Telecommunication Systems (2024) 86:769–783

https://doi.org/10.1007/s11235-024-01170-1

A novel algorithm for the development of a multipath protocol for

routing and energy efficient in IoT with varying density
Radwan S. Abujassar1

Accepted: 17 May 2024 / Published online: 1 June 2024

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

Abstract
Data transmission from sensor nodes is the biggest problem for IoT networks. Overusing communication power shortens
node lifespans. Thus, network issues including QoS, security, network heterogeneity, congestion avoidance, reliable routing,
and energy savings must be addressed. Routing protocols are essential for delivering data between organizations. Information
gathering and consolidation require data aggregation to minimize traffic congestion, operating expenses, energy usage, and
network lifespan. IoT data aggregation makes route planning dependable, energy-efficient, and difficult. Disjoint & Scalable
Multipath Routing (D&SMR) is a new routing system developed using NS2 simulation. The method estimates delivery
success using decision trees and neural networks. We evaluate characteristics such as (D&SMR) routing scheme predictability,
node popularity, power consumption, speed, and location while training the model. Simulation results show that (D&SMR)
outperforms a reliable routing system in terms of delivery success, lost messages, overhead, and hop count. The proposed
hybrid routing method involves cluster construction and intra- and inter-cluster routing. The study found that (D&SMR) beats
previous research in network resilience, packet transmission efficiency, end-to-end latency, and energy usage.

Keywords Internet of Things (IoT) · Energy efficiency · Network life time · Optimal path selection · Packet inquiry (PINQ)

1 Introduction wireless personal area networks (WPAN), wirelessfidelity

networks (WIFI), and numerous more, have been developed.
The Internet of Things (IoT) refers to a network that links To clarify, the IoT currently comprises a vast and diverse
a significant quantity of items. An object can be classified network [3–6].
based on several characteristics. Firstly, it can be categorized Within the IoT ecosystem, there exist several method-
as either virtual or physical. Second, an object is autonomous ologies and protocols that provide effective communication
in terms of its decision-making and functioning. Third, it and connectivity among items. In this context, one promi-
is capable of communicating with other devices. Fourth, nent strategy is the use of ad hoc mode [7, 8]. An ad hoc
an object is interactive and interoperable because it works network refers to a transient infrastructureless peer-to-peer
with a variety of devices. Lastly, an object is flexible, as (p2p) network framework whereby each individual device
it can interact with any other object, regardless of location undertakes the tasks of data collection, processing, storage,
or service [1, 2]. However, from a technological standpoint, and transmission [9]. In the context of wireless communica-
it is important to note that the IoT is not only a singular tion, it is observed that devices establish connections with
technology but rather a convergence of several networking nearby devices through direct wireless transmissions. How-
components. Various technologies, including wireless sen- ever, in cases where the distance between nodes exceeds
sor networks (WSN), radio frequency identification (RFID), the range of radio transmission, multi-hop transmissions are
smart ad hoc networks, vehicular ad hoc networks (VANET), employed to facilitate communication between these nodes.
In multihop communication, an entity that is transferring
data uses additional entities as relays to expand the range
B Radwan S. Abujassar
of its transmission. Hence, a device depends on interme-
[email protected]
diary entities to construct pathways and transmit packets
1 Faculty Information Technology and Computing, Arab Open to their intended destinations. Ad hoc networks have sev-
University-Kuwait, Alardiya Industrial, 830, 92400 eral advantages, including efficient and rapid deployment,
Al-Ardiya, Kuwait

123
770 R. S. Abujassar

Fig. 1 Mobile Ad-hoc Network

cost-effectiveness, adaptability, resilience, and support for Fig. 2 Network Topologies with NCH
mobility. Ad hoc networking is widely prevalent in many
IoT applications, including but not limited to disaster res- IoT devices and the recent advancements in interoperable
cue, cooperative intelligent transportation systems (CITS), protocols and middleware technologies, the dimensions of
e-health, logistics, military scenarios, and environmental MANETs are continuously expanding. Consequently, these
sensing, among others. The user’s text consists of three networks are becoming intricate compositions of diverse
numerical references: [10, 11]. Furthermore, the ad hoc mode devices, facilitating the transmission of critical data. In addi-
is a fundamental component of 5 G deployment modes, play- tion, the multi-hop transmission paradigm operates on the
ing a crucial role in increasing the network’s coverage area, assumption that devices exhibit complete cooperation and
maintaining service resilience, and improving the overall faithfully fulfill their roles in routing and forwarding. How-
user experience [5, 12, 13]. ever, there are several motives, such as the goal of conserving
Figure 1 illustrates the visual representation. Ad hoc energy, that might lead to nodes refraining from forward-
networks are self-governing networks that function indepen- ing and routing packets. Certain entities with self-centered
dently, or stub networks that establish connections with an motives may exhibit a reluctance to engage in collaborative
existing fixed network. It is not advisable to solely depend efforts. Conversely, in the case of individuals with malevo-
on pre-existing infrastructure. Therefore, the absence of an lent intent, their primary goal is to disrupt the network and
access point is evidenced by the decentralized nature of the manipulate packet contents. These behaviors have a signifi-
network, where each individual node functions as a router cant influence on the availability of connections in MANETs,
and assumes the role of forwarding packets on behalf of other leading to reduced connectivity and perhaps causing various
nodes. packet transmission issues. In the context of multi-hop trans-
Figure 1 illustrates an exemplar of a mobile ad hoc net- missions, it is important to note that each forwarding node
work (MANET) in the context of the IoT. On the other has the potential to engage in a Man in the Middle (MITM)
hand, Fig. 2 portrays a scenario in which a satellite dissem- assault. Furthermore, as a result of the inherent spontane-
inates data simultaneously to several destinations inside a ity of the IoT and MANETs, devices possess the freedom to
huge IoT network. The satellite facilitates the transmission join or depart from the network at any given moment. Conse-
and reception of communications, mostly directed towards a quently, individuals who are newly introduced to the network
nearby smart device located on an aircraft or vessel. These lack pertinent information, rendering them susceptible to
messages serve the purpose of disseminating auditory and manipulation by malicious actors. Additionally, the process
visual warnings. In the context of MANETs and the IoT, of joining and departing the network contributes to unstable
it is common for devices to often depart from and join the connectivity, leading to heightened traffic and subsequent
network or exhibit mobility, resulting in a dynamic alteration congestion. Furthermore, incidents necessitate updating the
of the network’s topology. This particular attribute presents a network topology.
multitude of issues, particularly in the context of routing pro-
tocols, as self-organizing protocols become indispensable in
these ecosystems [14, 15]. The essential components of the
IoT are self-organization, information exchange, and infor- 2 The main contribution of this work
mation routing. To facilitate these operations, energy- and
computation-limited devices employ routing protocols that The contributions of this study are significant and note-
prioritize optimization at the expense of security, in par- worthy. The contributions of this study are summarized as
ticular. Furthermore, due to the escalating proliferation of follows:

123
A novel algorithm for the development of a multipath protocol... 771

• In this study, we provide a novel Disjoint & Scalable & The following portions of this work are organized in the
Multipath Routing (D&SMR) algorithm and an adaptive follows manner: Sect. 3 presents a comprehensive review
multipath routing protocol specifically designed for IoT of the current body of research on safe routing inside the
and MANETs. IoT framework. Section 4 of this article provides a thorough
• In this study, we present a mathematical assessment examination of the secure routing protocol that we have cre-
model that encompasses key attributes of a node, includ- ated. In addition, we give a mathematical model to evaluate
ing mobility and uncooperative behaviors. This model is the technique that we have suggested. Section 5 of this study
designed to be applicable to many ad hoc routing proto- evaluates the evaluation of our technique and the results that
cols. were obtained. To summarize, Sect. 6 functions as the con-
• The proposed technique aims to improve node energy cluding part of the study and also outlines potential areas for
efficiency by optimizing the energy consumption of sup- further research.
plementary nodes in the provided topology that are not
part of the primary path. In scenarios where nodes within
a network experience migration or new nodes join, it 3 Background and related work
is feasible to employ an alternative node situated on a
distinct path from the primary one to alleviate any dis- The issue of mobility challenges in IoT routing protocols
turbances to the primary path. has been the subject of substantial research in recent years,
• If any node inside a multipath or disjoint network as evidenced by several scholarly works [16–19]. Numerous
becomes inactive or fails, our algorithm will initiate the studies have consistently highlighted the absence of stan-
creation of a new, distinct multipath or disjoint network dardized protocols in the domain of reliable routing for IoT
prior to any network events taking place. devices, with particular emphasis on ad hoc routing. Numer-
• The Internet of Things-enabled network (IOTN) is now ous previous studies [20–23] have focused on the develop-
a prominent field of research. While the IoT has been ment of routing protocols that prioritize improvements in
successfully used in several real-time applications, it is routing efficiency rather than mesh topology connections,
important to acknowledge that one significant limitation owing to the distinctive attributes and limitations of IoT
of IoT is its energy constraint. The management of energy devices. The IoTN has garnered considerable attention in
limitations is achieved through the use of several energy- academic studies because of its extensive potential for appli-
efficient strategies. Of all the actions conducted by the cation in several areas such as healthcare, surveillance, item
IOTN, routing stands out as the most energy-intensive tracking, and environmental monitoring, among others. The
task. In light of the aforementioned considerations, this author has provided a citation range, indicating that they are
paper presents a proposed route selection strategy that referencing multiple sources. An IoTN consists of a mul-
takes into consideration both the reliability score and titude of wireless sensors that possess the ability to detect,
the sensitiveness of the data. The nodes’ reliability score transmit, and analyze data. Wireless sensors are commonly
is computed using evidence theory, and users are given put in regions that are inaccessible to humans, posing chal-
the opportunity to specify the expected reliability score lenges in sensor management. The sensors are responsible for
for data forwarding. According to the specified criteria, monitoring the surrounding environment and gathering the
the route selection is prioritized based on the depend- necessary data. The data that has been gathered is transmitted
ability score, with secondary consideration given to the to a high-capacity node for additional data processing, facil-
route’s length. The suggested technique’s performance itated by data transfer. Data transfer can occur through two
is evaluated using common performance indicators, and distinct methods, namely direct and indirect data transmis-
the findings demonstrate that the proposed approach pro- sion. Data transmission encompasses two primary entities,
duces superior results. namely the source and destination nodes. The source and
• The objective of this study is to provide many pathways destination nodes may alternatively be referred to as sender
from a given source to a destination, with the aim of miti- and receiver nodes. When engaging in direct data transmis-
gating the impact of a single point of failure.The optimal sion, the source node is able to establish a direct connection
pathway selection is determined by taking into account with the destination node. Nevertheless, this is not feasible in
the level of sensitivity associated with the data being con- all instances. The most prevalent kind of data transfer is an
veyed. The calculation of route dependability entails the indirect method of communication. Indirect data transmis-
use of trust characteristics. The incorporation of the clus- sion is the conveyance of data from a source to a destination
tering idea has the effect of lowering energy usage and via one or more intermediate nodes. The present methodol-
enhancing manageability. ogy raises a number of concerns, as the forwarding behavior
of the intermediary nodes remains uncertain, and there is a
potential for malevolent intent on the part of the nodes. The

123
772 R. S. Abujassar

aforementioned factor has an impact on the velocity, uni- To assess their impact on network performance. Accord-
formity, and dependability of data transfer. Therefore, it is ing to the source, sensing devices are activated exclusively
important to take into account all of these elements in order through heat application. The REERS system offers energy
to guarantee the secure transfer of data. efficiency solutions for extensive implementations [27]. Effi-
Despite the manifold benefits of the IoT, including cient route selection for data aggregation is crucial in IoT
its capacity for self-governance, dynamic topology, cost- contexts characterized by extensive connectivity. The REER
effectiveness, and ease of deployment, it continues to face framework takes into account both real-time and non-real-
limitations in terms of energy consumption, memory use, time situations. A routing system that prioritizes energy
and processing capabilities [24]. Out of all the aforemen- efficiency utilizes cost-effective paths to enhance the per-
tioned concerns, the issue of energy restriction stands out formance of real-time applications by reducing latency and
as the most crucial problem that requires immediate atten- optimizing energy consumption. In their study, they con-
tion. The utilization of a sensor’s energy is facilitated by ducted data aggregation from designated monitoring stations,
three modules, namely sensing, processing, and communi- with a particular emphasis on optimizing the data gather-
cation. Out of all the many capabilities, communication is ing process. The scheme’s algorithm aggregates data from
the most energy-intensive. Routing is a pivotal procedure both collecting and loading regions. The proposed method-
that facilitates communication between sensors, and it is ology enhances the longevity of the network by evaluating
often recognized as the most energy-intensive operation [25]. the expense of the path between different regions and then
Due to the dynamic nature of the IoTN, determining opti- transmitting the load of each sensing node. Researchers
mal paths from the source to the destination is a significant have found that mobile data collection sinks exhibit energy
challenge. Furthermore, the Index of Orthodontic Treatment efficiency. Mobile devices can interact with data sensors
Need IoT demonstrates active engagement in many real-time to collect and combine data while navigating a specific
applications. Therefore, it is imperative for the routing mech- path. This process involves many stages, including initial-
anism to provide both reasonable data transmission time and ization, aggregation region selection, and path design, as
enhanced dependability. In order to ensure low energy use, outlined in reference [28]. The authors in [29] examine the
it is imperative that these fundamental standards are met. energy and dynamic spectrum considerations associated with
Attain satisfaction by using less energy. This article pro- cluster-based routing. The implementation of self-distributed
poses a reliable route selection technique based on clustering, clustering techniques results in a reduction in power con-
which effectively addresses the aforementioned difficulties sumption and the generation of optimal clusters. The system
while also ensuring energy efficiency. This study employs possesses a greater number of co-sensing channels. The sys-
a cluster-based method due to its energy-saving potential. tem transmits data across clusters by utilizing gateway nodes
The essence of the clustering strategy is in the management with higher energy levels and shared channels. It determines
of node members by the Node Cluster Head (NCH) node the head node by considering factors such as residual energy,
inside the clusters. The energy consumption of the NCH node channel availability, neighboring nodes, and distance from
is reduced, resulting in balanced energy consumption. This the sink. The proposed solution provides an efficient method
study ensures dependability by incorporating the fundamen- to transfer data from the root to the sink. Achieving green
tal principle of trust. The trust parameters play a crucial role routing in a wireless sensor network (WSN) poses signifi-
in characterizing the characteristics of sensor nodes. More- cant challenges. Wireless sensor networks (WSNs) possess a
over, the present study aims to examine a specific facet of finite amount of energy, which is allocated towards the exe-
data security, focusing on user preferences. This study cal- cution of crucial activities in remote locations. Consequently,
culates numerous routes between the source and destination the availability of energy is diminished, thereby resulting in
and ultimately selects the route with a higher dependabil- a reduction in the overall lifespan of the network.
ity score as the final route. In their publication, the authors In [30], double deep Q-learning dynamic arithmetic
provide a novel methodology known as the multi-hop, adap- reinforcement learning (DQARL) is proposed for energy-
tive, tree-based energy-balanced (DMATEB) method [26]. efficient clustering and scheduling in reliable routing. The
The algorithm determines the nearest node that possesses initial step suggests using Hunter Prey Optimization (HPO)
a substantial amount of remaining energy. The utilization to find the best head. The selection includes many fitness
of multi-hop routing, which incorporates the use of clus- functions. In the second phase, duty cycling extends network
ters, enhances both data collection efficiency and the overall life. This solution uses distributed scheduling to efficiently
lifespan of the network. The authors in [27] introduced the minimize sensor node energy usage. Schedulers examined
EACRA system as a means of mitigating the issue of peri- sleep, listen, and transmit modes. Weighted Practical Byzan-
odic duplicate data transfer. The data is transferred from tine Fault Tolerance (WPBFT) ensures stable routing and
sensing devices in response to contextual changes, result- reduces packet transmission time in the third phase. This
ing in reduced energy consumption and message exchange. method selects the optimal source-cluster head path. This

123
A novel algorithm for the development of a multipath protocol... 773

study investigates the power consumption of the compressor neighboring nodes. This approach reduces the costs associ-
during the 20-minute start-up of the HRF under certain fail- ated with the Low Energy Adaptive Clustering Hierarchy
ure scenarios, including blockage of the 3-way valve and (LEACH) protocol. In [38], we propose a network struc-
clogging of the capillary tube. The efficacy of these set- ture called the delay-aware data collection network structure
tings was evaluated in six failure scenarios. The compressor’s (DADCNS) in order to select tracks that have the least amount
power consumption is rapidly affected by both quantitative of latency in data gathering. This approach uses both top-
and qualitative changes in the refrigerant stream caused by down and bottom-up techniques to construct a hierarchical
these difficulties. The compressor power data clearly exhib- tree structure that governs the distances, rankings, and delays
ited noticeable discrepancies in relation to the analyzed of the nodes.
issues with refrigerant passage changes, serving as a sig- In [39], the reduction of energy utilization is achieved by
nal for fault detection [31]. The author introduced a novel the implementation of a multi-hop algorithm that balances
algorithm in the publication referenced as [32]. During the routing costs. The algorithm is capable of withstanding fail-
optimization process, several factors are carefully consid- ures in both head and sensor nodes. In order to optimize
ered, including channel fade margin, cross-correlation and energy dissipation, the SEED protocol is designed to min-
coherence time, spectral efficiency, interference level, power imize the occurrence of duplicate transmissions. However,
consumption, retransmission rate, access probability, and it is not compatible with various sensor data. To tackle the
propagation delay. To improve the scheduling efficiency of issue of heterogeneous data trafficking and energy utilization
the DE algorithm, many steps are meticulously designed: in sensor nodes, an additional approach was introduced. In
startup, mutation, crossover, fitness evaluation, selection the LEACH Vice Cluster Head (L-VH) [15] protocol, a sub-
for iteration evolution, and termination. The OMPS model ordinate cluster head with the second highest energy level is
assesses many measures including as latency, PDR, spectrum responsible for managing the workload. The VH system does
usage, interference level, energy efficiency, and the success not operate simultaneously with the NCH system; rather, it
rate of created paths. remains in a dormant state until the CH system uses a suffi-
The present technique employed by MDTA exhibits cient amount of energy. Upon awakening, VH assumes the
reduced energy consumption in comparison to prevailing responsibilities previously assigned to NCH and thereafter
energy-aware routing systems. Additionally, it evaluates the fulfills CH’s role. The reduction in the time required for
cost of data transmission in relation to the size of the net- selecting the next head and the decrease in the number of
work. The Agile data delivery system employs a very efficient messages processed by VH for communication management
data transmission mechanism that facilitates seamless com- are seen. The researchers in the referenced study propose
munication between several nodes. The MDTA algorithm a novel approach called Traffic and Energy-Aware Routing
enhances the formulation of the root-to-sink path by consid- (TEAR) to identify nodes with high traffic and low energy
ering both end-to-end latency and energy usage. This concept consumption. In this particular scenario, TEAR does not offer
is applicable to a wide range of sensor and vehicle networks redundant data filtering. The ETASA technique, which is a
[33]. Wireless Sensor Networks (WSNs) encounter chal- hybrid approach, offers enhanced load balancing by effec-
lenges in achieving energy-efficient data monitoring. The tively managing the energy consumption of sensing devices.
authors in [34] propose a scheduling approach that takes These devices are able to alternate between active and dor-
into account both sensor and network energy, with a focus mant states depending on considerations of energy levels and
on energy efficiency and quality. The effectiveness of data data transmission speeds. This paper aims to provide a rout-
fusion is contingent upon the costs associated with data for- ing method that prioritizes both route dependability and data
warding as well as the power capacity of the sensing nodes. sensitivity, drawing inspiration from previous research in the
The proposed methodology aims to reduce the number of field. The majority of current literature focuses on providing
active nodes while ensuring the integrity and dependability security measures; however, it fails to address the ability for
of the data. In their publication, the authors in [35] presented users to customize their route preferences. Nevertheless, this
a routing system that utilizes location-based information to study enables the author to establish the anticipated reliabil-
enhance the optimization of residual energy and network ity score of the route, taking into consideration the sensitivity
quality. The PEDAP algorithm was developed based on the of the data that is to be transferred. This concept serves to
concept of a near-optimal spanning tree with the objective of safeguard data against potential security risks while also mit-
enhancing the lifetime of a network [36]. The localized PEDP igating energy use. Furthermore, the work demonstrates an
system described in Reference [37] utilizes a distributed increased quality of service (QoS). The subsequent section
design, in contrast to other systems that adopt a centralized will provide an explanation of the functionality of the sug-
approach. In the seventeenth reference, the power-efficient gested strategy. A comparison has been shown in Table 1.
gathering in sensor information systems (PEGASIS) proto-
col is described as a method that exclusively sends packets to

123
774 R. S. Abujassar

Table 1 Comparison with

Method Strengths Weaknesses
existing Protocol
active trust [40] High network stability is achieved High communication cast, routing
by selecting reliable neighbors in delay
secure transmission using a simple
decision-making method
Optimum routing [41] Changes the major nodes and iden- High communication overhead,
tifies the movement that is occurring high time complexity
on the primary nodes
reliable routing [42] Using a hierarchical architecture, The failure is neither mitigated nor
taking into account energy effi- anticipated using this method. The
ciency and network security at the design procedure lacks considera-
same time, taking into considera- tion for clustering, has a subpar trust
tion the amount of energy that is mechanism, and is not scalable
still available in nodes while discov-
ering new pathways, and employ-
ing hybrid cryptography to ensure
that connections between nodes are
properly protected
Disjoint & Scalable Multipath In the event of a disconnection, There is a need for further enhance-
Routing (D&SMR) reroute the traffic, and there will be ment in order to achieve higher
no additional cost associated with speed and improved security during
enabling UDP live streaming traffic. the exchange of data packets
This will result in an improvement
in power usage

4 Proposed an optimal rout selection illustrates a schematic representation of the overall flow of
strategy the proposed work.
When a source node desires to send a message to a des-
Despite the multitude of advantages associated with the IoT, tination node, the proposed methodology yields numerous
its primary issue is energy limitations. In the majority of potential pathways from the source to the destination. Fur-
real-time applications, sensors are deployed in hostile envi- thermore, the user has the option to indicate the anticipated
ronments to facilitate the acquisition and analysis of data. As reliability of the chosen route. The selection process involves
a result, periodic recharging or replacement of sensors is not choosing the route that aligns with the desired level of trust-
feasible, potentially resulting in a node’s premature failure. worthiness from a set of available routes. As a result, this
The prevention of this issue can only be achieved through the study takes into account the data’s sensitivity and makes a
prudent utilization of the existing energy resources. Energy modest contribution to the field of data security. This study
consumption is inversely proportional to the duration of net- primarily emphasizes the aspects of energy efficiency, route
work operation. This study aims to explore several strategies reliability, and a little consideration of data security. The sub-
for minimizing energy usage, with a specific focus on the sequent section provides a comprehensive explanation of the
implementation of an energy-efficient and secure routing proposed research.
method. The energy consumption associated with the func-
tioning of sensors is mostly attributed to the routing process,
which surpasses all other processes in terms of energy uti- 4.1 Proposed D&SMR with its assumption
lization. Therefore, this study presents a routing policy for
the IoT that aims to enhance energy efficiency and ensure This paper is conducted based on specific assumptions, and
trustworthiness. The present study employs the cluster model the proposed methodology demonstrates superior perfor-
as a means to achieve the objective of enhancing energy mance when aligned with these assumptions. The assump-
efficiency. In this particular framework, the NCH has the tions are enumerated as follows:
responsibility of overseeing and controlling the activities of
the node members. The NCH undergoes alteration within
a designated time frame. In this manner, the utilization of • This study involves the implementation of a resilient
energy in the sensor nodes is distributed evenly, resulting in mobile base station (BS) that is provided with a battery
a consistent energy consumption pattern. The Algorithm 1 of substantial capacity.
• The nodes within the network stay immobile and have
limited mobility at low speeds or with no movements.

123
A novel algorithm for the development of a multipath protocol... 775

• All nodes are equipped with information about their indi- Table 2 Math Notation
vidual positions, notwithstanding the absence of Global Notation Definition
Positioning System (GPS) functionalities. The network is
composed of nodes that may be classified as either NCH n Nodes
or node members. t Trust node
• The NCH possesses the capability to establish communi- R Reliable Node
cation with both the Base Station (BS) and other Cluster R̄ Non-Reliable Node
Node Heads. The nodes possess the ability to communi- K Represent all nodes Reliable OR non-Reliable
cate only with their respective cluster heads. δ Probability for node trust
T Denoted to the shortest path
4.2 Topology formation with its NCH z Denoted to Next Hop
z̄ Denoted to Next Hop but in the main path
Clustering’s primary goal is to allocate a NCH for each clus- d Distance
ter, ensuring that the NCH manages node members. The IoT K_s Nodes in the route from source
lacks a centralized authority for overseeing node function-
ality, which is addressed by the introduction of a NCH to
resolve this concern. Given the significant function played by node is denoted by δ and may be expressed as follows 3 and
the NCH, it is imperative that it exhibit qualities of trustwor- 4:
thiness and honesty. As a result, it is critical to observe node
behavior and select the most appropriate nodes to serve as the z = {R}; z̄ = { R̄}; K = t;
NCH. It serves as the cluster’s governing body. Therefore, it is T 1(z) = δ; T 1(z̄) = 0; T 1(K ) = 1 − δ; (3)
imperative that the NCH organization maintain a high level of
trustworthiness in its functioning. Nevertheless, the process In this manner, the calculation of node reliability is per-
of assessing the reliability of a node is more complex than formed, and the computation of dependability scores by two
it initially appears. While it is possible to quantify a node’s distinct nodes that are adjacent is expressed as follows. The
trustworthiness, it is not advisable to rely solely on the trust equations provided below illustrate fictitious representation
evaluation provided by a single node. As a result, this study of the aforementioned events.
chose two distinct neighboring nodes to evaluate the sensor
node’s dependability score. The evidence theory is used to z = {R}; z̄ = { R̄};
measure the nodes’ dependability score. The use of evidence
K = t; T 1(z) = 0; T 1(z̄) = δ; T 1(K ) = 1 − δ; (4)
theory is advantageous because it does not require any prior
knowledge of probabilistic ideas. The real reliability score Equation (5) represents the proposal made by two neigh-
is obtained by combining the reliability scores computed by boring nodes. The node has a significant degree of reliabil-
two distinct neighborhood nodes; all notations are shown in ity. Equation (6) depicts the scenario when adjacent nodes
Table 2. As denoted by Eq. (1), it is possible to categorize sen- acknowledge the existence of unreliability in the node.
sor nodes as either trustworthy or non-trustworthy, as well as Ultimately, the node is classified as either dependable or
reliable or non-reliable. undependable. The significance of qualities is critical in the
decision-making process, and several approaches have been
suggested to determine the appropriate weights.
n : t = {R, R̄} (1)
1
The aforementioned Eq. (1) illustrates the potential trustwor- T 1(z) ⊕ T 2(z) = T 1(z)T 2(z)
d
thiness of sensor node n, which may be classified as either +T 1(z)T 2(K ) + T 1(K )T 2(z)
trustworthy or non-trustworthy. This equation serves as a rep- 1
resentation of this classification as following: T 1(z̄) ⊕ T 2(z̄) = T 1(z̄)T 2(z̄)
d
+T 1(z̄)T 2(K ) + T 1(K )T 2(z̄)
z = {R}; z̄ = { R̄}; K = t (2) 1
T 1(K ) ⊕ T 2(K ) = T 1(K )T 2(K ) (5)
d
In the above equation, the symbol R represents the concept
of reliability, whereas the symbol R̄ denotes non-reliability. where d is given by
Additionally, the symbol K represents the hypothesis pertain-
ing to the potential reliability or non-reliability of the node. d = T 1(z)T 2(z) + T 1(z)T 2(K ) + T 1(K )T 2(z)
The probability function denoting the trustworthiness of the +T 1( Z̄ )T 2( Z̄ ) +

123
776 R. S. Abujassar

T 1( Z̄ )T 2(K ) + T 1(K )T 2(z̄) + T 1(K )T 2(K ) (6) tematically evaluates each path based on its node count and
dependability score. The selection of a path is contingent
If there are numerous nearby nodes aside from the prin- upon the presence of a matching dependability score, which
cipal node, the following equation will be used to begin the is mentioned in conjunction with the path request message.
computation of the ideal weight based on the most advanta- When the dependability score is lower than expected, the
geous distance, using the D&SMR technique. As a result, the shortest path is disregarded. The dependability score for the
following equation explains how the primary approach deter- given path is computed using the node’s reliability score,
mines the most favorable weight for the nearby elements. We which is measured on a scale from zero to one. A score of
made each cluster head pick the next alternative hop based one signifies that the node is entirely trustworthy, while a
on the minimum distance, as indicated: ∀ CH→ NextHop = score of zero indicates that the node is entirely unreliable.
min(distances−d ) Until arriving at the destination, the route The intermediate value of 0.5 indicates that the node has
will serve as the foundation for the new routing table by cre- a certain degree of reliability. When the base station (BS)
ating the matrix_route as follows: M Rout equals min(∀ CHnh encounters a cluster, all the nodes that are part of the cluster
with minimum distance ∈ Cluster). transmit the dependability score together with their identi-
The node’s dependability score ranges from zero to one, fication to the BS. The base station (BS) selects the node
with a value of one indicating perfect reliability and a value with the highest dependability score and formally designates
of zero denoting complete unreliability. The presence of an it as the NCH. Subsequently, the remaining nodes transmit
intermediate value of 0.5 suggests that the node has limited the JOIN request to the NCH and operate under the CH’s
reliability. When the base station (BS) encounters a cluster, purview. As a result, the cluster selects the most dependable
all the nodes that are part of the cluster transmit the depend- node, the NCH, as shown in Algorithm 1 (Table 3).
ability score together with their identification to the BS. The
base station (BS) selects the node with the highest depend-
ability score and formally designates it as NCH. Following
that, the remaining nodes transmit the JOIN request to the Algorithm 1 Algorithm for NCH
NCH and operate under the NCH’s purview. As a result, 1: Cr eatalltoplogiesthenselect N C H
2: count = 0; Min = 100sec
the NCH is selected based on the identification of the most 3: (N C H → selected Basedon L S D B I n f o.Histor y)
dependable node within the cluster. When the nodes inside a 4: count E xchangeI nquir y Packetsoutbound f r omallnodes
cluster need to transmit data to nodes in another cluster, they 5: Allad jacent N odes ←∈ / mainpath(N C H &Dest)
accomplish this by utilizing the NCH node. The NCH plays 6: while N H ¬main Route do
7: get N H _i_d
a vital role in the process of routing, as it is responsible for 8: count + +
managing and maintaining two essential tables: the recent 9: Computether eliableallnodesnotinthemainpath
pathways database and the reliability scores table. When a 10: K _r = N H (K _r ) + N N H (K _r )
network node intends to transmit a message to another node, 11: if K _r > count then
12: Declar etheK istheN C H
it proceeds by forwarding the PINQ message to its desig- 13: end if
nated next-hop controller. The PINQ message consists of 14: if r out[K ][N Hid ] = ad jandr out[K ][N Hid ] = Primar y Ad j
three components: the source ID, the destination ID, and then
the predicted path reliability. The NCH verifies the destina- 15: N ext H op[K ] = arar out[K ][N H ]
16: i ← ID+1
tion identification (ID) and cross-references it with the most 17: Fomring Route[K ][N H ] =
up-to-date path database. The current path table comprises r out[N ext H op[N H _i_d]][Ad jacent]
the paths that have been most recently accessed. The table 18: end if
in question is designed to accommodate a maximum of 10 19: P_r = (NCH, V _i), (NCH, D)
20: N C H = Find(AD J _LeadT o_Dest(D)
entities, with a mechanism in place to automatically remove 21: ∈ / M N H , Long Distance)
older entities. This concept effectively preserves memory and 22: Check_its_K (Ad jacent_N ode)
facilitates the path-searching procedure. The routing method 23: H ead_Path_to_Destination = Head− > V _i =
is outlined in the following manner: Assuming that the table (V _i + 1, D)
24: V _i_Ad jacnet_Labeled As Available = (P_r out) ∈ /
includes the anticipated trajectory, the message’s dependabil- Destination∩
ity score is evaluated. If a certain route meets the specified 25: ∀Ad jacentsourr ounded As_alt =
condition, it is chosen for the purpose of transmitting data. Head− > V _i− > Destination
Conversely, in instances where the required path is absent 26: end while
27: r etur n N ode_has_ AD J W ith Low Dist_Build Path;
from the database, the process of path discovery is initiated,
leading to the retrieval of various pathways. The NCH sys-

123
A novel algorithm for the development of a multipath protocol... 777

Table 3 Algorithm Notation and Definition time it takes for a data packet to be transmitted and success-
Notation Definition fully received. Before evaluating the performance effects of
network topologies on the calculation of a backup path in
NH Next Hop MANET, it is essential to determine the network components
NNH Next Next Hop that could affect the quality of service (QoS) of the video
ADJ Adjacent Node data being sent. This research specifically examines three
Dest Destination distinct qualities. This study aims to elucidate the impacts of
alt Alternative Node video traffic strategies by specifically examining three key
S Source attributes. To evaluate the influence of node density, it is
NCH Node Cluster Head critical to recognize that higher densities are associated with
PINQ Packets Inquiry longer lifespans compared to lower or intermediate densities.
This difficulty occurs because the latter’s capacity to accu-
rately detect and maintain new pathways decreases when
the nodes spread out into different clusters, increasing the
5 Simulation experiment chances of creating a disconnected structure. When there
is less movement, regions consisting of nodes demonstrate
An evaluation was conducted using network simulation to increased stability, which allows for the provision of ser-
assess the effectiveness of the upgraded D&SMR protocol in vices for longer periods of time. After dividing the networks
NS2-simulated networks with both high and low node den- into many clusters, we conducted measurements to assess the
sities. The results derived from comparing the simulation packet delivery ratio, throughput, latency, and power con-
of the D&SMR protocol with other well-established proto- sumption.
cols, such as active trust [40], optimum routing [41], and
reliable routing [42], The NS2 simulation demonstrated the 5.1 Performance and analyses evaluation
potential of IoT technology to improve nodes’ preparedness
and agility in various connection situations. Radio propaga- In the simulation conducted, there were a total of 200 nodes
tion required a transmission capacity of 0.28 watts. Within when there was mobility and 300 nodes for a stable network.
a radius of 250 ms, the technology allowed nodes to send All nodes were distributed inside a square space of 1000 x
and receive data packets. To enable wireless transmission of 8002 . The distribution of all nodes is uniform and random.
multimedia files, the researcher included IEEE 802.11b tech- The placement of the sink node is positioned in the cen-
nology at both the physical and data-link layers. The NS2 troid of the area. Each node in the network is subject to a
utilized the Way-Point random mobility paradigm inside a random waypoint mobility (RWM) model, where the speed
roaming area measuring 1000 x 800m 2 . During the simu- of the nodes ranges from 0 to 10 m/s [43]. The simulation
lation, clusters of nodes emerged. The predicted beginning parameters have been described in Table 4. In this study,
velocity of 1 m/s was moderately low. However, the effects of we conducted a comparative analysis of D&SMR in rela-
this velocity selection will become apparent and substantial tion to active trust, reliable routing, and the optimum path.
in future research and relationships. Each simulation lasted The evaluation was based on many performance metrics,
for a length of 600 s. The experimental approach involved including PDR, energy consumption, average end-to-end
conducting one hundred trials, followed by the calculation (AE2E) latency, network lifespan, network energy consump-
of the mean value. The maximum length of a packet was tion, cluster formation time, and cluster lifetime.
1024B, although its capacity was restricted to 512 bytes. The
bit rate was set to 2Mb/sec. It is widely known that devices • CLUSTER Life TIME VS. NUMBER OF NODES
in an ad hoc configuration network have the ability to share The cluster building time refers to the processes of cluster
a wireless connection. The first simulated scenario consisted formation and NCH selection. A prolonged duration for
of an evenly distributed density of 200 to 300 nodes in each cluster setup implies that the procedure exhibits a high
cluster location. Throughout the experiment, a steady data level of complexity. Figure 3 demonstrates the influence
transmission rate of 512B per second was observed between of node density on the duration required for clustering
the source and destination nodes. The next part presents construction. As we can see, the proposed algorithm has
the results of the simulation, which are depicted visually in reduced the time by around 30% compared with reliable
the form of line graphs. The specific characteristics of the routing and 5% to 10%. The D&SMR algorithm was also
diagrams are provided in Table 4. The packet loss ratio is used for NCH selection, relying on the number of nodes
determined by dividing the number of discarded packets by in each cluster and the best energy for the nodes. Further-
the total number of sent packets. The average end-to-end more, based on the measurement distance, the sink node
latency is a statistical measure that calculates the average will be in the center of all cluster nodes.

123
778 R. S. Abujassar

Table 4 Simulation Parameter

Parameter Value

Simulation Time 600 sec

Simulator Network Simulator (NS2)
Number of Nodes 200–300
Wireless Channel Unit Disk Graph Medium (UDGM)
Max. range 250 m
Roaming area 1000 X 800 m2
Min. Speed 0 m/s
Max. Speed up to 10 m/s
Mobility model Random WayPoint Mobility
Threshold Value 1Mb
Initial Energy of Nodes 2J
Packet Size 512B
Propagation Model Free-Space
Compared Routing Protocol Reliable Routing,Active Trust, Optimal Routing
Transport Protocol UDP
CBR rate 2Mbps

Energy(i)
NC H LF = (7)
E total

Once the energy of a node drops below a certain thresh-

old, the algorithm must start the process of selecting
the NCH. In the topology, the variable i represents the
energy level of each node. A shorter cluster lifespan
results in higher energy usage because it necessitates re-
clustering and NCH selection. Figure 4 depicts a clear
inverse relationship between node densities and clus-
ter longevity, suggesting that as node densities increase,
cluster duration decreases. The frequent alterations in
network design, resulting from the movement of nodes,
have an adverse impact on the longevity of the cluster,
leading to a reduction in its length. The diagram illus-
trates the energy usage generated by the transmission and
Fig. 3 Cluster Building Time reception processes at both the source and intermediate
nodes that have successfully delivered the packet. It is
important to note that with D&SMR, The network dis-
plays lower energy consumption in comparison to other
In the case of a large number of nodes, the duration
networks. Reliable routing, active trust, and optimal rout-
required for cluster formation increases significantly
ing use energy by transmitting and receiving packets that
across the majority of clustering methods. In the D&SMR
are then lost due to buffer overflow and congestion. The
framework, the developed PINQ packet was employed to
route continues without achieving a successful delivery.
effectively decrease the clustering construction time. The
duration between cluster development and dissolution is • PDR VS. NUMBER OF NODES
an indicator of its longevity and a significant determinant
of network energy consumption. Within the context of The protocol’s performance was investigated by manip-
D&SMR, it is observed that a node possessing a greater ulating the network’s node density. In this investigation,
fitness value assumes the role of the NCH. In the present we varied the number of nodes within the range of 20
study, we examine residual energy as a primary determi- to 200. The node density has a substantial impact on the
nant for NCH selection. The cluster lifetime was defined PDR. Because of their low density, the nodes frequently
in Eq. (7). encounter link disconnection problems, resulting in a

123
A novel algorithm for the development of a multipath protocol... 779

Fig. 5 Packet Delivery Ratio

Fig. 4 Life Time

decrease in the PDR. The Fig. 5 illustrates the aggregate

number of packets lost per second in the network due to
buffer overflow and loss in the wireless channel. Consid-
ering the aforementioned variables, it is not unexpected
that D&SMR experiences a significantly lower rate of
packet loss at the buffer. The D&SMR technique employs
the modified PINQ packet policy, whereas active trust,
dependable routing, and optimal routing lack a mech-
anism for adapting the transmitting rate. However, the
quantity of lost packets stays constant. The levels of inac-
Fig. 6 Packet Over All Ratio Received
tive trust, dependable routing, and optimal routing are
higher in comparison to D&SMR, which has an impact on
the received PDR. The graphic illustrates that D&SMR, • AE2E DELAY VS. NUMBER OF NODES
active trust, reliable routing, and optimal routing transmit
2.47, 5.41, and 25.91 packets per second, respectively. Figure 7 illustrates the end-to-end latency, which repre-
Furthermore, the numerical value The packet loss rate sents the duration from when a packet is created at the
per second for active trust, reliable routing, and opti- source application to when it is successfully received
mal routing is 0.32, 0.44, and 0.56, respectively. After at the destination application. D&SMR has a reduced
increasing the number of nodes, there is a significant end-to-end latency in comparison to active trust, reliable
increase in the PDR. Regarding all the procedures, Our routing, and optimal routing. D&SMR initially scans for
proposed D&SMR demonstrates improved performance an uncongested area around the forward packets. If con-
when compared to active trust, reliable routing, and opti- gestion or other issues persist, the quantity of packets
mal routing strategies. This is because we have taken injected into the network is decreased by lowering the
into account the effectiveness of the two-hop neighbor- sending rates of the nodes. As a result, the buffer for
based clustering strategy, which is superior to the one-hop the neighboring nodes can quickly accept and transmit
neighbor-based clustering approach. Neighbor informa- packets without experiencing significant buffer delays.
tion improves the overall understanding of the network. However, active trust, reliable routing, and optimal rout-
The results illustrated in Fig. 6 demonstrate that an ing experience significant delays due to the periodic
increase in the number of nodes inside each cluster results increase and decrease of exchanged packets in each
in a proportional increase in the availability of paths topology, which is influenced by the network conditions.
between the source and destination. Consequently, the This process persists when there is congestion. As a
network’s performance is affected. result, the packets experience significant delays in the
nodes’ buffers. While active trust routes packets down
less crowded channels, it does not implement a mecha-
nism to decrease the sending speeds of nodes when buffer

123
780 R. S. Abujassar

Fig. 7 End to End Delay

Fig. 8 Latency analysis

drops still happen. As a result, packets will encounter sig-

nificant delays from one end of the network to the other
if the buffers are consistently filled. There is a clear pos-
itive correlation between the density of nodes and the
latency of AE2E. This is because the task of determining
the node’s location and creating routes becomes time-
consuming when the nodes are moving. Implementing
our two-hop neighbor strategy leads to a decrease in the
latency of node AE2E when compared to clustering and
routing protocols that rely on one-hop. Within the frame-
work of D&SMR, the application of NCH in clustering
is used to improve the stability of clusters.

• NODES LATENCY vs ENERGY CONSUMPTION

Fig. 9 Energy consumption vs. number of nodes
The experimental findings demonstrate the effectiveness
of the suggested technique, as evidenced by the observed
performance. Additionally, a latency analysis of the pro-
posed approach is presented in Fig. 8. The suggested
technique has a higher packet delivery rate because it con-
siders the trustworthy path rather than only prioritizing
the shortest route. Trustworthiness is prioritized above
the length of the journey. The suggested work demon-
strates an enhanced packet delivery rate. The latency
analysis assesses the average delay experienced by each
individual technique. Minimizing latency is crucial in
order to increase the quality of service (QoS). This study
demonstrates low latency due to the selection of reliable
sources and the utilization of the most efficient route. The
presence of malicious nodes along the route could poten-
tially cause packet-forwarding delays. This study adopts
a trustworthiness-based approach, resulting in reduced
delay. Fig. 10 Power consumption for the D&SMR within simulation
Figures 9 and 10 demonstrate the relationship between
energy usage and node density fluctuations. The D&SMR
protocol demonstrates a lower energy consumption, sav-
ing around 35% in comparison to other protocols that are

123
A novel algorithm for the development of a multipath protocol... 781

equivalent in nature. The consideration of node energy 6 Conclusion and future work
is deemed essential in the processes of clustering, NCH
selection, and routing, owing to its impact on achieving This article presents a method for choosing a way that consid-
an optimal number of clusters. In the suggested method- ers both the reliability of the road and the level of sensitivity
ology, we employ an analytical optimization technique of the data being transmitted. Certain data possesses a higher
to determine the optimal number of clusters for our level of sensitivity, which in turn raises the probability of
network’s spectrum power analysis. This optimization potential modifications or removals taking place through-
process is carried out throughout the simulation, as shown out the routing process. Each individual pathway has the
in Fig. 10. This is because the next alternative hop will be capacity to contain nodes that are either reliable or malicious.
in sleep mode until a failure or any disconnection occurs The presence of malevolent nodes might result in unfavor-
during the transmission. The relationship between total able behaviors, rendering the route that incorporates them
energy consumption and round work may be observed unsuitable for sending extremely sensitive data. This study
in the context of minimizing the number of single-node introduces a new method for choosing a path that allows users
clusters, leading to reduced energy consumption. Fig- to designate the desired level of dependability for the chosen
ure 10 illustrates the declining trend of the node’s residual path. The dependability score is calculated on a scale ranging
energy rate. The D&SMR protocol has a larger residual from 0 to 1. The path that satisfies the specified dependability
energy compared to other protocols due to its ability to score is further optimized based on its duration. The execu-
disperse energy use among nodes, resulting in a slower tion of these procedures is carried out using a NCH with
depletion of energy in individual nodes. Hence, in the the objective of conserving energy. This concept efficiently
context of D&SMR, a greater number of nodes have a preserves energy and maximizes the lifespan of the network.
higher likelihood of survival in comparison to alternative The objective of this study is to investigate the capacity of
techniques. mobile nodes to improve energy efficiency. Our future work
The graph depicted in Fig. 9 showcases the rate of will focus on integrating D&SMR with other IoT technolo-
energy consumption among the nodes. A negative cor- gies, including edge computing and 5 G/6 G networks. We
relation exists between the lifespan of a network and its will also modify or improve the algorithm to tackle the issues
energy use. A network’s energy consumption is inversely posed by next-generation IoT systems. The clustered design
proportional to its longevity, meaning that a longer lifes- of the head node will result in further power savings. It will
pan is correlated with lower energy consumption. After also facilitate the expansion of the network to accommodate
thorough analysis, it has been concluded that the energy up to 1000 nodes. Furthermore, security measures will be
consumption rate linked to the proposed methodology is enhanced to prevent any malicious packets or assaults dur-
significantly lower than the already utilized approaches. ing transmission.
Selecting the most efficient route is crucial for reducing
Funding No Funding.
energy use. When a node demonstrates slow packet for-
warding, as is frequently observed with malicious nodes, Data and materials availability The datasets generated during and/or
it leads to higher energy consumption. However, our anal- analysed during the current study are available from the corresponding
ysis chooses the path based on dependable nodes, as author on reasonable request.
indicated by Eq. (8).
Declarations
Ks
K s (Routi ) = (8) Conflict of interest We certify that there is no actual or potential Conflict
∀nodes of interest in relation to this article.

Integrating the concept of clustering helps reduce energy Ethical approval I have approved there is no Conflict of interest for this
use even more. The recommended work’s minimal study.
energy consumption results in a significant improve-
ment in the network’s durability. Overall, the simulations
demonstrate that the D&SMR algorithm outperforms References
the active trust, reliable routing, and optimal routing
approaches to a substantial degree. Furthermore, it is 1. Ghamari, M., Janko, B., Sherratt, R. S., Harwin, W., Piechockic, R.,
D&SMR clearly outperforms active trust, reliable rout- & Soltanpur, C. (2016). A survey on wireless body area networks
ing, and optimal routing in terms of energy usage, with for ehealthcare systems in residential environments. Sensors, 16(6),
831.
average improvements of more than 19.36%, 28.02%,
2. Crosby, G. V., Ghosh, T., Murimi, R., & Chin, C. A. (2012). Wire-
30.07%, 31.97%, 40.2%, 52.58%, and 62.35%, respec- less body area networks for healthcare: A survey. International
tively, during the whole simulation period. Journal of Ad Hoc, Sensor & Ubiquitous Computing, 3(3), 1.

123
782 R. S. Abujassar

3. Arafat, M. Y., & Moh, S. (2022). JRCS: Joint routing and charg- 22. Kardaş, S., Celik, S., Arslan, A., & Levi, A. (2013). An efficient and
ing strategy for logistics drones. IEEE Internet of Things Journal, private RFID authentication protocol supporting ownership trans-
9(21), 21751–21764. fer. In Lightweight cryptography for security and privacy: Second
4. Mu, J., Wei, Y., Ma, H., & Li, Y. (2020). Spectrum allocation international workshop, LightSec 2013, Gebze, Turkey, May 6-7,
scheme for intelligent partition based on machine learning for 2013, Revised Selected Papers 2, Springer. pp. 130–141.
inter-WBAN interference. IEEE Wireless Communications, 27(5), 23. Saxena, D., & Patel, P. (2023). Energy-efficient clustering and
32–37. cooperative routing protocol for wireless body area networks
5. Al-Turjman, F. (2017). Energy-aware data delivery framework for (WBAN). Sādhanā, 48(2), 71.
safety-oriented mobile IoT. IEEE Sensors Journal, 18(1), 470–478. 24. Lalitha, S., Sundararajan, M., & Karthik, B. (2023). Reliable multi-
6. Aledhari, M., Razzak, R., Qolomany, B., Al-Fuqaha, A., & Saeed, path route selection strategy based on evidence theory for internet
F. (2022). Biomedical IoT: Enabling technologies, architectural of things enabled networks. Measurement: Sensors, 27, 100795.
elements, challenges, and future directions. IEEE Access, 10, 25. Raja Basha, A. (2022). A review on wireless sensor networks: Rout-
31306–31339. ing. Wireless Personal Communications, 125(1), 897–937.
7. Dian, F. J., Vahidnia, R., & Rahmati, A. (2020). Wearables and the 26. Haseeb, K., Saba, T., Rehman, A., Ahmed, Z., Song, H. H., &
internet of things (IoT), applications, opportunities, and challenges: Wang, H. H. (2022). Trust management with fault-tolerant super-
A survey. IEEE Access, 8, 69200–69211. vised routing for smart cities using internet of things. IEEE Internet
8. Barakah, D.M., & Ammad-Uddin, M (2012). A survey of chal- of Things Journal, 9(22), 22608–22617.
lenges and applications of wireless body area network (WBAN) 27. Chandnani, N., & Khairnar, C. N. (2022). An analysis of archi-
and role of a virtual doctor server in existing architecture. In 2012 tecture, framework, security and challenging aspects for data
Third international conference on intelligent systems modelling aggregation and routing techniques in IoT WSNs. Theoretical
and simulation, IEEE. pp. 214–219. Computer Science, 929, 95–113.
9. Qu, Y., Zheng, G., Ma, H., Wang, X., Ji, B., & Wu, H. (2019). A 28. Reddy Yeruva, A., Saleh Alomari, E., Rashmi, S., Shrivastava,
survey of routing protocols in WBAN for healthcare applications. A., Kathiravan, M., & Chaturvedi, A. (2023). A secure machine
Sensors, 19(7), 1638. learning-based optimal routing in ad hoc networks for classifying
10. Li, S., Kim, J. G., Han, D. H., & Lee, K. S. (2019). A survey of and predicting vulnerabilities. Cybernetics and Systems, 1–12.
energy-efficient communication protocols with QoS guarantees in 29. Vellela, S. S.,& Balamanigandan, R. (2023). Optimized cluster-
wireless multimedia sensor networks. Sensors, 19(1), 199. ing routing framework to maintain the optimal energy status in the
11. Rani, S., Talwar, R., Malhotra, J., Ahmed, S. H., Sarkar, M., & WSN mobile cloud environment. Multimedia Tools and Applica-
Song, H. (2015). A novel scheme for an energy efficient internet of tions, 1–20.
things based on wireless sensor networks. Sensors, 15(11), 28603– 30. Kaythry, P., Kishore, R., & Avinash, E. (2024). Reliability based
28626. https://doi.org/10.3390/s151128603 multistage ARQ for wide area wireless sensor networks. Journal
12. Yadav, R. N., Misra, R., & Saini, D. (2018). Energy aware cluster of Engineering Science and Technology, 19(2), 374–389.
based routing protocol over distributed cognitive radio sensor net- 31. Papachary, B., Arya, R.,& Dappuri, B. (2024). Power-aware QoS-
work. Computer Communications, 129, 54–66. https://doi.org/10. centric strategy for ultra reliable low latency communication in 5G
1016/j.comcom.2018.07.020 beyond wireless networks. Cluster Computing, 1–14.
13. Cengiz, K., & Dag, T. (2017). Energy aware multi-hop routing 32. Dasari, R., & Venkatram, N. (2024). Optimizing multichannel path
protocol for WSNs. IEEE Access, 6, 2622–2633. scheduling in cognitive radio Ad HoC networks using differential
14. Xiao, K., Wang, R., Deng, H., Zhang, L., & Yang, C. (2019). evolution. Scalable Computing: Practice and Experience, 25(2),
Energy-aware scheduling for information fusion in wireless sensor 1199–1218.
network surveillance. Information Fusion, 48, 95–106. 33. Rocha, D., Teixeira, G., Vieira, E., Almeida, J., & Ferreira, J.
15. Altowaijri, S. M. (2022). Efficient next-hop selection in multi-hop (2023). A modular in-vehicle c-its architecture for sensor data col-
routing for IoT enabled wireless sensor networks. Future Internet, lection, vehicular communications and cloud connectivity. Sensors,
14(2), 35. 23(3), 1724.
16. Moussa, N., Hamidi-Alaoui, Z., & El Belrhiti El Alaoui, A. (2020). 34. Kaur, P., Kaur, K., Singh, K., Bharany, S., Almazyad, A. S.,
ECRP: An energy-aware cluster-based routing protocol for wire- Xiong, G., Mohamed, A. W., Shokouhifar, M., & Werner, F.
less sensor networks. Wireless Networks, 26, 2915–2928. (2023). Acoustic monitoring in underwater wireless sensor net-
17. Mehmood, A., Mauri, J. L., Noman, M., & Song, H. (2015). works using energy-efficient artificial fish swarm-based clustering
Improvement of the wireless sensor network lifetime using leach protocol (EAFSCP).
with vice-cluster head. Ad Hoc and Sensor Wireless Networks, 35. Wang, H., Li, Y., Zhang, Y., Huang, T., & Jiang, Y. (2023).
28(1–2), 1–17. Arithmetic optimization AOMDV routing protocol for FANETs.
18. Shagari, N. M., Idris, M. Y. I., Salleh, R. B., Ahmedy, I., Murtaza, Sensors, 23(17), 7550.
G., & Shehadeh, H. A. (2020). Heterogeneous energy and traf- 36. Manoharan, J. S. (2023). A metaheuristic approach towards
fic aware sleep-awake cluster-based routing protocol for wireless enhancement of network lifetime in wireless sensor networks. KSII
sensor network. IEEE Access, 8, 12232–12252. Transactions on Internet & Information Systems, 17(4).
19. Salunkhe, S. P.,& Patil, H. D. (2016). Delay efficient authenticated 37. Hadwa, S. M., Ghouraba, R. F., Kabbash, I. A., & El-Desouky,
anonymous secure routing for MANETs. International Journal of S. S. (2023). Assessment of clinical and radiographic efficiency
Computer Applications, 148(4). of manual and pediatric rotary file systems in primary root canal
20. Sbeiti, M., Goddemeier, N., Behnke, D., & Wietfeld, C. (2015). preparation: A randomized controlled clinical trial. BMC Oral
PASER: Secure and efficient routing approach for airborne mesh Health, 23(1), 687.
networks. IEEE Transactions on Wireless Communications, 15(3), 38. Ullah, S., Saleem, A., Hassan, N., Muhammad, G., Shin, J., Minhas,
1950–1964. Q. -A.,& Khan, M. K. (2023). Reliable and delay aware routing
21. Babbitt, T. A.,& Szymanski, B. K. (2016). Trust based secure protocol for underwater wireless sensor networks. IEEE Access.
routing in delay tolerant networks. In 2016 IEEE Conference on 39. Gopi, B., Ramesh, G., & Logeshwaran, J. (2022). The fuzzy log-
Computer Communications Workshops (INFOCOM WKSHPS), ical controller based energy storage and conservation model to
IEEE. pp. 542–547. achieve maximum energy efficiency in modern 5G communication.

123
A novel algorithm for the development of a multipath protocol... 783

ICTACT Journal on Communication Technology, 13(3), 2774– Radwan S. Abujassar currently

2779. holds the position of Associate
40. Sahu, M., Sethi, N., & Das, S. K. (2022). Secure data transmission Professor at the Faculty of Infor-
in wireless sensor networks with secure system for identification mation Technology and Comput-
of trusted route with node behavior analysis. Revue d’Intelligence ing at the Arab Open University-
Artificielle, 36(2). Kuwait, an institution associated
41. Gupta, S. K., & Singh, S. (2022). Survey on energy efficient with the prestigious University of
dynamic sink optimum routing for wireless sensor network and OU in the United Kingdom. Dr.
communication technologies. International Journal of Communi- Radwan is currently the faculty’s
cation Systems, 35(11), 5194. local dean and a member of the
42. Abbas, G., Ullah, S., Waqas, M., Abbas, Z. H., & Bilal, M. (2022). AOU-Kuwait scientific committee.
A position-based reliable emergency message routing scheme for Dr. Radwan earned his bachelor’s
road safety in VANETs. Computer Networks, 213, 109097. degree in computer science from
43. Valle, M. S., Casabona, A., Sapienza, I., Laudani, L., Vagnini, A., Applied Science University in
Lanza, S., & Cioni, M. (2022). Use of a single wearable sensor to Amman, Jordan, in 2004, and then
evaluate the effects of gait and pelvis asymmetries on the compo- acquired an M.Sc. degree from the New York Institute of Technology
nents of the timed up and go test, in persons with unilateral lower in 2007. In 2012, Dr. Radwan earned a Ph.D. from the University of
limb amputation. Sensors, 22(1), 95. Essex, UK, with a specialization in IP recovery in IGP and MANET
networks in the field of computer and electrical studies. He is inter-
ested in studying network architecture and control systems, particu-
Publisher’s Note Springer Nature remains neutral with regard to juris- larly focusing on routing protocols and IoT.
dictional 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.

123