The MANETs security
[7] is a big hindrance to the fine functioning of the network. For secured message trans-
mission, we are utilizing five types of crypto techniques and each fragment of message is
encrypted with a different encryption algorithm. By combining symmetric and asymmetric
encryption, hybrid cryptography leverages the performance of symmetric encryption for
larger data encryption and the security advantages of asymmetric encryption for secure
key exchange. This approach is commonly used in modern cryptographic protocols and
systems.

1.1 Key contributions

The primary contribution of this work is to structure a multi-routing protocol for MANETs
that prioritizes security, energy efficiency, stability, and successful data transmission. The
proposed protocol’s significant contribution lies in its ability to provide secure and stable
communication while enhancing the likelihood of successful data delivery.The major con-
tribution of the proposed work are:

(i) Route Discovery: Selection of efficient multiple paths from source to destination
using the Clustering technique.
(ii) Transmission of fragmented messages through multiple paths using cryptography
techniques. By randomly selected cryptography techniques to each fragmented mes-
sage which is sent through different paths.

1.2 Organization of the paper

The remaining sections of the paper are designed as below: The 2 segment explains
an overview of existing research conducted in the field. The 3 section outlines the

13
 Multimedia Tools and Applications

model proposed in this study, encompassing the clustering approaches for MANETs,
including cluster head identification, formation of clusters, cryptographic algo-
rithms, and secure fragmented message transmission. The Results segment explains
the experimental validation of the proposed methodology, comparing it with previous
approaches. Finally, the 5 part summarizes the findings of the study and discusses
potential future directions for research.

2 Related work

Numerous clustering schemes and routing protocols have been developed for MANETs to
tackle many network-related challenges, such as scalability, energy efficiency, and mobil-
ity. In line with our objectives, the research on MANET routing protocols focuses on three
categories: cluster-based techniques, multipath routing, and cryptographic techniques.
These categories are examined and studied to explore their applicability in secure MANET
applications.Cluster-based routing protocols are widely utilized in MANETs, including
energy-depended, connectivity-depended, mobility-depended, and weighted protocols [11].
In a study [12], a clustering protocol called SYN was proposed, focusing on energy effi-
ciency. This protocol utilized the Max-heap tree to represent the different energy grades
of the nodes. Additionally, in another study [6], the authors introduced a protocol based
on Particle Swarm Optimization (PSO) to optimize clusters and conserve energy in ad-hoc
networks.
The MBHC (Mobility-Based Hierarchical Clustering) method utilizes node mobility
patterns to form clusters [13]. The identification of a CH depending on the least reliability
value among neighboring nodes. The associated flexibility among two nodes is estimated
by considering the distance they travel over time, enabling the merging of clusters with
similar mobility patterns. Cluster formation takes into account the similarity in velocity
and direction, grouping nodes with lower mobility differences together in the same clus-
ter. This scheme offers the benefits of minimizing the number of clusters by considering
the mobility pattern and referring to scalability concerns for routing protocols in larger
MANETs. In another study [14], the cluster head is chosen as the node havingthe min-
imizedID that can be controlledby the remaining energy. The cluster head manages the
energy consumption within the cluster and adjusts the loads for transmitting depending on
the available nodes energy of the cluster.
Extensive studiesare conducted on networks formed by clustering that utilize multipath
routing protocol in MANETs. AOMDV (Ad-hoc On-demand Multipath Distance Vector)
[15] is an enhanced version of AODV (Ad-hoc On-demand Distance Vector) [11], specifi-
cally designed for multipath routing. It aims to discover loop-free routes while exhibiting
link disjoint characteristics. Another protocol, SMR (Split Multipath Routing), employs
maximum disjointness to establish multipath [16]. Load balancing becomes essential when
utilizing multiple paths. EIDM (Energy Interference and Drop rate-based Multipath) intro-
duces a novel weight metric for path selection. It combines factors such as available energy,
interference, rate of packet drop, and node mobility [17]. MEER (Multipath Energy-effi-
cient Routing) is an energy-efficient multipath routing protocol that employs energy control
mechanisms to prolong the network’s lifespan [18].
In a study [19], a user-selectable encryption approach was developed to enhance the
security of MANETs. This approach utilizes popular encryption algorithms such as
Data Encryption Standard (DES), Triple DES (3DES), Advanced Encryption Standard

13
Multimedia Tools and Applications

(AES), and the Diffie-Hellman Key Exchange (DHKE) protocol for key management.
The efficiency of the approach was carried out using the Network Simulator-2 (NS-2),
focusing on data transmission duration and throughput of network. The results highlight
the superior performance of AES compared to other encryption schemes.Another study
[20] introduced a novel hybrid cryptography methodology for secure data transmission
in MANET. The Ad Hoc On-Demand Vector Routing (AODV) protocol is employed to
maintain paths between nodes and ensure loop-free communication in the event of link
failures. The Secure AODV (SAODV) protocol, a trust-based secure routing technology,
enhances the security of MANETs. The designed cryptographic routing method was
implemented in the NS2 network simulation environment, demonstrating good packet
delivery ratio, throughput, and energy efficiency.
In a study [21], a generation of the GK method for intra-group connection was pre-
sented, with a specific focus on polynomial generation to establish secure inter-group
keys. The proposed approach allowed members of the group and controllers to share
the intra-group key without the need for encryption or decryption. These techniques
effectively reduced the number of re-keying messages during group changes by utilizing
polynomials. While the polynomial-based key generation provided enhanced security
for group communication, it incurred a higher computational rate. This was particularly
evident when nodes joined or left the operation, as the polynomials needed to be regen-
erated and the key refreshed.In a study by [22], they proposed a scalable and lightweight
approachto address authentication problems in multicast MANET. But, the challenges
regarding hop-to-hop connectivity and integrity left unresolved. Another approach by
[23] involved the design of a Trust-based clustering algorithm to detect malicious nodes
through reduced communication and enhance the routing process in MANETs. How-
ever, the security level of their solution was deemed insufficient.

3 Proposed work

The proposed methodology is separated into two phases.

In Phase-1: Initially,the cluster is formed and Cluster Head Selection Algorithm

(CHSA) is adopted for the CH selection. Based on the CH multiple paths are discovered
by combining the route discovery mechanism of AODV and the mobility-based path
selection of MBOMRP multipath routing protocol. Detailed steps are explained in Sec-
tion 3.1 with a block diagram in Fig. 1.

Deploy Nodes Evaluate total Find CH of

Initial Source
Initiate node number of Source Code
properties Cluster

Apply Routing Broadcast R-

Multiple Path
Protocol REQ
Discovery
AODV+MBOMRP

Fig. 1  Block diagram of clustered multipath routing algorithm

13
 Multimedia Tools and Applications

In Phase-2: To transmit fragmented messages through multiple paths. Each message

isfragmented andusing Round-Robin Algorithm each fragment is distributed among
paths. As a next step, we are randomly selecting a cryptographic technique for each
fragment. Four different techniques are used namely AES, RSA, SHA-256, and Blow-
fish cryptography techniques are applied. Detailed steps are explained in Section 3.2
with a block diagram in Fig. 3. As the final step, the efficiency of the proposed method-
ology is measured using different parameters.

3.1 Phase 1:Cluster‑based multipath routing algorithm

The proposed methodology is shown in the below block diagram Fig. 1. The network
is separated into clusters, each comprising Cluster Heads (CHs) and cluster member
nodes, with each node having a unique cluster ID.
During initialization, all network nodes broadcast HELLO messages to update their
status values, including energy and mobility information. The primary need of exchang-
ing HELLO messages is to determine the range of transmitting nodes. By measuring
the signal strength within the pair ofnodes, the distance between them can be estimated,
every node are allowed to identify its associated nodes fallingwithin the specified range
of transmission. The steps included in the initiation steps are as follows:

Cluster Formation

1. Initially, network nodes exchange periodic HELLO messages to inform their neighbors about their exist-
ence. These messages include the node’s ID, energy level, and mobility value
2. Neighbour discovery involves each node compiling a set of its associates based on the HELLO messages
accepted from other nodes
3. Once the neighbours have been discovered, the procedure of selecting the Cluster Head commences

3.1.1 Selection of cluster head

The proposed methodology incorporates a Cluster Head Selection Algorithm (CHSA)

for choosing the CH [24]. The algorithm determines the CH by selecting the reliabili-
typair factor and the node’s maximum energy. The node with the highest energy level is
termed as the CH, while the subsequent nodes are assigned as substitute Cluster Heads.
The alternate cluster heads serve two purposes: firstly, when the CH energy diminishes,
the substituteCH automatically takes over as the new CH; secondly, a second SCH is
employed to handle CH failures.

Cluster Head Selection Algorithm (CHSA)

• The Cluster Head (CH) is determined by computing the parameters NE (Node Energy), ND (Node Dis-
tance), and FRP (Fault Ratio Probability) laying base station and the chosen higher energy node CH ( NE
,ND,FRP)
• Once the CH is identified, a request message (REQ) is broadcastedto every node of the network. If nodes
are within a similar path and accept the REQ message, they respond to the CH accordingly
rem rem
• Compute FRP = ((Ep(d(p,q,t))
,Eq )

Where d representsthe distance between node p and node q at timestamp

13
Multimedia Tools and Applications

Cluster Head Selection Algorithm (CHSA)

• After the CH is determined, it identifies analternative cluster head (SCH) depending on the computation
of CH > SCH > MN (CH has a higher priority than SCH, which has a higher priority than group nodes)
• The CH updates the routing table in the base station, which contains information about the cluster mem-
bers
• In the event that the CH becomes vacant, the substitute CH takes over the responsibility and manages the
energy, distance, and stability pair factor of the cluster members
• The efficient path is determined by considering the node’s maximum availability of energy and with a
lesser hop count from the base station and the CH as well as among the CH and nodes of group

3.1.2 The routing protocol

By combining the route discovery mechanism of AODV and the mobility-based path
selection of MBOMRP, the proposed algorithm can provide multiple paths with improved
stability and reliability. AODV is a flat routing protocol and it does not need any central
administrative system to handle the routing process. It is also loop free, self-starting. It
supports large number of mobile nodes , it has lower setup delay for connections and detec-
tion of latest route to the destination. The MBOMRP algorithm focuses on selecting the
traffic free available route and reducung traffic. Here route selection is made by considering
the lowest remaining energy of the nodes in paths.

A detailed explanation of the combined AODV and MBOMRP algorithm for multiple path discovery:

1. Initialization:
• This step involves initializing the parameters and data structures required for both AODV and MBOMRP
• Parameters include routing tables, sequence numbers, timers, and mobility information
• Data structures store routing information, mobility data, and packet queues
2. Route Discovery (AODV):
• The route discovery process begins when the source node broadcasts a Route Request (RREQ) packet to
its associate nodes
• The RREQ packet includes essential details such as the location of source nodes, the location of the
destination address, and a unique identifier
• Upon receiving the RREQ packet, each node verifies its routing table to identify if it already framed the
path to reachthe destination or relevant mobility information
• If the node doesn’t possess a route or has better mobility information, it rebroadcasts the RREQ packet to
its neighboring nodes
• This process repeats iteratively until a route to the destination is established
3. Mobility-Based Path Selection (MBOMRP):
• Nodes continuously monitor their own mobility and that of their neighbors
• As nodes receive RREQ packets, they update their mobility information for both the source and destina-
tion nodes
• This information includes speed, direction, stability metrics, or any other relevant parameters
• Based on the mobility information, nodes evaluate the stability and reliability of potential routes
• Paths with higher stability and reliability are selected for multiple path establishment
• The selection criteria can include factors such as lower speed, higher stability, and lower probability of
node disconnections
4. Route Maintenance (AODV):

13
 Multimedia Tools and Applications

A detailed explanation of the combined AODV and MBOMRP algorithm for multiple path discovery:
• AODV handles the route maintenance flow to ensure the availability of established paths
• If a node identifies a failure in route, such as link breakage or node movement, it sends a Route Error
(RERR) packet resends to the source node
• Upon receiving the RERR packet, the source node recommends a new route discovery procedure to find
an alternative path
• The route maintenance mechanism of AODV ensures that paths remain up-to-date and usable even in the
presence of network changes
5. Packet Forwarding:
• When data packets are generated at the source node, they are transmitted through the selected multiple
paths
• Each intermediate node makes forwarding decisions based on its routing table, sequence numbers, and
the selected paths
• The source and destination nodes handle the packet reassembly and acknowledgment processes to ensure
ordered and reliable packet delivery

A reactive routing protocol is adopted to establish a route in response to the source

node’s request. Key control messages used in this process are the Route Request (RREQ)
and Route Reply (RREP). These messages contain important properties, such as the
sequence number ofthe destination, which helps assess the uniqueness of the route [25]. To
reduce computation delay during data communication, cluster-oriented routing is imple-
mented. To illustrate the routing scheme, we present a working simulation setup with 80
nodes using NS2 simulation environment as depicted in Fig. 2, showcasing the network
deployment. Here mesh topology is considered for implementation purpose,

Fig. 2  Deployment of network with mechanism of cluster head selection

13
Multimedia Tools and Applications

Table 1  Routing loop

Source Destination CHs

Member belonging to CH5 (Node Member belonging to CH8 (Node Nodes 48, 60, 9,8,20,62,76, and 50
11) 28)

Fig. 3  Coverage calculation of CH for route discovery

In this deployment, a network consisting of 80 nodes is clustered into 8 groups

depending on their capability of coverage. Each group is assigned a cluster head (CH),
represented by red nodes in Figure 2. The CH nodes are numbered 48, 60, 9, 8, 20, 62,
76, and 50.
In this particular instance, Source node 11 and destination node 28. The node represent-
ing the Source forms under CH5, while CH8 represents the destination. A mechanism for
pathdiscovery is required to facilitate the transmission of data packets from the source to
the destination node.
To achieve this, a valid and efficient routing protocol is utilized to identify paths from
the source to the destination node via the respective on-demand CHs. The "Request
(RREQ) - Reply (RREP)" routing mechanism is employed, minimizing transmission delays
by avoiding routing loops, as described in Table 1.
In this scenario, the source node is a member node of CH5 and is responsible for trans-
mitting packets of data to the Source node (CH5). Subsequently, the Source node initiates

13
 Multimedia Tools and Applications

Table 2  Selected CH and their Selected CH Coverage CHs

coverage
CH 1 CH 3 and CH 7
CH 2 CH 4 and CH 6
CH 3 CH 8
CH 4 CH 2 and CH 6
CH 5 (includes Source) CH 4 and CH 7
CH 6 CH 2 and CH 4
CH 7 CH 2 and CH 3
CH 8 (includes Destination) CH 2 and CH 3

the distribution of an RREQ message packet to neighboring CHs within its limit of cov-
erage. The coverage of each CH, as depicted in Figure 3, is calculated and presented in
Table 2.
In this scenario, CH5 begins transmitting a RREQ info packet to its neighboring CHs,
specifically CH4 and CH7. The broadcast structure includes crucial information such as
the source CH, the number of hops, and the destination Cluster Head. With the present
simulation example, this structure would be represented as (CH5, hops, CH8).
CH4 and CH7 receive this packet and respond back to CH5 with a RREP message,
which contains the information [CH5, Hop-Count, CH4/CH7]. The RREQ message

Fig. 4  Coverage calculation of CH for route discovery

13
Multimedia Tools and Applications

serves as a "hello message" that carries details about the source CH, destination CH,
and their respective hop counts. Upon receiving responses from the neighboring nodes,
the RREP message is generated, indicating the ID’s of the nodes.
These stepscontinue iteratively until the destination CH is discovered, and the com-
plete mechanism of route discovery is illustrated in Figure 4.
The cluster heads of Node 11 initiate the transmission of packets by issuing a rout-
ing request, RREQ, for all the neighboring Cluster Heads nodes for identifying the effi-
cient path. The path selection depends on the computation of overall consumption of
power and delay in transmission. Upon accepting the RREQ, the intermediate Cluster
Headresponds by generating a reverse route/path to the Cluster Head of Node 11, using
RREP packets.This procedure repeats as the RREQ message is transmitted to neighbor-
ing CHs still it reaches the CHs of Node 28 destination. Simultaneously, the destination
Cluster Head of Node 28 creates an RREP message and passes it back to the Cluster
Head of the source Node 11, following the established reverse path. Upon receiving
message RREP packets via different formed routes, the CH of the source Node 11
selects a forward route based on criteria such as lesser consumption of power and delay
in transmission. In this proposed routing mechanism, a total of 4 different paths are
formed between the source nodes to the destination node, utilizing the Cluster Heads.
Referring to Figure 4, a routing table is provided in Table 3, showcasing the available
routes for data transmission.

3.2 Phase 2— Secured fragmented message transmission using multiple

Cryptographic techniques

3.2.1 Encryption Algorithm

AES AES algorithm is used for security and speed improvement. It employs around 10
repetitions of encryption, each including 4 steps: sub bytes, shift rows, mix columns, and
add round key. The encryption process transforms information into ciphertext, making it
unintelligible [26, 27]. The tenth round is similar to the previous rounds, except it excludes
the mix columns step.

Table 3  Routing table Possible routes Member nodes

Route 1 [Node 11, CH 5, CH 7, CH 3, CH 8, Node 28]

Route 2 [Node 11, CH 5, CH 7, CH 2, CH 8, Node 28]
Route 3 [Node 11, CH 5, CH 4, CH 2, CH 8, Node 28]
Route 4 [Node 11, CH 5, CH 4, CH 6, CH 2, CH 8, Node 28]

13
 Multimedia Tools and Applications

Algorithm: AES
Input: Plaintext message M, Key K
Output: Encrypted message C
Procedure AES_Encryption(plaintext, key):
1. Set the number of rounds (Nr) based on the key size:
- For a 128-bit key, Nr = 10
- For a 192-bit key, Nr = 12
- For a 256-bit key, Nr = 14
2. Initialize the round keys (w) from the original key (K).
3. Perform the initial round of operations:
- AddRoundKey(plaintext, w[0:3])
4. Perform Nr - 1 rounds of operations:
- SubBytes(plaintext)
- ShiftRows(plaintext)
- MixColumns(plaintext)
- AddRoundKey(plaintext, w[i*4:(i+1)*4])
5. Perform the final round of operations:
- SubBytes(plaintext)
- ShiftRows(plaintext)
- AddRoundKey(plaintext, w[Nr*4:(Nr+1)*4])
6. Return the encrypted ciphertext.

RSA The term RSA is derived from the initials of the three researchers who designed it.
Public key encryption is a fundamental aspect of RSA, ensuring secure data transmission.
In RSA, a public key is utilized for encrypting data, during a corresponding private key is
required for the process of decryption. The RSA security parameter depends on the factori-
zation of two large prime numbers, a key component in the algorithm [28, 29].

13
Multimedia Tools and Applications

Algorithm: RSA

Input: Plaintext message M, Public key (e, n)

Output: Encrypted message C
Procedure RSA_Encryption(fragment, e, n):
1. Convert the fragment into a numeric representation (e.g., using ASCII or UTF-8 encoding).
2. Ensure the numeric representation is smaller than n (the modulus).
3. Apply modular exponentiation to the fragment using the public key (e) and modulus (n).
4. Obtain the encrypted fragment as the result of the modular exponentiation.

SHA‑256 The SHA-256 (Secure Hash Algorithm) is a cryptographic-based hash function

havinga digest length of 256 bits. This hash function is performed without the key for gen-
erating a Manipulation Detection Code (MDC). The algorithm processes the message in
blocks of size 512 bits (16X32) and requires 64 rounds for each block [30, 31].

Algorithm : SHA-256

Input: Fragmented message fragments[]

Output: Hash value H

Procedure SHA256(fragment):
1. Initialize the SHA-256 constants (K) and initial hash values (H0).
2. Prepare the message schedule array (W) with the fragment data.
3. Initialize working variables (a, b, c, d, e, f, g, h) with the initial hash values (H0).
4. Perform the SHA-256 rounds:
a. Iterate through the message schedule array:
i. Compute the intermediate values based on the current working variables.
ii. Update the working variables based on the intermediate values.
5. Calculate the final hash value H by concatenating the working variables.

Blowfish Blowfish is a high-speed block cipher, except when it comes to key changes.
Whenever a new key is used, pre-processing has to be done using 4 kilobytes of words
which impacts to same as performing encryption. Developed by Bruce Schneier, Blowfish
is a symmetric block cipher.Blowfish employs a variable key size, ranging from 32 to 448
bits. The key is utilized for generating 18 number of 32-bit subkeys and 4 number of 8 × 32
S-boxes, which together hold a sum of 1024 32-bit entries [32].

13
 Multimedia Tools and Applications

Algorithm: Blowfish

Input: Fragmented message fragments[], Encryption key K

Output: Encrypted message ciphertext[]

Procedure Blowfish_Encryption(fragment, K):

1. Initialize the Blowfish S-boxes and the P-array using the encryption key K.
2. Divide the fragment into data blocks (e.g., 64 bits per block).
3. For each data block:
a. Perform the Blowfish encryption rounds:
i. XOR the data block with the initial P-array value.
ii. Iterate through the Blowfish rounds:
- Apply the Feistel function using the S-boxes and P-array.
- Update the data block based on the Feistel function result.
b. XOR the final P-array value with the data block.
4. Obtain the encrypted data block as the result.

3.2.2 To transmit fragmented messages through multiple paths

By randomly selecting the encryption algorithm each fragmented message is encrypted and
transmitted to the multiple route obtained from phase-1. Its block diagram is shown in the
Fig. 5 and its detailed steps are explained below:

1. Fragmentation: Split the original message into smaller fragments. Determine the size
of each fragment based on your requirements or constraints.

To split the original message into smaller fragments during the fragmentation process,
you can follow these steps:

Assign Fragments Encrypt the

Fragment the Efficient Path
to paths Fragmented
Message Selection
(Round-Robin) Messages

Decryption and Receive and Transmit the

Reassembly Reassembly Encrypted
Fragments

Fig. 5  Proposed block-diagram

13
Multimedia Tools and Applications

Fragmentation of message

• Determine the Fragment Size: a fragment size of 1000 bytes

• Calculate the Number of Fragments: Divide the total size of the original message by the fragment size
to determine the number of fragments needed. Keep in mind that the last fragment may be smaller if the
original message size is not evenly divisible by the fragment size. For example, if the original message
size is 4500 bytes and the fragment size is 1000 bytes, you would need 5 fragments
• Split the Message into Fragments: Start from the beginning of the original message and extract consecu-
tive portions of data equal to the fragment size. Copy each portion into a separate fragment. Repeat this
process until the entire message has been split into fragments
• Assign Fragment Identification
• Include Fragment Headers
• Store or Transmit Fragments

2. Multiple path discovery using phase-1.

3. Assign fragments to paths: Distribute the fragments among the selected paths using
Round-Robin Algorithm

The round-robin algorithm [33] is a simple and widely used technique for evenly dis-
tributing tasks or items among a set of resources in a cyclic manner. The round-Robin algo-
rithm is explained below:

Algorithm Round-Robin
1. Input:
- `resources`: List of available resources
- `fragments`: List of fragments to be assigned
2. Initialize:
- `num_resources` = length of `resources`
- `current_resource_index` = 0
3. For each fragment in `fragments`:
a. Assign the current fragment to the resource at `current_resource_index` in `resources`.
b. Increment `current_resource_index` by 1.
c. If `current_resource_index` exceeds `num_resources - 1`, reset it to 0 to wrap around.
4. Output:
- The assignments of fragments to resources, indicating which fragment is assigned to which resource
(e.g., path).

This algorithm ensures that each fragment is assigned to the next available resource
in a cyclic manner. It starts by assigning the first fragment to the resource at index 0,
then moves to the next resource for the next fragment, and so on. Once it reaches the
last resource, it wraps around to the first resource, continuing the round-robin assign-
ment pattern.

13
 Multimedia Tools and Applications

4. Encryption: Apply a cryptographic technique to each fragment. The advantages of

hybrid cryptographic technique is that a plain text can be any bit string of any length. It
is also easier to construct secure hybrid encryption schemes than the direct construction
of public-key encryption schemes. We are randomly selecting a cryptographic technique
for each fragment. Procedure is explained below:

Randomly selecting a cryptographic technique from the available options

• Define a list or an array of the available cryptographic techniques that you want to choose from. avail-
able_techniques = [’AES’, ’RSA’, ’SHA-256’, ’5’]
• Use a random number generator to select an index from the list of available techniques. The method for
generating random numbers will depend on the programming language you are using. Here’s an example
using Python’s `random` module: import random selected_technique = random.choice(available_tech-
niques) ```
• The `selected_technique` variable will now contain a randomly chosen cryptographic technique from the
`available_techniques` list

5. Path transmission: Transmit the encrypted fragments through their assigned paths
simultaneously or sequentially. Ensure that the network infrastructure supports multiple
paths and can handle the transmission requirements.
6. Receiving and reassembly: At the receiving end, the fragments are received through
multiple paths. Collect all the fragments, regardless of the path they arrived on.
7. Decryption and reassembly: Decrypt each fragment using the appropriate decryption
algorithm based on the cryptographic technique applied during encryption. Once all the
fragments are decrypted, reassemble them to reconstruct the original message.

4 Results and discussion

This section focuses on evaluating the performance of both the existing mechanism
and the proposed mechanism using several factors, including Control Packet Overhead,
Packet Delivery Ratio (PDR), average end-to-end delay, and throughput. The evaluation
is carried out in two phases. Parameters used in the simulation are tabulated in Table 4.
In the first phase, the discovery of multiple routes is considered. The performance of
this aspect is assessed and compared with existing approaches.

Table 4  Parameters used in the Parameters Value

simulation
Network dimension 1000 × 1000 m2
The initial energy of nodes 100 mJ
Nodes 50 10 300
Packet size 2000 Bytes/P
Routing Protocol AODV + MBOMRP

13
Multimedia Tools and Applications

Table 5  The multipath route discovery

Route Average throughput Average PDR Average consumed Delay in seconds
energy in mJ

[3, 5, 7, 8, 12, 25] 4172.2104 0.4821042 2.03476877 2.03981893

[2, 5, 7, 8, 12, 25] 3055.72407 0.3557034 1.74527255 2.84326352
[2, 4, 5, 8, 12, 25] 3182.23878 0.3821787 1.98308396 3.03353644
[2, 4–6, 8, 12, 25] 3216.56652 0.3162526 1.74699070 2.62577356

In the second phase, the evaluation focuses on secure message transmission through
multiple paths. The proposed mechanism is compared to existing approaches to analyse
and compare factors such as a throughput, PDR, energy consumed, and delay.
The network allows for a total of 4 possible routes, each based on different QoS
parameters. These routes are treated as candidates and are presented in the Table 5
below.Possible routes are [2, 2–5, 5, 5, 7, 7, 8, 8, 8, 12, 12, 12, 25, 25, 25] and [2, 4–6,
8, 12, 25] here nodes 11 and 28 are source and destination, respectively.
The evaluation of the proposed work focuses on two parameters: throughput and delay.
A comparison is made between the proposed methodology and other existing techniques.
As observed from the table, it is observedby increasing the number of nodes, all algo-
rithms demonstrate an increase in throughput. This is expected as the amount of carriers for
data packets also increases with more network nodes. However, the proposed methodology
stands out with a significantly higher throughput compared to other existing techniques.The
notable increase in throughput can be attributed to the architecture of the proposed meth-
odology, which incorporates a clustering technique to enhance the route selection policy.

Table 6  Throughput comparison by varying number of nodes

Numberofnodes Throughputproposed ThroughputAdiletal [34] ThroughputJana- Throughput-
kiramanetal [35] Tongetal [36]

50 3245 3101.4322 3089.663 3081.7748

100 4080.65244 3812.00244 3674.8769 3401.97335
150 5876.14379 5532.5628 4424.02818 4425.70629
200 7853.86691 7626.83521 5709.70214 5354.57379
250 10,518.6957 8754.0543 7702.78221 6962.35297
300 14653.2772 12548.9491 9005.73384 7823.11966

Table 7  Delay comparison by varying number of nodes

“Node “Delayproposedin “DelayAdiletal.in “DelayJankiramanetal.in “DelayTongetal.”

count” ms” ms”[34] ms”[35] [36]

50 32.2873679 36.932468 39.1538966 39.2373847
70 33.1606471 37.7698548 38.351566 39.8051611
90 30.9266673 34.9384715 37.3427305 39.3521663
120 27.3980742 27.9429802 36.465805 36.9402624
150 25.4706602 26.8721583 33.5027698 32.9370081
170 23.0747941 24.2530223 32.1434133 31.9786625

13
 Multimedia Tools and Applications

Throughput
16000
14000
12000
Throughput

10000
8000
6000
4000
2000
0
50 100 150 200 250 300
No.of Nodes

Throughput proposed Throughput Adil et al. [22]

Throughput Janakiramanet al. [32] Throughput Tonget al. [33]

Fig. 6  Throughput comparison with existing methods

Delay
50
40
30
Delay

20
10
0
50 70 90 120 150 170

No.of Nodes
“Delay proposed in ms” “Delay Adil et al. in ms” [22]
“Delay Jankiraman et al. in ms” [32] “Delay Tong et al.” [33]

Fig. 7  Delay comparison with existing methods

This allows network nodes to have more lifespan in the network, resulting in a higher aver-
age throughput for the proposed simulation setup (Tables 6 and 7) (Figures 6 and 7).
By comparing the values of throughput and delay with varying nodes with the existing
approaches, our proposed methodology is observed to be more efficient.

13
Multimedia Tools and Applications

60
50
End-to-End Delay

40
30
20
10
0
10 15 20 25
Bandwidth

Proposed Methodology Fuzzy NB KmeansNB Naïve Bayes

Fig. 8  Bandwidth vs end-to-end delay

Table 8  Outcome of bandwidth Bandwidth Proposed Fuzzy NB KmeansNB Naïve Bayes

vs end-to-end delay (Mbps) methodol-
ogy

10 49.1023 47.73 45.75 43.141

15 47.234 44.75 44.74 39.5134
20 34.194 31.69 31.98 24.4621
25 24.732 22.15 19.14 15.754

The performance validation parameters, including throughput, end-to-end delay, deliv-

ery ratio, average path lifetime, and routing control overhead, are utilized to assess the
effectiveness of the algorithms. By comparing the outcomes of the proposed methodology
with existing algorithmssuch as: K-means NB [37], Naive Bayes [38], and NBTrust [39].,
we can gauge its performance and determine its superiority.

1. Bandwidth vs. end-to-end delay

Figure 8 depicts the bandwidth and end-to-end delay comparison with the proposed
methodology with other existing models. As seen in Table 8, our proposed methodology
has minimized the loss of data when compared to other existing models.

13
 Multimedia Tools and Applications

120
100
Delivery Rao

80
60
40
20
0
5 25 45 65 85
Throughput

Proposed Methodology Fuzzy NB KmeansNB Naïve Bayes

Fig. 9  Throughput vs. delivery ratio

Table 9  Outcome of throughput Through- Proposed Fuzzy NB KmeansNB Naïve Bayes

vs. delivery ratio put (Mbps) methodol-
ogy

5 42.76 47.14 54.46 70.52

25 52.77 58.95 66.85 79.64
45 66.79 72.49 80.18 86.92
65 73.85 80.77 85.18 90.43
85 84.76 89.46 94.76 97.75

5
ROUTING OVERHEAD

4
3
2
1
0
2 3 4 5 6 8 12
AVERAGE PATH LIFETIME

Proposed Methodology Fuzzy NB

KmeansNB Naïve Bayes

Fig. 10  Average path lifetime vs. routing protocol overhead

13
Multimedia Tools and Applications

Table 10  Results of average Lifetime(ms) Routing overhead (104 packets)

path lifetime vs. routing protocol
overhead Proposed Fuzzy NB KmeansNB Naïve Bayes
methodol-
ogy

2 3.9167 3.6968 2.7562 2.4465

3 3.1139 2.8512 2.506 2.206
4 2.7314 2.6235 2.476 2.015
5 2.4623 2.3267 2.156 1.8076
6 1.9034 1.8523 1.7421 1.5364
8 1.448 1.363 1.287 1.01253
12 1.265 1.186 1.1254 0.9478

2. Throughput vs. delivery ratio

Figure 9 explainsns the Throughput and Delivery Ratio comparison with the proposed
methodology with other existing models. As seen in Table 9, our proposed methodology
results with a rise in the delivery ratio for higher throughput when compared to other exist-
ing models.

3. Lifetime with routing control overhead

Figure 10 shows the Average path lifetime and Routing protocol overheadcomparison
with the proposed methodology with other existing models. Table 10 displays the proposed
methodology achieves lesser routing overhead comparing with the existing methods.

45
40
35
30
25
Delay

20
15
10
5
0
5 25 45 65 85
Throughput

Proposed Methodology Fuzzy NB KmeansNB Naïve Bayes

Fig. 11  Throughput vs. delay

13
 Multimedia Tools and Applications

Table 11  Results of throughput Throughput Delay (ms)

vs. delay (Mbps)
Proposed Fuzzy NB KmeansNB Naïve Bayes
Methodol-
ogy

5 18.0289 16.891 13.753 9.629

25 20.7113 18.568 14.652 12.225
45 27.5823 24.652 21.754 16.7367
65 30.5823 28.9534 26.756 24.0473
85 39.0034 36.8232 32.7836 30.9577

4. Throughput vs. delay

Figure 11 express the Throughput and Delaycomparison with the proposed methodol-
ogy against other existing models. As shown in Table 11 our proposed methodology deliv-
ers a decline in loss of packet loss, minimized delay, and rise in throughput.

5 Conclusion

This research introduces an efficient mechanism that ensures secure message transmission
through multiple paths, leading to increased throughput in MANETs. The main focus of
this work is on message transmission confidentiality and authentication. To achieve this,
a combination of AODV and MBOMRP routing protocols is employed, utilizing cluster
heads for path selection. This approach helps prevent packet loss and reduce delay time
during transmission. Additionally, an encryption mechanism is implemented, where the
original fragmented message is encrypted using randomly chosen encryption algorithms.
Each fragment is encrypted with a different algorithm, enhancing the overall security of the
packet. While simulating, different metrics are utilized to evaluate the performance of the
proposed methodology. A comparison is made between the proposed approach and existing
techniques to demonstrate its effectiveness. The practicality of implementing this can be
used in SMS / Multimedia messages on android. It can also be used for cloud computing
security. The outcome indicates that the proposed framework is more efficient when com-
pared with other methods. In the future, we aim to address new challenges of MANETs
using the proposed algorithm, further improving the system’s resilience.

Data availability The data required to reproduce the above findings cannot be shared at this time as the data
also forms part of an ongoing study.

Declarations
Conflict of interest The authors have no conflicts of interest to declare that are relevant to the content of this
article.

13
Multimedia Tools and Applications

References
1. Sakhaee E, Jamalipour A (2008) Stable clustering and communications in pseudolinear highly mobile
ad hoc networks. IEEE Trans Veh Technol 56(8):3769-3777 2
2. Silva L et al (2021) Computing paradigms in emerging vehicular environments: a review. IEEE/CAA J
Autom Sin 8(3):491–511. https://​doi.​org/​10.​1109/​JAS.​2021.​10038​62
3. Huiyao A, Xicheng L, Wei P (2004) A cluster-based multipath routing for MANET. National Univer-
sity of Defense Technology, Changsha, China, Computer School, pp 405–413
4. Huang H, Huang C, Ma D (2020) A method for deploying the minimal number of UAV base stations
in cellular networks. IEEE/CAA J Autom Sin 7(2):559–567. https://​doi.​org/​10.​1109/​JAS.​2019.​19118​
13
5. Chatterjee M, Das SK, Turgut D (2009) A weight based distributed clustering algorithm for mobile ad
hoc networks. HiPC 2000, LNCS 1970, pp. 511–521, 2000.
6. Asif W, Qaisar S (2011) “Energy and path aware clustering algorithm for mobile ad hoc networks.
ICCSA, Part IV, LNCS 6785:133–147
7. Basu P, Naidu W (2012) A mobility based metric for clustering in mobile ad hoc networks," in Interna-
tional conference on Distributed computing systems workshop. pp 413–418
8. Zhong J, Huang Z, Feng L, Du W, Li Y (2020) A hyper-heuristic framework for lifetime maximiza-
tion in wireless sensor networks with a mobile sink. IEEE/CAA J Autom Sin 7(1):223–236. https://​
doi.​org/​10.​1109/​JAS.​2019.​19118​46
9. Shukla N (2013) Mobile Ad-Hoc Network (MANET): Security Issues Regarding Attacks. In Inter-
national Journal of Computer Applications (0975–8887), National Conference on Recent Trends in
Engineering and Management “NCRTEM-2013
10. Sudha K, Ranjith JP, Ganapathy S, Sasidharan MSR (2015) Secure transmission over remote group:
a new key management prototype. Int J Comput Sci Netw Secur (IJCSNS) 15(1):102–105. http://​
paper.​ijcsns.​org/​07_​book/​201501/​20150​118.​pdf
11. Sung S, Seo Y, Shin Y (2006) Hierarchical clustering algorithm based on mobility in mobile ad hoc
networks. ICCSA, LNCS 3982:954–963. https://​doi.​org/​10.​1007/​11751​595_​100
12. Morris R, Jannotti J, Kaashoek F, Li J, Decouto D (2010) CarNet: A Scalable Ad Hoc Wireless
Network System. Proc. of 9th ACM SIGOPS European Workshop, Kolding, Denmark
13. Grossglauser M, Tse D (2001) Mobility increases the capacity of ad-hoc wireless networks. Proc.
of INFOCOM’01. pp 1360–1369
14. Steenstrup M (2001) Cluster-Based Networks. Chapter 4, Ad Hoc Networking, edited by C. E. Per-
kins, Addison-Wesley
15. Neethu V, Singh AK (2015) Mobility aware loose clustering for mobile ad hoc network. Procedia
Comput Sci 54:57–64
16. Loutf A (2014) An energy aware algorithm for OLSR clustering. Ann Telecommun 69:201–207
17. Basurra SS, De Vos M, Padget J, Ji Y, Lewis T, Armour S (2015) Energy efcientzonebased routing
protocol for MANETs. Ad Hoc Netw 25:16–37
18. Sp RM, Bhattacharya S, Maddikunta PKR, Somayaji SRK, Lakshmanna K, Kaluri R, Hussien A,
Gadekallu TR (2020) Load balancing of energy cloud using wind driven and frefy algorithms in
internet of everything. J Parallel Distrib Comput 142:16–26. https://​doi.​org/​10.​1016/j.​jpdc.​2020.0.​
2.​010
19. Kulkarni SB, Yuvaraju B (2016) Challenges and issues of cluster based security in MANET. IOSR
J Comput Eng (IOSR-JCE) 18:01–05
20. Kaur I, Rao A (2017) A framework to improve the network security with less mobility in MANET.
Int J Comput Appl 167(10):21–24. https://​doi.​org/​10.​5120/​ijca2​01791​4390
21. Jhaveri RH, Patel NM (2015) A sequence number based bait detection scheme to thwart grayhole
attack in mobile ad hoc networks. Wireless Netw 21(8):2781–2798
22. Singh M, Gagangeet S (2015) A secure and efficient cluster head selection algorithm for MANET. J
NetwCommunEmergTechnol (JNCET);2. www.​jncet.​org
23. Chakeres ID, Belding-Royer EM (2004) AODV routing protocol implementation design. ICDCSW
’04: Proceedings of the 24th International Conference on Distributed Computing Systems Work-
shops - W7: EC (ICDCSW’04) 7:698–703
24. Panke T Clustering based certifcate revocation scheme for malicious nodes in MANET. IJSRP,
Volume 3, Issue 5, May 2013 Edition [ISSN 2250–3153]
25. Dhenakaran DS, Parvathavarthini A (2013) An overview of routing protocols in mobile ad-hoc net-
work. Int J Adv Res Comput Sci Softw Eng 3:251–259
26. Taneja S (2010) A survey of routing protocols in mobile ad hoc networks. Int J Innov Manag Tech-
nol 1(3):279

13
 Multimedia Tools and Applications

27. ALGhafran L, Yusof ZBM (2015) Anefcient load balancing clustering algorithm for mobile ad hoc
networks. Int J Comput Commun Instrum Eng (UCCIE) 2(2):215–9
28. Chatterjee M, Das SK, Turgut D (2002) WCA: a weighted clustering algorithm for mobile ad hoc
networks. Clust Comput 5:193–204
29. Grover D, Saini S (2015) A survey on unicast routing protocols in mobile ad-hoc networks. Int J
Adhoc Ubiquitous Comput 5:697–702
30. Djemili F, Tolba C (2013) A mixed approach load balancing and efcient energy for multi-path rout-
ing in mobile ad hoc networks. In: SENSORS, IEEE, Baltimore, MD, USA, IEEE, pp 1–4
31. Spyropoulos T, Psounis K, Raghavendra C (2010) Performance analysis of mobility-assisted rout-
ing. In Proc. MobiHoc ’06: 7th ACM International Symposium on Mobile Ad Hoc Networking and
Computing. pp 49–60
32. Xinming Z, Dong S, Fengfu Z (2010) Integrated routing metric for mobile ad hoc networks. J
Softw 20(11):3077–3085
33. He G (2002) Destination-Sequenced Distance Vector (DSDV) Protocol. Hels Univ Technol Finl Mag 135:1–9
34. Kavitha P, Mukesh R (2018) Detection of impersonation attack in MANET using polynomial reduc-
tion algorithm. Int J Netw Secur 20(2):381–389. https://​doi.​org/​10.​6633/​IJNS.​201803.​20(2).​19
35. Camp T, Boleng J, Davies V (2002) A survey of mobility models for ad hoc network research. Wirel
Commun Mob Comput 2(5):483–502
36. Zhang X (2016) A street-centric opportunistic routing protocol based on link correlation for urban
manets. IEEE Trans Mob Comput 1:1–1
37. Subba B, Biswas S, Karmakar S (2016) Intrusion detection in mobile ad hoc networks: Bayesian game
formulation. Eng Sci Technol 19(2):782–799
38. Marchang N, Datta R, Das SK (2017) A novel approach for efcient usage of intrusion detection system
in mobile ad hoc networks. IEEE Trans Veh Technol 66(2):1684–1695
39. Singh O, Singh J, Singh R (2018) Multi-level trust based intelligence intrusion detection system to
detect the malicious nodes using elliptic curve cryptography in MANET. Cluster Comput 21:51–63

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a
publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript
version of this article is solely governed by the terms of such publishing agreement and applicable law.

Authors and Affiliations

Jayantkumar A. Rathod1 · Manjunath Kotari2

* Jayantkumar A. Rathod
[email protected]; [email protected]
Manjunath Kotari
[email protected]
1
Alva’s Institute of Engineering and Technology, Moodubidire 574225 Mangalore, India
2
Visvesvaraya Technological University, Belagavi 590018, India

13