A Study to Enhance the Route Optimization Algorithm for the Internet of Vehicle

Abstract

In the Internet of Things (IoT), the advancement of the new modern era of Internet-off driving, known as the Internet of Vehicles (IoV), is aimed at the Intelligent Transportation System (ITS) to improve road safety and traffic. Therefore, IoV is better for traffic management. But perhaps, the main challenges in this type of network are that timely decision to make adequate decisions by a driver under certain conditions of rapid change in topology, high vehicular mobility, and frequent link failures is hard to improve road safety. Therefore, an optimized and congestion-free route is required to collect real-time data from vehicles. Thus, some of the latest bioinspired optimization routing algorithms in the IoV environment are presented in this study. Hence, monitoring speed limits, pollution checks, and emergency responses to traffic accidents should also be considered while performing vehicle routing to avoid traffic problems. In the previous decades, many route-directing protocols for the IoV environment have been proposed that can handle the requirements of reliability and security. But such routing of protocols suffers from high complexity and scalability limitations in big-scale networks, routing overlays, etc. Therefore, their strengths, weaknesses, and critical characteristics are compared using various criteria for such algorithms. Then, a suggestion is made to propose a prescribed combined model of a multimodular bioinspired approach to IoV routing. Finally, the main future directions of research in this sector are highlighted.

Full text

Review Article
A Study to Enhance the Route Optimization Algorithm for the
Internet of Vehicle
Ritesh Dhanare ,
1
Kapil Kumar Nagwanshi ,
1
and Sunita Varma
2
1
Department of CSE, Amity School of Engineering and Technology, Amity University Rajasthan, Jaipur, India
2
Department of Computer Engineering, Shri G. S. Institute of Technology and Science, Indore, India
Correspondence should be addressed to Kapil Kumar Nagwanshi; dr.kapil@ieee.org
Received 16 February 2022; Accepted 9 April 2022; Published 27 April 2022
Academic Editor: Nitish Pathak
Copyright © 2022 Ritesh Dhanare et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
In the Internet of Things (IoT), the advancement of the new modern era of Internet-offdriving, known as the Internet of Vehicles
(IoV), is aimed at the Intelligent Transportation System (ITS) to improve road safety and traffic. Therefore, IoV is better for traffic
management. But perhaps, the main challenges in this type of network are that timely decision to make adequate decisions by a
driver under certain conditions of rapid change in topology, high vehicular mobility, and frequent link failures is hard to improve
road safety. Therefore, an optimized and congestion-free route is required to collect real-time data from vehicles. Thus, some of
the latest bioinspired optimization routing algorithms in the IoV environment are presented in this study. Hence, monitoring
speed limits, pollution checks, and emergency responses to traffic accidents should also be considered while performing vehicle
routing to avoid traffic problems. In the previous decades, many route-directing protocols for the IoV environment have been
proposed that can handle the requirements of reliability and security. But such routing of protocols suffers from high
complexity and scalability limitations in big-scale networks, routing overlays, etc. Therefore, their strengths, weaknesses, and
critical characteristics are compared using various criteria for such algorithms. Then, a suggestion is made to propose a
prescribed combined model of a multimodular bioinspired approach to IoV routing. Finally, the main future directions of
research in this sector are highlighted.
1. Introduction
In recent years, there has been a growing interest in the
transport industry and academic researchers in improving
road safety by delivering fast and reliable information to
vehicle drivers and navigation entities [1, 2]. Integrating
road traffic data over mobile or wireless networks with the
minimum infrastructure offered by the Intelligent Transpor-
tation System (ITS) is one approach to enhancing road
safety. ITS provides a wide range of real-time applications
like real-time diversion route calculation; blind crossing;
trafficflow control; monitoring traffic; collision avoidance;
vehicle safety; finding the nearby restaurant, gas station, or
hotel; and automatic toll payment [3, 4].
Despite much progress in sustainable development traf-
fic, transmitting data to a broadcast medium like mobile gate-
ways has limitations like high asset prices and geographic
confines, particularly in rural areas such as deserts, moun-
tains, and islands. As a result, many telecommunication
services are severely limited in their ability to cover a wide
range of networks. Moreover, due to the speed of the traffic,
the connections offered by these basic network connections
are affected by the unreliable connectivity of the network
and are more prone to interruptions [5]. Furthermore,
infrastructure-based vehicle networks have a significant dis-
advantage in terms of the expense of deploying several fixed
devices and units on the street.
Because of these vehicular network restrictions, all cars
must participate in a multihop or unicast routing procedure
with each other as well as with fixed and stationary gateways.
Therefore, vehicular ad hoc networks (VANETs) have been
developed to provide traffic users with reliable network cov-
erage, extensive connectivity, and minimum communication
costs [6].
Hindawi
Wireless Communications and Mobile Computing
Volume 2022, Article ID 1453187, 20 pages
https://doi.org/10.1155/2022/1453187

Several surveys addressing various elements of routing in
vehicular networks have been suggested in previous
research. The majority of them are concentrated based on
the density, velocity, direction of motion, and so on, whereas
other categorizations of VANET routing protocols are based
on information dissemination mode [7]. In particular, in [8],
the authors categorized the protocols devoted to VANETs
based on V2I and V2V structures, and they also discussed
the future potential of considering vehicle density in the
routing algorithm. But, they ignore the hybrid construction
that employs both V2I, V2V, and V2O, whereas in [9], the
authors proposed a conceptual taxonomy based on various
factors of the unicast protocol. They first divide them into
two major categories. The route-dependent and route-
routing systems vary in data forwarding and path selection.
In addition, to highlight the advantages and shortcomings
of protocols, the authors proposed a conceptual taxonomy
based on fundamental functions such as relay selection,
recovery strategy, inference logic, probing objective, probing
scope, and routing techniques. In [10], a taxonomy of
VANET routing protocols is presented, with categorization
based on data fusion, opportunism, clustering, hybridity,
and geography.
The VANET routing protocols described in [11] are
divided into two primary categories: geographic-based rout-
ing and topology-based routing. In the paper presented here
[12], the authors are interested in position-based routing
methods for VANETs and examine position determination
issues. They present some relevant work on the advance-
ment of position assessment system solutions and list the
position-based protocols, along with their benefits and a
worldwide qualitative comparison of the protocols. From
this study, the authors classify routing protocols that employ
decision metrics. Furthermore, they provided privacy, secu-
rity, and delay in view.
The network topology of VANETs is continually and
rapidly changing, which creates several challenges and
unclear instructions like bandwidth scarcity, channel insta-
bility, and communication delays when the network is
implemented on a broad scale. While a greater number of
nodes create a more stable route, a too dense population
might induce network congestion and packet collisions. To
reduce delays and enhance the efficiency and dissemina-
tion distance, several techniques such as cluster-based
routing, geographic position, reactive, proactive, machine
learning, bioinspired, probabilistic, broadcast, and unicast
are utilized [13, 14].
However, due to a lack of processing capacity for global
information by VANET and vehicle telematics, there is a
need for an open-integrated network framework system.
So, nowadays, VANET, vehicle telematics, and other vehicle
networks are associated with IoV [15]. IoV is considered to
be a large network capable of providing services for large
cities and countries [16]. IoV is particularly an open and
incorporated network framework system that consists of
many more users, vehicles, network organizations, and
things in accommodation with controllability and high
management. The key goal of IoV is to interact with the
human-vehicle environments and things to limit social cost,
ensure human satisfaction, promote efficient transportation,
and improve the service levels in cities [17]. The basic model
of IoV is shown in Figure 1.
Advances in IoV technology can reduce the travel time
of passengers or drivers by finding a shorter route. There
are numerous routing protocols for a vehicle to find a better
route path. But some of the paths become invalid as slight
changes occur in their topology. To mention the prediction
of a path failure, before finding a new path, a novel estima-
tion of the path duration routing protocol is necessary to
enhance the performance of the network [18]. Several
classic-type routing protocols and the vehicle’s properties
by some geographic-type protocols that come from MANET
studies are used, such as Destination Sequenced Distance
Vector (DSDV), Dynamic Source Routing (DSR), Ad hoc
On-Demand Distance Vector (AODV), Greedy Perimeter
Stateless Routing (GPSR), and Greedy Perimeter Coordina-
tor Routing (GPCR). However, it should be noted that the
majority of routing algorithms are only suitable for a limited
scale [19]. For such issues, an optimization algorithm is the
best way to find the shortest path and the fastest conver-
gence speed. Many models inspired by nature have been
developed to solve optimization problems that arise in
different areas of life. Bioinspired approaches such as swarm
intelligence (SI), genetic algorithms (GA), and neural net-
works like artificial neural network (ANN) are used to solve
complex optimization problems [20, 21]. An evolutionary
system deals with the collective behavior of groups of
worms, birds, ants, and bees. Individuals in the group
develop traits based on their specific characteristics, i.e.,
where to feed and where to breed. Bees and ants use this
information to find new food sources. Cuckoo Search,
Bat Algorithm, firefly algorithm (FA), etc. are swarm intel-
ligence established methods utilized to solve complex real-
life issues [22–24]. Ant Colony Optimization (ACO)
makes extensive use of routing paths. The Quality of Ser-
vice (QoS) multicast problem is solved by the bee colony
routing optimization algorithm. All of evolutionary com-
puting has found a wide range of solutions to many
real-world problems that include graph colorization prob-
lems, vehicle routing problems (VRP), back-packing prob-
lems, and work-planning problems [25].
With the popularisation of nature-inspired optimization
algorithms over the last decade, our survey offers a solution
V2N V2V
V2I
V2P
Figure 1: Architecture of IoV.
2 Wireless Communications and Mobile Computing

to validate recent progress in the bioinspired VANET based
on emerging research issues that can be further expanded to
contribute to optimal routing in VANET.
The organization of this survey is structured in Figure 2.
In Section 2, we provide the existing review of IoV. The
preliminaries of review are explained in Section 3. The
taxonomy of optimization algorithms in IoV is discussed
in Section 4. We also discussed the challenges and applica-
tions of this survey there. Finally, at the end of the paper,
the conclusion to this review is provided in Section 6.
2. Literature Review
Numerous sorts of examinations have been conducted in the
field of IoV in ITS. From that, a small portion of the overall
challenges of finding the shortest route path to reach the
endpoint by the existing routing algorithms is clarified
underneath.
Hussain et al. [26] have reviewed the end-to-end service
in the IoV scenario and the quality service of IoV. The
agglomeration is critical in IoV for vehicle communication,
processing, integrating, and sensing. Due to the additional
devices connected and communicating, there will be a huge
vehicle load on the network. Hence, IoV needs accurate
end-to-end delivery without any compromise and with
QoS parameters such as latency, jitter, delay, and bandwidth.
In IoV, QoS is utilized to enhance the network service and
satisfy customer needs. The parameters of QoS provide ded-
icated bandwidth and reduce the characteristics of packet
loss, which are some of the main goals of IoV. In IoV, since
the vehicles are dynamic and move from one place to
another at a very high speed, the network requirements are
unpredictable. Yet, it is very difficult to manage and
IoV Study
Section1:
Introduction
Background
of IoV
Historical
evaluation of
optimization
algorithm
Section 2:
Literature
Review
Related work
of IoV
Section 3 :
preliminaries
Motivation
Signicant
contribution
Research
source
Section 4:
taxonomy of
optimization
techniques
Bio-inspired
algorithms
CI algorithm
Section 5:
challenges &
applications
Vehicle
Routing
mangemnet
Infotainment
and comfort
Safety
Section 6 :
conclusion
Figure 2: Organization of survey.
3Wireless Communications and Mobile Computing

integrate all these features into the network while providing
the IoV service, which is dynamic in nature. The main chal-
lenge in IoV is that when the vehicles move from one place
to another network, there could be packet loss when the
vehicle network is autonomous. The end-to-end services
are in collaboration with the IoV network. The vehicles are
treated as nodes and change dynamically with varying
conditions. From this review, they conclude that they can
develop the optimal solution using QoS parameters and
their measurements.
Cheng et al. [27] have surveyed the details of IoV-based
routing algorithms. For this purpose, the IoV uses a general
protocol for the horizontal networking process in order to
realise IoV. Initially, they categorize the transmission
scheme into three types, such as broadcast ones, unicast
ones, and geocast ones. In much of the work, they discussed
the routing protocols referring to the unicast category. The
application of this category is emergency vehicle preemption
and information about the road conditions. The balance
type of routing algorithms is used in the last category. Based
on these details, they classify the information into four cate-
gories: map, position, path-based one, and topology for the
communication process in vehicular networks. Yet, the
nature of the geographic nature of these protocols prevents
them from forecasting the topology holes in the distributing
nodes. Moreover, many routing protocols only work in
one- or two-dimensional scenarios and have the worst
performance in the real-time third-dimensional scenario.
They also discussed the reliability of the communication
between the vehicles. For this communication, a heteroge-
neous vehicular network process is needed. Therefore, the
researcher needs a set of well-defined routing protocols,
which are used to address the issues caused by the hetero-
geneous nature of IoV and large scale. Finally, they suggest
the researchers certify the studies in small-scale homoge-
neous networks. Furthermore, the large-scale heterogeneous
networks make the vehicles dependent on road information
after the social environment.
Kayarga and Kumar [28] presented the details of the IoV
network model, and they reviewed IoV technologies in dif-
ferent applications. Mostly, they concentrate on the bioin-
spired algorithms, which are used between vehicles, things,
and humans. For these different applications, the IoV con-
tains various steps to take the vehicles into the IoV inte-
grated network by using wireless access technology. To
maintain this technology, the IoV has aspects such as encod-
ing, virtual network, and data awareness that play the main
role in switching the control message to the IoV. With this
type of maintenance, each topology is also maintained in
intervehicle networks. The main routing topology is an ad
hoc network-based topology. This technology is more effec-
tive for wireless vehicle communication. It is also essential
for a credible and manageable IoV. The solution of IoV con-
siders vehicle communication in the complex city. Sophisti-
cated technology is utilized to develop the ability of IoV
communication by using a variety of technologies for IoV.
Hence, vehicular communication over big data can enhance
the network burden. Effective optimization techniques are
also needed to develop the service to enhance the suitability
of mobile nodes in the global network. Furthermore, many
challenges are focused on the VANET of IoV from multiple
vehicles, things, and users, because IoV goes beyond tele-
matics, intelligent transportation, and ad hoc networks by
integrating sensors, vehicles, and mobile devices to enable
different services.
Tuyisenge et al. [29] have surveyed the IoV on the mar-
ket opportunities in the field of city transport communica-
tion systems. It faces huge challenges and problems such as
traffic congestion, safety problems, pollution, and commer-
cialization. This is all forecasted, and it is evident that the
IoV component will create a huge amount of data and huge
real-time IoV applications and requires fast routing. They
also reviewed the existing protocols, namely, IEE, 3GPP,
and VANET, which are used for IoV vehicle communica-
tion, and further, they used protocol stack analysis. From
this survey, they finally provide some information about
the future work related to the mechanism of IoV between
various networks.
Ksouri et al. [30] conducted a review of suitable VANET
routing techniques and multioptimization algorithms. They
suggest new criteria for the routing process in the dynamic
environment. The survey shows that new innovations in
IoV applications and the presentation of autonomous vehi-
cles in networking provide new security and QoS challenges
in the VANET network. They classify the routing algorithms
based on the forward criteria to maintain the macro classifi-
cation. In the routing protocol, the macro classification is a
third family. It is named “hybrid routing,”which is a combi-
nation of both geographically based routing and topology-
based modeling. The function of this process represents the
merging of both the routing processes. Secondly, they dis-
cussed the geographical routing techniques. This VANET
routing model is capable of taking specific route materials
based on the vehicle’s environment. This model is more
suitable for large-scale vehicular communication network
processes and the performance of improving the node’s
security and safety. Then, they survey the forward process.
It allows the packets to move from one node to another. This
stage is an important core part of the IoV routing process. It
has two substages, such as the forward selection stage and
the packet relaying stage. Then, they review the optimization
techniques used for routing to enhance the overall routing
process, such as relay, path discovery, cluster head selection,
and routing FR. The routing optimization algorithm is more
effective than the IoV routing algorithm. Finally, they dis-
cussed the large-scale IoV routing techniques, including
the marine domains, terrestrial and aerial, for future study.
Many traditional optimization algorithms for traffic
management in IoV to optimize deterministic problems
have been proposed. But those algorithms have not shown
their ability to tackle the inherent randomness in traffic
systems. Therefore, to handle such efficient and random
situations efficiently, a better bioinspired optimization algo-
rithm is needed, because bioinspired optimization algorithm
has an ability to combine biological and natural characteris-
tics that can be easily adapted to frequently changing envi-
ronments. Table 1 provides the comparative analysis with
an existing survey on vehicular routing techniques.
4 Wireless Communications and Mobile Computing

3. Preliminary of the Survey
This section discusses the motivation for conducting this
review, as well as the contribution of the primary survey.
The main motivation of the review is to show the following
issues and challenges present in performing a vehicle routing
algorithm in the IoV: to achieve that, some sources of papers
are selected by using some string words to select papers
based on our work.
3.1. Motivation of the Study. The main motivation for this
study is as follows:
(i) To find an optimal path between the network nodes
to find the shortest routes, thus reducing the travel
time
(ii) To gain real-time road traffic information for pro-
viding optimal routing to improve travel comfort
for drivers
(iii) Hence, the use of bioinspired optimization routing
algorithms for the IoV environment boosts the
robustness of the IoV network while performing
optimal path routing at the time of network disrup-
tions. The chosen path should satisfy the routing
constraints such as packet delivery ratio, distance,
congestion, connection probability, cost, transmis-
sion delay, and variance introduced by the algorithm
3.2. Contribution. The main contribution of this review is to
recognize the sorts of issues that occur when trying to find
an optimized path to travel from a source to a destination,
and that chosen path should satisfy the following needs to
mitigate them: packet delivery ratio, distance, congestion,
connection probability, cost, transmission delay, and vari-
ance. The main purpose of IoV routing is to afford a correct
path of routing direction by a network in avoidance of traffic
to their final destination. So the contribution of the review
falls within a study of the IoV routing protocol in an extreme
or complex urban environment. This review paper addresses
the need for bioinspired optimization routing algorithms in
IoV, which can be able to tolerate low-/high-density traffic
networks with little throughput and delay variation.
3.3. Search Criteria. The papers collected are manually
searched, particularly journal papers and conference pro-
ceedings since 2001.
3.3.1. Selection of Sources. For the systematic review of exist-
ing issues, we searched the Springer, IEEE, Elsevier, and
Wiley Online Library databases using the following search
string words: IoV, route optimization, challenges and issues,
VANET, and machine learning. The search engine selection
for this survey is listed in Table 2, and the search strategy is
illustrated in Figure 3.
3.3.2. Exclusion Criteria. The articles are excluded on the
following basis:
(i) Publication not related to IoV routing
(ii) If the entire content of the paper is not available
Table 1: Comparative analysis with existing survey.
Author Year Taxonomy Application Findings
Hussain et al. [26] 2019 QoS-based routing Multimedia application
It is critical to provide best end-to-end services
with networks because the cars treated as nodes
change constantly in response to changing
network circumstances. As a result, there is a need
to design efficient methods for dealing with
mobile nodes, such as automobiles.
Cheng et al. [27] 2015
Topology-based, position-
based, map-based, and,
path-based routing
Road condition, accident
warnings, and
infotainment applications
To make IoV function in the real world,
the authors recommend that researchers test
their findings not just in small-scale
omogeneous systems but also in large-scale
heterogeneous networks.
Kayarga and
Kumar [28] 2021 Genetic algorithm
Traffic message
management,
routing decision
The hybridization of machine learning and swarm
intelligence gives better routing efficiency.
Tuyisenge et al. [29] 2018 Architecture-based
routing
—
It is projected and obvious that IoV components
would generate large amounts of data at high
rates, extremely sensitive to delay, and demand
rapid big data processing and dependable fast
response. As a result, efficient and dependable
network designs that support efficient IoV
installation in real-time are required.
Ksouri et al. [30] 2020
Geographical routing,
optimization-based
routing
Prediction-based
applications
Information interchange across the various IoV
domains like marine, aerial, and terrestrial should
be facilitated to allow for greater collaboration and
the adoption of innovative solutions.
5Wireless Communications and Mobile Computing

(iii) The papers published before 2001
(iv) Same papers available in different journals
(v) Papers not written in English
4. Taxonomy of Optimization
Techniques of IoV
This section holds the taxonomy of overall bioinspired
optimization techniques and solutions used to overcome
the issues in IoV routing. The overall taxonomy of IoV opti-
mization techniques is illustrated in Figure 4.
4.1. Taxonomy of Bioinspired Optimization Algorithms in
IoV. There are some optimization algorithms to handle such
problems, where the correct path is determined by the actual
solution rather than the best solution. Nature is such an
inspiring, immense set of solutions that deal with all real-
time problems. This has motivated many researchers to use
nature-inspired algorithms to solve such complex-related
issues. Therefore, algorithms inspired by nature are also
called “bioinspired algorithms”[31]. Such algorithms are
nowadays used to optimize the routing problem in network-
ing. Hence, these algorithms are considered to find the
global optimal solution. Bioinspired route optimization algo-
rithms carry an assortment of optimization algorithms in
terms of using biological principles to cover a wide range
of networks. Such types of algorithms include Ant Colony
Optimization (ACO), Bat Algorithm (BA), Cuckoo Search
(CS), Particle Swarm Optimization (PSO), bee colony opti-
mization (BCO), firefly algorithms (FA), Flower Pollination
Algorithm (FPA), and the Krill Herd (KH) algorithm.
Nowadays, a variety of new classes of algorithms are
applied, such as genetic algorithms (GA), evolutionary
algorithms (EA), neural networks (NN), and simulated
annealing (SA) [32].
The following types of optimization methods are used
for solving NP-hard and NP-complete optimization prob-
lems: the hard optimization approach presents a huge num-
ber of candidate solutions that can attain a global optimal
solution at a reasonable time. Hence, finding an optimal
solution by bioinspired means in the case of a substantial
solution is well known.
The bioinspired algorithms are extra flexible in operating
larger-scale vehicular ad hoc networks because of some
similarity in the manner of finding routes to satisfy their
natural needs. As a result, one of its main advantages is that
it generates less complexity, making it suitable for use in
computational problems [33]. Also, it is improved to main-
tain a satisfactory routing performance over time despite
network disruptions [29]. As a result, bioinspired route opti-
mization techniques are bringing in more essentials in the
IoV. The classification and subclassification of algorithms
are shown in Figure 3. Accordingly, algorithms are classified
into three foremost categories: Swarm Intelligence Algo-
rithms (SIA), EA, and other biologically inspired algorithms.
Their detailed explanations are discussed below.
The first category of routing algorithms is motivated
by swarm intelligence like birds, bees, and ants, and the
three subclasses of ACO, BCO, and PSO are explained.
Most of the VANET routing protocols use ACO routing,
inspired by the behavior of ants while searching for the
shortest route to reach the destination. This is done by a
chemical substance called pheromone. Ants communicate
with each other via pheromone trails. In [33], they intro-
duced a mobility-aware ACO routing protocol in VANET
to find optimal routes between source and destination
nodes while estimating each vehicle’sflexibility in terms
of spot and speed [34].
The second category of the EA is inspired by the compu-
tational natural behavior of mutation, crossover, and inher-
itance, and aimed at finding the shortest route between the
transmitter and the receiver. From this, two subclasses are
carried out: Parallel Genetic Algorithms (PGA) and Sequen-
tial Genetic Algorithms (SGA). The SGA is aimed at evalu-
ating only one objective taken into consideration by the
geographical routing protocol based on intersection [31] to
increase the possibility of connection among vehicles and
the Internet. PGA optimizes multiobjective transmission
benchmarks such as power consumption and by performing
automatic configuration using the routing protocol of Opti-
mized Link State Routing (OLSR) to VANET. Evolution for
bioinspired-based vehicular routing algorithms is shown in
Figure 5.
4.1.1. Performance of Swarm Intelligence-Based Bioinspired
Algorithms in IoV for Vehicle Routing. The purpose of
bioinspired optimization algorithms is to exist in different
applications. This subsection is primarily concerned with
investigating its application in IoV at vehicle routing as
those discussed below.
(1) Grey Wolf Optimization Algorithm (GWO). The aim of
GWO is to follow the natural predation behavior of grey
wolves, such as hunting their food in a cooperative way.
GWO is based on the clustering algorithm in IoV. It copy-
cats fully the social and hunting performance of grey wolves.
Clustering is done in VANETs and IoV for certain features
such as vehicle location, speed, position, and direction. A
GWO-based clustering algorithm for VANETs is imple-
mented [35]. In this approach, the social behavior of the
wolves is pictured into four levels: alpha, beta, gamma, and
omega. An alpha is such a dimension that it has a position
of basic leadership in the pack. Beta, the subordinate wolf’s
terms, as the second dimension, helps in settling a choice
for alpha. Delta is the subordinate wolf’s term for the
Table 2: Selection of search engines for this study.
Search engine Source address
IEEE Xplore http://ieeexplore.ieee.org/
Springer http://link.springer.com/
Elsevier http://www.elsevier.com/
Wiley Online Library http://www.WileyOnlineLibrary.com/
Hindawi https://www.hindawi.com/
ACM https://dl.acm.org/
6 Wireless Communications and Mobile Computing

third dimension used for classification. Omega is the most
minimal dimension which helps the grey wolves remem-
ber the prey position capacity. Then, hunting for prob-
lems over the network is converted into a mathematical
model to easily understand the behavior of the algorithm.
From this concept, an appropriate form of cluster is
extracted to optimize the mentioned problem. Finally,
the alpha, beta, and gamma direct the hunting while
their position ought to be by the omega wolves by three
best-arrangement considerations [36].
Articles
Included n = 107 Excluded n = 94
Exclusion criteria
Publication not
related to IoV
routing.
Papers not written
in English
If the entire content
of the paper is not
available.
e papers
published before
2001.
Same papers that
are available in
different journals.
Search source
IEEE xplore Springer Elsevier Wiley online
library Hindawi ACM
Search criteria: 2001-2022
Keyword: IoV
Route
optimization,
challenges
Issues VANET Machine
learning n = 8180
Database
Google scholar
Figure 3: Search strategy.
7Wireless Communications and Mobile Computing

(2) Honey Bee Algorithm (HBA). HBA admits to being inter-
ested in the nature of bees. In a period of looking for food,
the bees leave their hives. When the food is situated at the
hour of voyaging, it passes on the message to different honey
bees in a type of waggle movement. This cycle of passing on
is considered a neighborhood (shifty) and worldwide
(explorative) look. Researchers in [37] have developed a
HBA-propelled QoS directing convention for VANETs and
taught about the value transmission of data among convey-
ing hubs and vehicles to suggest honey bee’s versatility.
There are four unique types of honey bee-roused calcula-
tions in tackling vehicle directing issues [38]. Such calcula-
tions are isolated into two subclasses dependent on honey
bee conduct and are depicted as follows. Among them, one
is the Artificial Bee Colony (ABC) calculation, and later,
some adjustments have been made by Basturk and Karaboga
in order to tackle numeric capacity streamlining issues. The
ABC model derives full motivation for planning from the
scavenging conduct of bees. The first ABC to tackle vehicle
steering issues was finished by Banharnsakun, who picked
the Traveling Salesman Problem as the fundamental issue.
Then, at that point, Karaboga utilized the ABC approach
as a universally useful streamlining model to take care of
the vehicle directing issue [39–42]. The accompanying
ABC calculation works by utilizing three sorts of specialists
(honey bees, for example, are passerby honey bees, utilized
honey bees, and scout honey bees). Each one of them is
utilized for explicit stages in the calculation. Next, honey
bee-roused calculation of BCO is proposed by Nakrani and
Tovey. To take care of a way portion issue in the
Taxonomy of IoV optimization
algorithm
Bio-inspired algorithms
(31-34)
Computational intelligence
(76)
Swarm
intelligence
(34)
Evaluation
optimization
(57)
Bacterial foraging
optimization
(72-73)
Ant colony algorithm
(54-56)
Firey optimization
algorithm
(47-53)
Bee optimization algorithm
(37-46)
Grey wolf algorithm
(35-36)
Genetic algorithm
(58-69)
Genetic
programming
(70-71)
Computing system of
microbial interaction and
communication (COSMIC)
(74)
Rule based bacterial
modelling (RUBAM)
(75)
ANN
(77-78)
Fuzzy
(79-81)
Machine learning
(82-83)
Supervised
(82)
Unsupervised
(60)
SVM
(84)
Neural network
(85)
Naïve bayesian
(87) (88)
HMM
(89)
Figure 4: Taxonomy of various vehicle routing optimization algorithm used in IoV (give the complete types for the taxonomy).
8 Wireless Communications and Mobile Computing

organization and a little alteration made in the first calcula-
tion to tackle the directing issue by Wong and Zhou later,
the Marriage in Honeybee Optimization (MHBO) model is
presented by Abbass. The MHBO conveys four types of spe-
cialists at performing enhancement, including sovereigns,
robots, broods, and workers [43–46].
(3) Firefly Algorithm. Fireflies are otherwise called lightning
bugs. The function of the firefly approach depends on the
conduct of fireflies. Fireflies act as a sign, and they deliver
short and cadenced blazes to speak with different fireflies.
The protocol for each group is based on the routing in IoV
and was planned by researchers in [47] dependent on the
firefly model. The firefly approach is used for spreading
messages among vehicles during a crisis, so reasonable con-
nections are chosen for quick message dissemination.
The FA is a BI algorithm dependent on the social mating
conduct of fireflies [48]. This approach has a place in the
class of multitudes of knowledge procedures that depend
on the bioluminescence blazing conduct of fireflies, which
goes about as a flagging framework to draw in different fire-
flies created by Yang [49]. With this approach, each of the
fireflies streaks a certain level of brilliance. This makes other
fireflies adjoin, and each of their fascinations is impacted by
their distance [50]. The two fireflies that are near one
another are considered to have a higher appreciation of
one another. Each firefly represents having a certain point
in a pursuit space, and therefore, the target work is indicated
by the engaging quality level of every firefly. The real natural
behavior of each firefly needs to be moved towards its neigh-
bors by its most noteworthy fascination. One of the funda-
mental parts of FA is the distinction of light force, and the
other is by means of their drawing quality. From this view,
the fascination of every firefly is estimated by its light
intensity [51–53].
(4) Ant Colony Optimization (ACO). This kind of ACO algo-
rithm is a type of bioinspired algorithm, used to mimic the
natural behavior of ants. An ant can find a small path from
the source location to the food location by discharging a
2020
2018
2016
2014
2012
Genetic algorithm
2010
Ant colony algorithm
Firey algorithm
Evolutionary algorithm
Swarm intelligence algorithm
Bacterial foraging
algorithm
Honey bee algorithm
Computational intelligence
2022
Genetic programming
Grey wolf algorithm
Supervised learning
Unsupervised learning
Fuzzy
Articial neural network
Firey+levy distribution
Advanced greedy algorithm
Modied ACO+rey algorithm
Figure 5: Evolution of bioinspired and CI algorithms in vehicular routing.
9Wireless Communications and Mobile Computing

pheromone along the path. The ACO technique for optimiza-
tion can be used to improve vehicle routing problems in IoV.
The function of ACO is based on the routing algorithm
efficiency for VANET and IoV. It depends on the nature of
ants. Their behavior is based on evaporation and pheromone
redeposition. Ants are not interested in traveling along sim-
ilar paths more than once, so circular dependence between
vehicles is developed. The main benefit of ACO is that it
can reduce the overall delay and also increase the packet
delivery.
Kumar et al. [54] have proposed the IoV-based traffic
management method, which is used to prevent traffic con-
gestion and accidents. Here, they separate the street map
into small submaps. Initially, they used the ACO algorithm
for each submap to find the optimal route path. The optimal
solution is based on the rod length and moving vehicles with
the help of IoV technology. In ACO, the forward ant’s pro-
cess is utilized to find the optimal solution. Its functional
equation is given below:
where ∂IJ denotes the pheromone value of ant in nodes I
and J,ηIJ denotes the instantaneous fuzzy function, A
denotes the weight value of ∂IJ,Bis the weight value of
ηIJ,nJdenotes the number of the nearest node, TABUK
denotes the node set not visited by the ant. Then, the
fuzzy function is designed to calculate the intensity of
the vehicle in heavy traffic. The experimental result of this
method is compared with some existing methods.
Nguyen et al. [55] have proposed addressing the dynamic
IoV traffic routing issues with multisource and multidestina-
tion in an IoV environment. This proposed model is
designed to solve the vehicle traffic in the IoV. Initially, they
proposed the decentralized ACO routing model to connect
the vehicles. By using this algorithm, the coloured ants are
utilized for trafficflow between the sources and destination.
This ACO algorithm design is used to develop communica-
tion and is also proposed to exchange data among infrastruc-
ture and connected vehicles. The functional equation of ACO
is given below:
ΔυλD
tIJ T
ðÞ
=1
lυλD
IJ T
ðÞ
+1
kυλD
IJ T
ðÞ
+1
dυλD
IJ T
ðÞ
,ð2Þ
where lυλD
IJ ðTÞis the length of each edge, kυλD
IJ ðTÞis the vehicle
traveling time, and DυλD
IJ ðTÞrepresents the density of vehi-
cles. The simulation result of this proposed method was per-
formed with a multi-intersection scenario in the NetLogo
platform. This result shows that the ACO algorithm is better
for the routing process than the non-ACO algorithms.
In the IoV model, Jabri et al. [56] present multiaccess
edge-based vehicular fog computing. The goal of this model
is to exploit both the infrastructure equipment and vehicles
to bring the users close to the cloud. The function of this
proposed model has two main steps. Firstly, the Multiaccess
Edge Computing (MEC) technology is used to provide
centralized control of the vehicular fog. They used various
modules in this architecture. But here, they particularly
focus on the selection gateway module. Then, in the context
of the vehicular cloud, they access the MEC server. It will
allow all vehicles to connect directly to the cloud, increasing
radio resources and increasing traffic congestion and colli-
sion rates. Initially, they utilized fuzzy logic for the gateway
selection process based on some parameters. The values of
each parameter are collected through the IoV communica-
tion between vehicles. Then, they use the ACO algorithm
to solve the gateway selection with the uncovered vehicles.
Therefore, the issues are divided into two steps: first, the
connection of the covered node to the gateways and sec-
ondly, connecting the uncovered nodes to the covered node.
This proposed model provides an important node ratio, par-
ticularly for static vehicles. The functional comparison of
SIA is discussed in Table 3.
(5) Hybrid Algorithm. To address the multicast routing
problem, a firefly with Levy distribution (FF-L) technique
was proposed by [57]. The Levy distribution is used in the
FF algorithm to prevent it from being trapped in local
optima. A set of tests on three distinct scenarios is used to
validate the FF-L algorithm in VANET. This method
showed less cost, jitter, and delay compared to other
methods. To improve the performance of IoV, a bioinspired
advanced greedy hybrid routing protocol was proposed in
[58], where a bee colony optimization combined with a
greedy algorithm is used to choose the best route with high
service quality and select a route with the minimum over-
flow. The simulation outcomes show that this protocol
works well in both V2I and V2V environments and has a
significant impact on enhancing delay and packet delivery
ratio while maintaining a minimum hop count across all
vehicles and reasonable overhead. However, various chal-
lenges exist, such as determining the quickest path between
the source and the final destination, excessive delays, poor
connection, congestion, and low packet delivery ratios. To
overcome these issues, the work presented in [59] provides
a hybrid optimization strategy that combines firefly optimi-
zation and modified ant colony to determine the average
K
IJ T
ðÞ
=A∂IJ

+B1−ηIJ

∑H∉TABU KA∂IJ

+B1−ηIJ

×1
1+ 1/nJ

!
,
(if J∉TABU K,
0, otherwise,
ð1Þ
10 Wireless Communications and Mobile Computing

speed and discover the optimum path to the destination.
Here, the algorithm used pheromones and attractiveness to
select the best path and cut trip time.
4.1.2. Performance of Evolutionary Algorithm-Based
Bioinspired Algorithms in IoV for Vehicle Routing. In the
taxonomy of bioinspired algorithms, the evolutionary algo-
rithm is the first type of approach. It is utilized to solve prob-
lems in the diverse domain of science and real-time
applications. It is used to obtain an exact optimal solution
for a multimodal function. The function of evolutionary
algorithms is loosely based on the metaphors of biological
processes. The function of EA is based on a random search
approach with some metaheuristics. This technique consists
of population-based functions; it represents the search point
and feasible solution, and it is unprotected from the collec-
tive learning process, which is performed from one genera-
tion to the next. The EA population is randomly initialised
and subjected to selection; then, the recombination process
and mutation process will be performed over several gener-
ations [60]. The classification techniques of EA are discussed
below.
(1) Genetic Algorithm. Because the vehicle routing problem
(VRP) is a complex NP-hard problem, it has attracted the
attention of many scholars from various fields. Using a
traditional method is unsatisfactory for VRP to get a global
optimal solution. Hence, a global method based on genetic
theory and selection theory of genetic algorithms has been
introduced. GA is a natural selection-based process that
carries out three operations: selection, crossover, and muta-
tion [61]. Therefore, the researchers presented a genetic-
based routing protocol for IoV to forward the data with
the least amount of rate from source to destination vehicle.
But GA itself has some problems, such as slow convergence
and premature termination. Hence, improving its conver-
gence speed and, at the same time, maintaining population
diversity in solving VRP are considered a main goal
[62, 63]. Therefore, they proposed an improved GA to
solve VRP by including SA into the GA in terms to avoid pre-
mature and convergence issues. Zirour introduced a GA to
solve the vehicle routing problem with time windows
(VRPTW) to reduce the aggregate distance and delay time
of each customer’s vehicle path under time violation, whereas
utilizing GA algorithms to solve such a problem includes two
steps. First, they use the initial solution to reduce the
expenses in two steps, and then, the expenses in the next step
will fail. Hence, this research proved the solutions that are
close to the best solutions. Then, GA with a crossover opera-
tor was used to solve the VRPTW problem. Then, the same
improved GA-based crossover operator algorithm was used
[64] that was used there. The crossover operator chooses an
individual from the whole population to generate offspring.
Rather than using a traditional method of presenting off-
spring, the swap node operator has been used for random
selection. After testing with other algorithms, the improved
GA algorithm presented the best optimal solution and illus-
trated that their outcomes were competitive. There have been
multiple objective functions in VRPTW. From the following
objective function, some of the common considerations of
the objective function used for minimization are (i) total
travel distance of the vehicles, (ii) total number of vehicles,
(iii) total travel time, and (iv) maximum QoS. In [65], we
chose five objective functions for multiobjective routing
problems, such as total delay time, total waiting time, total
travel distance, and makespan. To deal with such problems,
nondominated sorting GA (NSGA-II) will be introduced
to deal with vehicle routing problems (VRP). Then,
Zhong-yue et al. and Castro-Gutierrez [66, 67] applied
such an optimization algorithm in VRP with two kinds
of demand, focusing on the number of vehicles and make-
span. The optimization process considers the similarity set
of nondominated solutions. Later, Jozefowiez proposed an
improved version of NSGA-II optimization by including
two enhancement processes: parallelization and elitist
diversification. This type of approach is applied in VRP
problems to tackle where to limit the travel distance
between the shortest and longest routes. Next, Wei passed
this algorithm in a real-world test case in the USA, con-
sidering three objectives: distribution time, travel distance,
and set of vehicles. Bullinaria and Garcia-Najera used
NSGA-II performance, which is related to another algo-
rithm, an MOEA that includes two of the same measures:
edit distance and Jaccard parallel coefficient [68–70]. The
Table 3: SI-based bioinspired algorithm advantage, simulation tool, evaluation parameter, and application.
Swarm intelligence-based
bioinspired optimization
algorithm
Advantage Simulation
tool Evaluation parameter Application
ACO It provides highly optimized routing and is
fault-tolerant NS2 Packet delivery ratio, loss,
and throughput Routing
FA The delivery ratio of each packet proposed
protocol is quite improved NS2 End-to-end delay and
propagation delay Routing
HBA Outperformance of this approach is better
than the protocols like AODV and DSR NCTUns6.0 Throughput and packet loss Routing
GWO The outcome of this model illustrates that
performance of cluster heads is good MATLAB Transmission range, grid
size, and number of nodes Routing
Hybrid Selects the best path and cut trip time MATLAB Delay, packet delivery ratio,
connectivity Routing
11Wireless Communications and Mobile Computing

performance was assessed using the travel distance and
number of vehicles. A GA-based parallelization method
for the Traveling Salesman Problem (TSP) in VRP is
presented. The approach has been used to solve time-
constrained issues of IoV routing and autonomous vehi-
cle control in IoV [71, 72].
(2) Genetic Programming (GP). GP is a populace-based
inquiry calculation propelled by normal development. It
begins with creating a populace of people (as a rule, indis-
criminately) who are IoV for the target issue [73]. Then, at
that point, every individual is assessed and doled out with
a wellness score that demonstrates how well this up-and-
comer tackles or verges on taking care of the current issue.
Until an end standard is fulfilled, new populations are cre-
ated iteratively by utilizing determination, hybrid, and trans-
formation administrators, as in normal advancement. These
hereditary administrators are used to provide better services
to the new population. The vast majority of bioadvanced cal-
culations used in impromptu organizations are classified
into two types: centralized with global (online) information
and decentralized with local (offline) information. The exec-
utives of every vehicle handled by VANETs assess the trust
esteem dependent on the neighborhood data while continu-
ing on with the organization. The most trusted board models
are fundamentally used in impromptu organizations to
ensure secure and solid correspondence. As a result, the
GP for introducing trust in the executive model in VANET
deals with the powerful changing geography and events [74].
4.1.3. Performance of Bacterial Foraging-Based Bioinspired
Algorithms in IoV for Vehicle Routing. Bacterial foraging
optimization (BFO) utilizes the regular scavenging proce-
dures of Escherichia coli bacterium cells and makes an
essential move to augment the energy used per unit of time
spent on searching. From that point forward, BFO has been
applied effectively to multitarget issues. A versatile BFO is
introduced for addressing VRPTW. The VRPTW is defined
as a collection of vehicles used to collect the user’s time win-
dows [75]. It starts at the warehouse and ends at the station.
Thus, every one of the clients ought to be served precisely
once. In the event that vehicles show up before the time
window “opens”or after the time window “closes,”there will
be hanging tight for costs and late expenses to limit the total
expense of fulfilling all correspondence and disparity
requirements.
The objective function of total cost is taken into consid-
eration. The variables have various pieces of information,
such as distribution strategist, vehicle routing, and the
arrival time for every customer. The dimension of each
variable is based on the vehicles [76].
(1) Computing System of Microbial Interactions and Com-
munication (COSMIC). COSMIC was created in 2004 by
Saunders et al. The guidelines for this model are based on
the processing framework. It was mainly introduced for the
bacterial simulation process for COSMIC-Rules by Saunders
et al. in the year of 2006. The models of the following devel-
opment changes occur in cells while utilizing discrete time
waves and a pseudoconstant space model. Each of the cells
contains genomes that are associated with real time, etc.
COSMIC contains three remarkable levels of stages, includ-
ing the genome, the cell, and the climate occupied by cells.
The fundamental aim is to contemplate the advancement
or variation of microscopic organisms and to foresee the
conduct of pathogenic microorganisms [77].
(2) Rule-Based Bacterial Modeling (RUBAM). RUBAM was
designed by [78]. The function of RUBAM is a combination
of GA and GP algorithms. The function of RUBAM is based
on the bacteria. It acts as a data unit process which maps
action and message. The data from the real-time applica-
tions is mapped to take them again. This process is not static
because it makes it possible to continuously change the envi-
ronment, ensure real life, and also enhance the reproduction
process. The main goal of this model is to grasp the bacteria
and their behavior and features. The model can balance the
computational demands and complexity. Moreover, it has a
set of interconnection mechanisms and a collection of artifi-
cial organisms that also set the operators for the evolution
process. The overall knowledge contains artificial organisms.
In multimodal search space, this model can get optimal
results. But the other optimization algorithms can interact
with the real-time environment applications and also calcu-
late the fitness value. The best outcomes are shared with var-
ious processes [79]. The feature comparison of bioinspired
algorithms is illustrated in Table 4.
4.2. Taxonomy of Computational Intelligence-Based IOV
Algorithms. Computational intelligence (CI) is based on arti-
ficial intelligence and computer systems. The methodology
of CI is a biooriented computation model. It is used for some
real-time applications in dynamic and static environments
in which mathematical or traditional modeling and reason-
ing are too complex and are uncertain, and the process will
be stochastic in nature [80]. The CI process is controlled in
a modified way. The main principles of the CI technique
are listed in Figure 3.
4.2.1. Artificial Neural Network (ANN). ANN is a CI-based
technique; its functions are based on biological networks.
Data flows through the affected ANNs as a neural network
(NN) changes—or learns in a certain way—based on the
input and output. ANN is a framework to design a biologi-
cally inspired algorithm. Various kinds of networks have
been used, including self-organizing networks, single-layer
networks, recurrent networks, and multilayer networks. It
is effectively applied in many fields to find vehicular com-
munication details such as predictions, pattern classifica-
tions, traffic density estimations, and robotics. In addition,
this is one of the most effective methods for nonlinear,
uncertain, and time-based models. However, it is also uti-
lized for short-term route forecasting systems [81]. The
NN has the ability to predict various traffic routing parame-
ters, as well as these parameters are based on the minimum
and maximum changes in the lane, length of the road seg-
ment, vehicle flow, etc. The ANN has been applied success-
fully to reduce and solve several problems in the IoV [82].
12 Wireless Communications and Mobile Computing

Table 4: Summarize of bioinspired algorithm features.
Author/year Bioinspired algorithm Features Protocol
Mobility Scalability Delay Packet delivery ratio Energy awareness
Fahad et al. [35] (2018) GWO algorithm ✓✓_✓__
De Rango et al. [36] (2020) GWO algorithm ✓_✓_ _ Routing protocol
Jung and Mu’azu [37] (2014) Bee optimization algorithm ✓_✓✓ _ Routing protocol
Sachdev et al. [47] (2016) Firefly algorithm ✓_✓✓ ✓ _
Ramachandran Pillai and Arock [53] (2021) Firefly algorithm _ _ ✓_ _ Routing protocol
Nguyen and Jung [55] 2020 ACO algorithm _ _ ✓✓ __
Jabri et al. [56] 2019 ACO algorithm ✓_✓✓ ✓Routing protocol
Khan et al. [69] (2019) Genetic algorithm _ _ ✓✓ ✓Routing protocol
Hadded et al. [71] (2015) Genetic algorithm ✓__ ✓_ Routing protocol
Aslan and Sen [74] (2019) Genetic programming ✓_✓___
Mehta et al. [76] (2016) Bacterial foraging ✓_✓✓ _ Routing protocol
13Wireless Communications and Mobile Computing

4.2.2. Fuzzy. “Fuzzy”is an uncertainty within “crisp logic”in
which it has two reasons, i.e., true or false, while “fuzzy
logic”reasons approximately or to a degree of true or false.
Fuzzy logic has arisen from the growth of the fuzzy set. A
fuzzy system is a type of CI technology that uses fuzzy theory
to solve issues in many fields [83].
It is applied for the successful control system, home
applications, vehicle communication, etc. Nowadays, it is
mainly used for IoV prediction and density estimation.
Fuzzy logic is often utilized to address different issues in
control, system identification, and signal processing. Fuzzy
modeling is one of the most crucial issues in blazing
research. In the previous years, research and theoretical
study focused on research systems from the application-
original perspective [84].
Figure 6 illustrates the functional structure of a fuzzy
system. It has four stages: fuzzification, fuzzy rule base, fuzzy
inference engine, and defuzzification. In the fuzzification
phase, each section of the input data converts one or more
member functions into member degrees by one view. There
are rules that are based on unclear logic, including inappro-
priate relationships between inputs and outputs. These fuzzy
rules are based on IF-THEN statements. The fuzzy release
engine takes into consideration all the intricate rules of the
zero rule base and learns how to change a set of measures
for the relevant results [48]. In the end, the resulting fuzzy
logic outflows are transformed from an incomplete exhaust
engine to a series of defuzzification processes [85].
4.2.3. Machine Learning (ML). Machine learning is an
important learning technique that is utilized in a real-time
environment. It is the subsection of AI for the development
and understanding of techniques and algorithms that allow
machines to learn [86]. The learning method provides good
IoV, with high-density dependencies of explicit variables
and dramatic downtime for vehicle communication on
major networks. Thus, it outperforms and makes better
predictions than other learning methods. ML is categorized
into supervised and unsupervised learning [87]. The func-
tional overview of ML algorithms and the features of each
algorithm are compared in Tables 5 and 6.
(1) Supervised Learning (SL). SL inputs and their required
outputs and a general rule that maps the entries to outputs.
SL is widely used for classification problems. The SL
methods are divided into classification and regression. It is
one of the important techniques for decision tree networks
and NN training processes. The SL method is broadly used
for IoV. SL involves introducing extra layers between input
and output that change the manual functions with some
algorithms and the extraction of hierarchical properties
[86]. The taxonomy method determines an error in the
NN and minimizes it. The SL algorithm takes a set of IoV
routing input data and trains the model to make forecasts
for responding to new data.
(2) Support Vector Machine (SVM). An SVM is a supervised
technique used for classification. SVM defines the dimen-
sional hyperplane to separate the dataset into two sets.
Network weight can be achieved by solving a quadrature
programmatic problem that has simple limitations, rather
than removing arbitrary migration problems, as well as
training for basic neural networks. The SVM model has a
large amount of memory usage and terms of time. In
contrast, the converting approaches NN, SVM, and non-
parametric regression are utilized for IoV routing path fore-
casting because of their complex arcades and self-learning
characteristics. SVM can also achieve a globally optimal
solution without limiting it to a local minimum point, which
gives strong resistance to unnecessary problems and high
alert performance [88].
(3) Neural Network (NN). Neural networks have been widely
utilized in the IoV area. In many cases, the traffic network
will have a single output variable, although, in the case of
multistate classification issues, this may relate to multiple
output units. There are various applications of NN in the
IoV forecasting area: multilayer perceptrons, time-delayed
recurrent neural networks, multirecurrent neural networks,
state-space neural networks, etc. There are also some hybrid
neural network applications in the field of traffic forecasting.
Neural networks are the most widely applied models to the
traffic prediction problem because they are capable of
modeling nonlinear and dynamic processes well. Many
extensions of the basic concept have been implemented to
improve prediction accuracy and/or reduce computational
effort [89]. Among NN, backpropagation networks are
widely applied because of their ability to model complex
nonlinear relationships among continuous variables. These
multilayer networks perform supervised learning. NN’s
architecture is based on models and predictions of the evolu-
tion of traffic congestion based on global positioning system
Fuzzification Defuzzification
Interference
Rules
Input
data
Fuzzy input sets
Fuzzy output
sets
Outpu
t
data
Figure 6: Block diagram of the fuzzy system.
14 Wireless Communications and Mobile Computing

data. The backpropagation (BP) algorithm is an effective
algorithm, but it has some issues that can be resolved by it.
The backpropagation model has become one of the vital
NN models [90].
(4) Naive Bayesian Network (NBN). The NBN is classified
using Bayes theory and forecasting and is frequently obvious
to the problem solver. This method is widely used in practice
due to its low computational effort and ease of use. A small
amount of training data is required to estimate the parame-
ters required for classification. NBN can be trained in a
controlled learning system. In a number of real-time appli-
cations, the IoV routing paths are predicted for the NBN
model, which uses the highest probability method [91]. In
the NBN, the data from adjacent links is considered infor-
mative for the link currently under investigation. However,
this method is used in the real-time IoV system. The process
of NBN is related to the adjacent road links and statistical
data. From this NBN, the nearest node can be used in rea-
soning about some features of the present node. The NBN
explicitly adopts the data from the adjacent links to provide
IoV communication and prediction. The trafficflow forecast
here can be considered as a reference to the Bayesian net-
work [92].
(5) Unsupervised Learning (UL). UL is the training of a CI
algorithm using IoV data that is neither categorized nor
labeled and allows the algorithm to act on that information
without guidance. If it has no results given to the UL algo-
rithms, then it finds the structure in its input. The aim of
UL is to detect hidden patterns in the data. In today’s envi-
ronment, ML approaches are mainly applicable to transpor-
tation issues. In this review, we divided the techniques under
the UL base algorithm for the IoV routing process [63].
(6) Hidden Markov Model (HMM). An HMM is based on a
Markov chain, which is only partially noticeable in the state.
Observation functions are associated with the state of the
system, but they are usually not enough to accurately deter-
mine the state. In HMM-based IoV estimation approaches,
the communication state on a road segment is a latent state
that can be estimated in accordance with the conditions of
road sections having similar traffic characteristics. The
HMM was developed and calibrated using a large amount
of real-world IoV vehicle communication surveillance data.
It is broadly used for the routing process. The IoV has been
defined as the first- and second-rate statistics on vehicular
communication observations, as well as within the shortest
time frame, local variations, and traffic trends [93].
Table 5: Comparison of machine learning algorithms.
Techniques Overview Limitation Application
SVM
The SVM performs the coordination of individual
observation. The result of this model creates an
optimal hyperplane that separates classes.
The training and testing function of
SVM is slow.
Context-aware
security of IoV
Neural network It is an ANN biological NN and is organized in layers
that are made up of several interconnected “nodes.”
It takes a large time to process and
requires a large dataset.
Speed forecasting
in IoV
Naive Bayesian It is a bioinspired neural system organized in layers
that are connected with each other node.
The function of Bayesian is dependent
on hardware and is repeatedly faced
with inexplicable behavior.
Fast vehicular
communication
K-nearest
neighbor (KNN)
This algorithm is used to utilize a database to
categorize into several classes for forecasting the
sample point.
Its learning process is low. Communication
of IoV is fast
HMM The function of HMM is processed only on the
partially observable.
It needs a large amount of
computational time.
Best routing
process
Table 6: Summary of CI algorithm features.
Author/year Bioinspired
algorithm
Features
Protocol
Mobility Scalability Delay Packet delivery
ratio
Energy
awareness
Karabulut et al. [82] (2019) ANN algorithm _ _ ✓✓ __
Alsarhan et al. [84] (2019) Fuzzy model _ ✓✓ ✓ _Routing
protocol
Bylykbashi et al. [85] 2019 Fuzzy model _ _ ✓_✓Routing
protocol
Wu et al. [88] (2018) SVM algorithm ✓_✓___
Xu et al. [89] (2020) Neural network ✓__ _ ✓_
Yin et al. [90] (2020) Neural network ✓_✓___
Yao et al. [93] (2017) HMM ✓_✓✓ __
15Wireless Communications and Mobile Computing

5. Challenges and Application of This Survey
IoV is the combination of hardware pieces with different
types of networks, which is used to allow pedestrians, cars,
and other units under RSU to interchange real-time infor-
mation. IoV is one of the super versions of VANET. In
general, conventional VANET allows cars to form a wireless
connection with other vehicles. Therefore, all vehicles
nearby IoV have a reliable connection to the local infrastruc-
ture. The main working principle of IoV technology is to
form a social IoV with smart infrastructure. The five types
of IoV infrastructure are vehicle-to-roadside (V2R) unit,
vehicle-to-human (V2H), vehicle-to-sensors (V2S), vehicle-
to-vehicle (V2V), and vehicle-to-infrastructure (V2I). In
addition, in IoV, many issues are encountered when the
routing protocols for vehicular communication and traffic
enhancement are created. Some of these challenges are quick
variation of topology, high mobility, and network fragmen-
tation. Other issues are related to IoV, such as storage, as
the huge number of connected vehicles and the large volume
of data processing take place; thus, managing the network is
a challenging task, which can be solved [94].
The application of IoV is categorized into three types:
safety applications, infotainment and comfort, and manage-
ment of vehicle routing and efficiency.
5.1. Safety-Related Application. The aim of the IoV applica-
tion is to be utilized to reduce and avoid road accidents.
The safety application is used to provide a head warning to
the vehicle drivers. Also, this application is utilized to avoid
mishaps from occurring in any case.
(i) Traffic signal violation warning: it is used to manage
the infrastructure for vehicles’routing to pass the
message between vehicles in any hazard case
(ii) Intersection collision avoidance: based on dedicated
short-range communication, it is used to reduce
accidents and improve road safety. Moreover, it
establishes secure links to provide prior information
in any accident, thus alerting the driver to take
appropriate action
5.2. Management of Vehicle Routing and Efficiency (MVRE).
The MVRE application is designed to prevent road mishaps
by optimizing the trafficflow and ignoring the traffic conges-
tion. In this application, the vehicles forward the notification
about the traffic situation based on the Roadside Unit (RSU)
and On-Board Unit (OBU), and accordingly, the drivers
might change their vehicle routes.
(i) Intersection management: this is a technique for
improving the efficiency of vehicle intersections. In
smart cities, vehicles passing through the intersec-
tion area are dangerous. During that period, spared
roadside-to-vehicle communication will provide
more notifications about the trafficflow prediction
and the current traffic situation
(ii) Vehicle routing traffic management: this manage-
ment is used to manage the flow of vehicle traffic
and reduce congestion in traffic routing. In this turn,
it increases the capacity of routing efficiency and
prevents vehicle traffic jams
5.3. Infotainment and Comfort. The comfort and infotain-
ment applications are designed to improve the vehicle’s driv-
ing comfort. This IoV application can be tolerated for
information losses and delays. In IoV, the unicast routing
is used for infotainment applications, which are used to
search for the nearest locations and pass the information
between vehicles to vehicles, etc. [57–59, 95, 96].
6. Conclusion
In the new era of IoT, the evolution of Internet-based
driving, named IoV, supports ITS, telematics, and ad hoc
vehicular networks. The bioinspired approach in IoV has
proven to be more efficient for large-scale vehicular net-
works in performing routing in terms of greater scalability
and less complexity. In addition, it is more robust and
flexible and ensures better routing performance even when
the networks drop down. Therefore, this article will dis-
cuss a comparative study of various bioinspired algorithm
techniques in IoV routing to support ITS applications.
After discussing the background of the IoV routing, the
basic concept and operations of some recently imple-
mented bioinspired algorithms are mentioned in the tax-
onomy section, which is classified into three categories:
evolutionary algorithms, swarm intelligence, and bacterial
foraging. These three categories include some subclasses
of algorithms such as PSO, BCO, fuzzy logic, ACO, and
GA. Research projects were reviewed, classified, and com-
pared for each category using the essential scheme and
routing criteria, such as complexity, robustness, scalability,
and mobility model employed QoS. To find the conver-
gence in bioinspired applications in the IoV, a suggestion
to propose a unified formal model for managing multiple
solutions is discussed. From this consideration, the sug-
gested model is better suited for a bioinspired approach
than the existing procedures, which can address different
routing aspects. The final part of the review ends with
some lessons learned, future trends, and opportunities for
the IoV route.
Notations
A: Weight of ∂IJ
B: Weight value of ηIJ
∂IJ: Pheromone value of ant in nodes Iand J
ηIJ: The instantaneous state of fuzzy
nJ: Number of nearest nodes
TABUK: Node set not visited by the ant
ΔυλD
tIJ ðTÞ:Pheromone dropped by the vehicle
lυλD
IJ ðTÞ:Length of the edge
kυλD
IJ ðTÞ:Traveling time of vehicle
DυλD
IJ ðTÞ:Density of vehicles.
16 Wireless Communications and Mobile Computing

Data Availability
No data were used to support this study.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
References
[1] N. Omar, N. Yaakob, Z. Husin, and M. Elshaikh, “Design and
development of GreedLea routing protocol for Internet of
Vehicle (IoV),”IOP Conference Series: Materials Science and
Engineering, vol. 767, no. 1, article 012034, 2020.
[2] Z. Guo, Y. Zhang, J. Lv, Y. Liu, and Y. Liu, “An online learning
collaborative method for traffic forecasting and routing opti-
mization,”IEEE Transactions on Intelligent Transportation
Systems, vol. 22, 2021.
[3] M. A. Gawas and S. S. Govekar, “A novel selective cross layer
based routing scheme using ACO method for vehicular net-
works,”Journal of Network and Computer Applications,
vol. 143, pp. 34–46, 2019.
[4] F. Yang, S. Wang, J. Li, Z. Liu, and Q. Sun, “An overview of
internet of vehicles,”China Communications, vol. 11, no. 10,
pp. 1–5, 2014.
[5] L. Sumi and V. Ranga, An IoT-VANET-based traffic manage-
ment system for emergency vehicles in a smart city, Springer,
2018.
[6] U. Z. Hamid, H. Zamzuri, and D. K. Limbu, Internet of Vehicle
(IoV) applications in expediting the implementation of smart
highway of autonomous vehicle: a survey, Springer, 2019.
[7] A. Awang, K. Husain, N. Kamel, and S. Aissa, “Routing in
vehicular ad-hoc networks: a survey on single-and cross-
layer design techniques, and perspectives,”IEEE Access,
vol. 5, pp. 9497–9517, 2017.
[8] B. T. Sharef, R. A. Alsaqour, and M. Ismail, “Vehicular com-
munication ad hoc routing protocols: a survey,”Journal of Net-
work and Computer Applications, vol. 40, pp. 363–396, 2014.
[9] S. A. Shah, M. Shiraz, M. K. Nasir, and R. B. Noor, “Unicast
routing protocols for urban vehicular networks: review, taxon-
omy, and open research issues,”Journal of Zhejiang University
SCIENCE C, vol. 15, no. 7, pp. 489–513, 2014.
[10] A. Dua, N. Kumar, and S. Bawa, “A systematic review on rout-
ing protocols for vehicular ad hoc networks,”Vehicular Com-
munications, vol. 1, no. 1, pp. 33–52, 2014.
[11] I. Wahid, A. A. Ikram, M. Ahmad, S. Ali, and A. Ali, “State of
the art routing protocols in VANETs: a review,”Procedia
Computer Science, vol. 130, pp. 689–694, 2018.
[12] J. Liu, J. Wan, Q. Wang, P. Deng, K. Zhou, and Y. Qiao, “A sur-
vey on position-based routing for vehicular ad hoc networks,”
Telecommunication Systems, vol. 62, no. 1, pp. 15–30, 2016.
[13] R. Brendha and V. S. Prakash, “A survey on routing protocols
for vehicular ad hoc networks,”in 2017 4th International Con-
ference on Advanced Computing and Communication Systems
(ICACCS), Coimbatore, India, 2017.
[14] H. H. Saleh and S. T. Hasson, “A survey of routing algorithms
in vehicular networks,”in 2019 International Conference on
Advanced Science and Engineering (ICOASE), Zakho, Duhok,
Iraq, 2019.
[15] A. Ghaffari, “Hybrid opportunistic and position-based routing
protocol in vehicular ad hoc networks,”Journal of Ambient
Intelligence and Humanized Computing, vol. 11, no. 4,
pp. 1593–1603, 2020.
[16] M. T. Abbas, A. Muhammad, and W. C. Song, “SD-IoV: SDN
enabled routing for internet of vehicles in road-aware
approach,”Journal of Ambient Intelligence and Humanized
Computing, vol. 11, no. 3, pp. 1265–1280, 2020.
[17] A. Baykasoğlu and Ş. Akpinar, “Weighted superposition
attraction (WSA): a swarm intelligence algorithm for optimi-
zation problems - part 1: unconstrained optimization,”
Applied Soft Computing, vol. 56, pp. 520–540, 2017.
[18] H. Fatemidokht and M. K. Rafsanjani, “F-Ant: an effective
routing protocol for ant colony optimization based on fuzzy
logic in vehicular ad hoc networks,”Neural Computing and
Applications, vol. 29, no. 11, pp. 1127–1137, 2018.
[19] S. Yaqoob, A. Ullah, M. Akbar, M. Imran, and M. Shoaib,
“Congestion avoidance through fog computing in internet of
vehicles,”Journal of Ambient Intelligence and Humanized
Computing, vol. 10, no. 10, pp. 3863–3877, 2019.
[20] C. Archetti, F. Guerriero, and G. Macrina, “The online vehicle
routing problem with occasional drivers,”Computers & Oper-
ations Research, vol. 127, article 105144, 2021.
[21] H. Liu, Z. Guo, and Z. Zhang, “A hybrid multi-level opti-
misation framework for integrated production scheduling
and vehicle routing with flexible departure time,”Interna-
tional Journal of Production Research, vol. 30, pp. 1–8,
2020.
[22] D. Zhang, D. Li, H. Sun, and L. Hou, “A vehicle routing prob-
lem with distribution uncertainty in deadlines,”European
Journal of Operational Research, vol. 292, 2021.
[23] S. Kaur, T. C. Aseri, and S. Rani, QoS-aware routing in vehicu-
lar ad hoc networks using ant colony optimization and bee col-
ony optimization, Springer, 2019.
[24] M. Elhoseny and K. Shankar, Energy efficient optimal routing
for communication in VANETs via clustering Model, Springer,
2020.
[25] W. Wang, F. Xia, H. Nie et al., “Vehicle trajectory clustering
based on dynamic representation learning of internet of vehi-
cles,”IEEE Transactions on Intelligent Transportation Systems,
vol. 22, 2020.
[26] S. A. Hussain, K. M. Yusof, S. M. Hussain, and A. V. Singh, “A
review of quality of service issues in Internet of Vehicles
(IoV),”in 2019 Amity International Conference on Artificial
Intelligence (AICAI), Amity University Dubai, 2019.
[27] J. Cheng, J. Cheng, M. Zhou, F. Liu, S. Gao, and C. Liu, “Rout-
ing in Internet of Vehicles: a review,”IEEE Transactions on
Intelligent Transportation Systems, vol. 16, no. 5, pp. 2339–
2352, 2015.
[28] T. Kayarga and S. A. Kumar, “A study on various technologies
to solve the routing problem in Internet of Vehicles (IoV),”
Wireless Personal Communications, vol. 119, no. 1, pp. 459–
487, 2021.
[29] L. Tuyisenge, M. Ayaida, S. Tohme, and L. E. Afilal, “Network
architectures in internet of vehicles (IoV): review, protocols
analysis, challenges and issues,”in International Conference
on Internet of Vehicles, Paris, France, 2018.
[30] C. Ksouri, I. Jemili, M. Mosbah, and A. Belghith, “Towards
general Internet of Vehicles networking: routing protocols sur-
vey,”Concurrency and Computation: Practice and Experience,
vol. 21, article e5994, 2022.
[31] X. S. Yang, Ed., Nature-Inspired Algorithms and Applied Opti-
mization, Springer, 2017.
17Wireless Communications and Mobile Computing

[32] P. Agarwal and S. Mehta, “Nature-inspired algorithms: state-
of-art, problems and prospects,”International Journal of Com-
puter Applications, vol. 100, no. 14, pp. 14–21, 2014.
[33] H. Zang, S. Zhang, and K. Hapeshi, “A review of nature-
inspired algorithms,”Journal of Bionic Engineering, vol. 7,
no. S4, pp. S232–S237, 2010.
[34] D. Rai and K. Tyagi, “Bio-inspired optimization techniques,”
ACM SIGSOFT Software Engineering Notes, vol. 38, no. 4,
pp. 1–7, 2013.
[35] M. Fahad, F. Aadil, S. Khan et al., “Grey wolf optimization
based clustering algorithm for vehicular ad-hoc networks,”
Computers & Electrical Engineering, vol. 70, pp. 853–870,
2018.
[36] F. De Rango, M. Tropea, P. Raimondo, and A. F. Santamaria,
“Grey wolf optimization in vanet to manage platooning of
future autonomous electrical vehicles,”in 2020 IEEE 17th
Annual Consumer Communications & Networking Conference
(CCNC), Las Vegas, NV, USA, 2020.
[37] L. T. Jung and A. A. Mu'azu, “Bee inspired QoS routing in
VANET,”in 2014 4th World Congress on Information and
Communication Technologies (WICT 2014), Melaka, Malaysia,
2014.
[38] T. A. Masutti and L. N. de Castro, “Bee-inspired algorithms
applied to vehicle routing problems: a survey and a proposal,”
Mathematical Problems in Engineering, vol. 2017, Article ID
3046830, 20 pages, 2017.
[39] D. Karaboga, “An idea based on honey bee swarm for numer-
ical optimization,”Tech. Rep. tr06, Erciyes University, Engi-
neering Faculty, Computer Engineering Department, 2005.
[40] B. Basturk and D. Karaboga, “On the performance of artificial
bee colony (ABC) algorithm,”Applied Soft Computing Journal,
vol. 8, no. 1, pp. 687–697, 2008.
[41] A. Banharnsakun, T. Achalakul, and B. Sirinaovakul,
“ABCGSX: a hybrid method for solving the traveling salesman
problem,”in Proceedings of the 2010 2nd World Congress on
Nature and Biologically Inspired Computing, NaBIC 2010,
Kitakyushu, Japan, 2010.
[42] D. Karaboga, B. Gorkemli, C. Ozturk, and N. Karaboga, “A
comprehensive survey: artificial bee colony (ABC) algorithm
and applications,”Artificial Intelligence Review, vol. 42, no. 1,
pp. 21–57, 2014.
[43] S. Nakrani and C. Tovey, “On honey bees and dynamic server
allocation in internet hosting centers,”Adaptive Behavior,
vol. 12, no. 3-4, pp. 223–240, 2004.
[44] L. P. Wong, M. Y. H. Low, and C. S. Chong, “A bee colony
optimization algorithm for traveling salesman problem,”in
Proceedings of the Second Asia International Conference on
Modelling & Simulation, pp. 818–823, Kuala Lumpur, Malay-
sia, 2008.
[45] X. Lu and Y. Zhou, “A novel global convergence algorithm: bee
collecting pollen algorithm,”in Advanced Intelligent Comput-
ing Theories and Applications. With Aspects of Artificial Intel-
ligence. ICIC 2008, D. S. Huang, D. C. Wunsch, D. S. Levine,
and K. H. Jo, Eds., vol. 5227 of Lecture Notes in Computer Sci-
ence, Springer, Berlin, Heidelberg, 2008.
[46] H. Abbass, “MHBO: marriage in honey bees optimization-a
haplometrosis polygynous swarming approach,”in Proceed-
ings of the 2001 IEEE Congress on Evolutionary Computation,
Seoul, Korea (South), 2001.
[47] A. Sachdev, K. Mehta, and L. Malik, “Design of protocol for
cluster based routing in VANET using fire fly algorithm,”in
2016 IEEE international conference on engineering and tech-
nology (ICETECH), Coimbatore, India, 2016.
[48] S. Łukasik and S. Zak, “Firefly algorithm for continuous con-
strained optimization tasks,”in Computational Collective
Intelligence Semantic Web, Social Networks and Multiagent
Systems, Springer, Berlin/Heidelberg, Germany, 2009.
[49] X. S. Yang, Nature-Inspired Metaheuristic Algorithms, Luniver
Press, Beckington, UK, 2010.
[50] A. M. Altabeeb, A. M. Mohsen, and A. Ghallab, “An improved
hybrid firefly algorithm for capacitated vehicle routing prob-
lem,”Applied Soft Computing, vol. 84, article 105728, 2019.
[51] D. Trachanatzi, M. Rigakis, M. Marinaki, and Y. Marinakis, “A
fireflyalgorithm for the environmental prize-collecting vehicle
routing problem,”Swarm and Evolutionary Computation,
vol. 57, article 100712, 2020.
[52] A. M. Altabeeb, A. M. Mohsen, L. Abualigah, and A. Ghallab,
“Solving capacitated vehicle routing problem using coopera-
tive firefly algorithm,”Applied Soft Computing, vol. 108, article
107403, 2021.
[53] R. Ramachandran Pillai and M. Arock, “Spiking neural firefly
optimization scheme for the capacitated dynamic vehicle rout-
ing problem with time windows,”Neural Computing and
Applications, vol. 33, no. 1, pp. 409–432, 2021.
[54] P. M. Kumar, G. Manogaran, R. Sundarasekar,
N. Chilamkurti, and R. Varatharajan, “Ant colony optimiza-
tion algorithm with internet of vehicles for intelligent traffic
control system,”Computer Networks, vol. 144, pp. 154–162,
2018.
[55] T. H. Nguyen and J. J. Jung, “ACO-based approach on
dynamic MSMD routing in IoV environment,”in 2020 16th
International Conference on Intelligent Environments (IE),
Madrid, Spain, 2020.
[56] I. Jabri, T. Mekki, A. Rachedi, and M. B. Jemaa, “Vehicular fog
gateways selection on the internet of vehicles: a fuzzy logic
with ant colony optimization based approach,”Ad Hoc Net-
works, vol. 91, article 101879, 2019.
[57] M. Elhoseny, “Intelligent firefly-based algorithm with Levy
distribution (FF-L) for multicast routing in vehicular commu-
nications,”Expert Systems with Applications, vol. 140, article
112889, 2020.
[58] R. Attia, A. Hassaan, and R. Rizk, “Advanced greedy hybrid
bio-inspired routing protocol to improve IoV,”IEEE Access,
vol. 9, pp. 131260–131272, 2021.
[59] R. Dhanare, K. K. Nagwanshi, and S. Varma, “Enhancing the
route optimization using hybrid MAF optimization algorithm
for the internet of vehicle,”Wireless Personal Communica-
tions, vol. 2, pp. 1–21, 2022.
[60] J. Cheng, C. Cao, M. Zhou, C. Liu, S. Grao, and C. Jiang, “A
dynamic evolution mechanism for IoV community in an
urban scene,”IEEE Internet of Things Journal, vol. 8, no. 9,
pp. 7521–7530, 2020.
[61] S. Sivanandam and S. Deepa, “Genetic algorithms,”in Intro-
duction to Genetic Algorithms, pp. 15–37, Springer, 2008.
[62] D. Gupta and R. Kumar, “An improved genetic based routing
protocol for VANETs,”in 2014 5th International Conference
Confluence The Next Generation Information Technology
Summit (Confluence), pp. 347–353, Noida, India, 2014.
[63] J. Zhang, “An improved genetic algorithm for vehicle routing
problem,”in International Conference on Machine Learning
and Big Data Analytics for IoT Security and Privacy, Shanghai,
China, 2020.
18 Wireless Communications and Mobile Computing

[64] M. Zirour, “Vehicle routing problem: models and solutions,”
Journal of Quality Measurement and Analysis JQMA, vol. 4,
no. 1, pp. 205–218, 2008.
[65] H. Nazif and L. S. Lee, “Optimized crossover genetic algorithm
for vehicle routing problem with time windows,”American
Journal of Applied Sciences, vol. 7, no. 1, pp. 95–101, 2010.
[66] S. Zhong-yue, G. Zhong-liang, and W. Qin, “An improved
adaptive genetic algorithm for vehicle routing problem,”in
2010 International Conference on Logistics Systems and Intelli-
gent Management, Harbin, China, 2010.
[67] J. Castro-Gutierrez, D. Landa-Silva, and J. M. Pérez, “Nature of
real-world multi-objective vehicle routing with evolutionary
algorithms,”in 2011 IEEE International Conference on Sys-
tems, Man, and Cybernetics, Anchorage, AK, USA, 2011.
[68] T. Murata and R. Itai, “Multi-objective vehicle routing prob-
lems using twofold emo algorithms to enhance solution simi-
larity on non-dominated solutions,”in Evolutionary Multi-
Criterion Optimization, vol. 3410 of LNCS, pp. 885–896,
Springer, Berlin/Heidelberg, 2005.
[69] A. A. Khan, M. Abolhasan, W. Ni, J. Lipman, and
A. Jamalipour, “A hybrid-fuzzy logic guided genetic algorithm
(H-FLGA) approach for resource optimization in 5G
VANETs,”IEEE Transactions on Vehicular Technology,
vol. 68, no. 7, pp. 6964–6974, 2019.
[70] C. H. Lee, K. G. Lim, M. K. Tan, R. K. Chin, and K. T. Teo, “A
genetic algorithm for management of coding resources in
VANET,”in 2017 IEEE 2nd International Conference on Auto-
matic Control and Intelligent Systems (I2CACIS), Kota Kina-
balu, Malaysia, 2017.
[71] M. Hadded, R. Zagrouba, A. Laouiti, P. Muhlethaler, and L. A.
Saidane, “A multi-objective genetic algorithm-based adaptive
weighted clustering protocol in vanet,”in 2015 IEEE congress
on evolutionary computation (CEC), 2015.
[72] M. Abbasi, M. Rafiee, M. R. Khosravi, A. Jolfaei, V. G. Menon,
and J. M. Koushyar, “An efficient parallel genetic algorithm
solution for vehicle routing problem in cloud implementation
of the intelligent transportation systems,”Journal of Cloud
Computing, vol. 9, no. 1, pp. 1–4, 2020.
[73] K. C. Huang, Y. W. Wen, and C. K. Ting, “Enhancing
k-nearest neighbors through learning transformation func-
tions by genetic programming,”in 2019 IEEE Congress on Evo-
lutionary Computation (CEC),Wellington, New Zealand,
2019.
[74] M. Aslan and S. Sen, “Evolving trust formula to evaluate data
trustworthiness in VANETs using genetic programming,”in
International Conference on the Applications of Evolutionary
Computation (Part of Evo Star), Leipzig, Germany, 2019.
[75] B. Niu, H. Wang, L. J. Tan, L. Li, and J. W. Wang, “Vehicle
routing problem with time windows based on adaptive bacte-
rial foraging optimization,”in International Conference on
Intelligent Computing, Huangshan, China, 2012.
[76] K. Mehta, P. R. Bajaj, and L. G. Malik, “Fuzzy bacterial forag-
ing optimization zone based routing (FBFOZBR) protocol for
VANET,”in 2016 International Conference on ICT in Business
Industry & Government (ICTBIG), Indore, India, 2016.
[77] V. A. Saunders, R. Gregory, and J. R. Saunders, Individual-
based modeling of bacterial genetic elements, IGI Global,
2009.
[78] C. Vlachos, R. C. Paton, J. R. Saunders, and Q. H. Wu, “A rule-
based approach to the modelling of bacterial ecosystems,”Bio-
systems, vol. 84, no. 1, pp. 49–72, 2006.
[79] J. R. Faeder, M. L. Blinov, B. Goldstein, and W. S. Hlavacek,
“Rule-based modeling of biochemical networks,”Complexity,
vol. 10, no. 4, pp. 22–41, 2005.
[80] N. Hegde and S. S. Manvi, “Implementation of VANET rout-
ing using computational intelligence,”in 2016 International
Conference on Communication and Signal Processing (ICCSP),
Melmaruvathur, India, 2016.
[81] C. Dutta and D. Singhal, “A hybridization of artificial neural
network and support vector machine for prevention of road
accidents in VANET,”Engineering and Technology, vol. 10,
no. 1, 2019.
[82] M. A. Karabulut, A. F. M. Shahen Shah, and H. Ilhan, “Perfor-
mance Optimization by Using Artificial Neural Network Algo-
rithms in VANETs,”in 2019 42nd International conference on
telecommunications and signal processing (TSP), pp. 633–636,
Budapest, Hungary, 2019.
[83] K. Ozera, K. Bylykbashi, Y. Liu, and L. Barolli, “A fuzzy-based
approach for cluster management in VANETs: performance
evaluation for two fuzzy-based systems,”Internet of Things,
vol. 3-4, pp. 120–133, 2018.
[84] A. Alsarhan, Y. Kilani, A. Al-Dubai, A. Y. Zomaya, and
A. Hussain, “Novel fuzzy and game theory based clustering
and decision making for VANETs,”IEEE Transactions on
Vehicular Technology, vol. 69, no. 2, pp. 1568–1581, 2019.
[85] K. Bylykbashi, D. Elmazi, K. Matsuo, M. Ikeda, and L. Barolli,
“Effect of security and trustworthiness for a fuzzy cluster man-
agement system in VANETs,”Cognitive Systems Research,
vol. 55, pp. 153–163, 2019.
[86] A. Uprety, D. B. Rawat, and J. Li, “Privacy preserving misbe-
havior detection in IoV using federated machine learning,”
in 2021 IEEE 18th Annual Consumer Communications & Net-
working Conference (CCNC), Las Vegas, NV, USA, 2021.
[87] S. A. Siddiqui, A. Mahmood, W. E. Zhang, and Q. Z.
Sheng, “Machine learning based trust model for misbeha-
viour detection in Internet-of-Vehicles,”in International
Conference on Neural Information Processing, Sydney, NSW,
Australia, 2019.
[88] S. Wu, W. Xia, W. Cui et al., “An efficient offloading algorithm
based on support vector machine for mobile edge computing
in vehicular networks,”in 2018 10th International Conference
on Wireless Communications and Signal Processing (WCSP),
Hangzhou, China, 2018.
[89] L. Xu, H. Wang, and T. A. Gulliver, “Outage probability per-
formance analysis and prediction for mobile IoV networks
based on ICS-BP neural network,”IEEE Internet of Things
Journal, vol. 8, no. 5, pp. 3524–3533, 2020.
[90] K. L. Yin, Y. F. Pu, and L. Lu, “Hermite functional link artifi-
cial-neural-network-assisted adaptive algorithms for IoV non-
linear active noise control,”IEEE Internet of Things Journal,
vol. 7, no. 9, pp. 8372–8383, 2020.
[91] P. Gu, R. Khatoun, Y. Begriche, and A. Serhrouchni, “k-nearest
neighbours classification based Sybil attack detection in vehic-
ular networks,”in 2017 Third International Conference on
Mobile and Secure Services (Mobi Sec Serv), Miami Beach, FL,
USA, 2017.
[92] X. Guo, Y. Chen, L. Cao, D. Zhang, and Y. Jiang, “A
receiver-forwarding decision scheme based on Bayesian for
NDN-VANET,”China Communications, vol. 17, no. 8,
pp. 106–120, 2020.
[93] L. Yao, J. Wang, X. Wang, A. Chen, and Y. Wang, “V2X rout-
ing in a VANET based on the hidden Markov model,”IEEE
19Wireless Communications and Mobile Computing

Transactions on Intelligent Transportation Systems, vol. 19,
no. 3, pp. 889–899, 2017.
[94] T. T. Dandala, V. Krishnamurthy, and R. Alwan, “Internet of
Vehicles (IoV) for traffic management,”in 2017 International
conference on computer, communication and signal processing
(ICCCSP), Chennai, India, 2017.
[95] X. Wang, Z. Ning, and L. Wang, “Offloading in Internet of
Vehicles: a fog-enabled real-time traffic management system,”
IEEE Transactions on Industrial Informatics, vol. 14, no. 10,
pp. 4568–4578, 2018.
[96] S. Ansari and K. K. Nagwanshi, “An empirical study of soft
computing approaches in wireless sensor networks,”Journal
of Cases on Information Technology (JCIT), vol. 24, 2022.
20 Wireless Communications and Mobile Computing