Received: 21 March 2019 Revised: 16 April 2019 Accepted: 19 April 2019

DOI: 10.1002/dac.4005

RESEARCH ARTICLE

Software defined vehicular networks: A comprehensive

review

Jitendra Bhatia1,2 Yash Modi2 Sudeep Tanwar1 Madhuri Bhavsar1

1
Department of Computer Science and
Engineering, Institute of Technology, Summary
Nirma University, Ahmedabad, Gujarat, The recent breakthroughs in the automobile industries and telecommunication
India
2
technologies along with the exceptional multimodal mobility services brought
Computer Department, Vishwakarma
Government Engineering College, focus on intelligent transportation system (ITS), of which vehicular ad hoc
Ahmedabad Gujarat Technological networks (VANETs) gain much more attention. The distinctive features of
University, Gujarat, India
software-defined networking (SDN) leverages the vehicular networks by its state
Correspondence of the centralized art having a comprehensive view of the network. Its potential
Sudeep Tanwar, Department of Computer to bring the flexibility, programmability and other extensive advancements to
Science and Engineering, Institute of
Technology, Nirma University, vehicular networks has set the stage for a novel networking paradigm termed as
Ahmedabad, Gujarat, India. software-defined vehicular networks (SDVNs). Many researchers have demon-
Email: [email protected]
strated the SDN-based VANETs with the various configuration of the SDN
components in VANET architecture. However, a compilation of the work on the
SDN-based VANET system as a whole, incorporating its architecture, use-cases,
and opportunities, is still inadequate. We start with the summary of the recent
studies that exist on the SDVNs, followed by the comprehensive explanation of
its components. Next, we present the taxonomy of SDVN based on the archi-
tecture modes, protocols, access technologies, and opportunities with trending
technologies. Finally, we highlight the challenges, open research issues, and
future research directions.

1 I N T RO DU CT ION

Recent advancements in wireless communication and automotive technology have been witnessing the vehicular ad hoc
networks. Vehicular networks have attracted a lot of interest in the past decade from both academia and industry. Vehicles
have become dominant and smarter as they are equipped with a wide variety of sensors, onboard computers, and different
devices supporting navigation and communication. Vehicular networks support the plethora of applications ranging from
safety to non-safety application via enabling vehicles to communicate with infrastructure (ie, infrastructure-to-vehicle,
I2V), with each other (ie, vehicle-to-vehicle, V2V), and hybrid mode (ie, jointly termed as V2X). The Federal Commu-
nications Commission (FCC), within the United States devoted 75 MHz of the recurrence range in the range 5.850 to
5.925 GHz to be utilized for a vehicle to vehicle and vehicle to roadside correspondence, known as dedicated short-range
communications (DSRC).1,2 DSRC based on IEEE802.11p has gained increasing attention for its easy deployment, low
cost, and capacity to naturally supports V2V communications. Vehicular networks are expected to provide various kinds of
services such as vehicle safety, traffic alert dissemination, route planning, context-aware infotainments, and surveillance.
Although vehicular ad hoc networks (VANETs) are deployed in reality offering several such services; many challenges still
exist in the current system because of difficulties in both management and deployment. The need for efficient resource
utilization, balancing the traffic flow among multi-path topology, delay constraints due to intermittent disconnectivity,

Int J Commun Syst. 2019;32:e4005. wileyonlinelibrary.com/journal/dac © 2019 John Wiley & Sons, Ltd. 1 of 22
https://doi.org/10.1002/dac.4005
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 22 BHATIA ET AL.

TABLE 1 Nomenclature used Acronym Meaning

SDN software-defined networking
SDVN software-defined vehicular network
VANET vehicular ad hoc network
ITS intelligent transportation system
API application programming interface
V2V vehicle to vehicle
V2I vehicle to infrastructure
V2R vehicle to RSU
TE traffic engineering
OBU on board unit
RSU road side unit
QoS quality of service
LTE long term evolution
CDMA code division multiple access
DSRC dedicated short range communication
WAVE wireless access vehicular environment
CH cluster head
PDR packet delivery ratio
MAC media access control
SNR signal to noise ratio
ORUR optimal resource utilization routing
IGA improved genetic algorithm
ROFS roadside OpenFlow switch
CRM cloud resource management
PRETTI predictive time dependent multicast
DDoS distributed denial of service
IoV internet of vehicle
NFV network function virtualization
OVN overlay virtual network

and high mobility are still the challenges because of exponential growth in the number of vehicles and seems difficult to
be solved with the conventional architecture. Moreover, to cope up with the wide variety of services needed by the stake-
holders, VANET expects the flexibility, scalability, and efficient decision-making systems in the current architecture. It
led us to the key requirements of an open and flexible vehicular architecture. The current architecture does not meet the
various requirements such as flexibility and scalability as required by ITS applications. We focus on software-defined net-
working (SDN) to address these challenges. The summary of various acronyms used throughout this paper is mentioned
in Table 1.

1.1 Motivation
The underlying motivation behind this work is to introduce network virtualization in a VANET environment leveraged
by SDN technology. SDN is a revolution in the field of computer networking that gives assurance about the flexibility
and simplicity of a network.3 It can dynamically change, manage, and control network behaviour with the help of cen-
tralized software components. The primary objective of SDN is to make provision for the administrator and network
engineers to respond quickly as per the changing business requirements. It also allows network engineers to shape the traf-
fic from a single centralized controller without taking care of how the devices are connected with each other.3 It is open,
reprogrammable and provides the way for network resource virtualization on the demand bases.4 The key feature of
the SDN is to disassemble the control plane and forwarding plane (ie, data plane) so, the network can perform tasks
of controlling and forwarding of packets individually. The control plane acting as the brain of the network5 decides the
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 3 of 22

TABLE 2 Progress of software-defined vehicular networks (SDVNs)

Author Year Objective Limitation(s)
Ku et al8 2014 Integration of SDN into VANETs with new Advancements in SDVNs are not covered
architecture. and open issues are missing.
Zhu et al9 2015 Use of SDN paradigm in VANET along with Short survey. Architecture and protocol
opportunities and challenges. implementation is not covered.
He et al10 2016 Survey of SDVN with case studies detailed Provided brief survey. Use cases are
architecture. lacking.
Mendiola et al11 2017 Full-fledged survey on SDN protocols for TE Focused only on one aspect of VANET, ie,
traffic engineering.
Yaqoob et al12 2017 Survey on SDVN with new technologies, use cases, Explanation of SDVN protocol is lacking.
and challenges,
Zekri et al13 2018 360 degree survey of heterogeneous VANETs. SDN technology is lacking.
Proposed Survey 2019 Comprehensive survey of SDN enabled VANET -
with taxonomy, use cases, and challenges.
Abbreviations: SDN, software-defined networking; TE, traffic engineering; VANET, vehiclar ad hoc network.

policy for the movement of packets from one place to another place in the network.6 The control plane consists of an
SDN-enabled controller for handling various applications. The controller logic is implemented in the server as a software
component while the data plane logic is executed in the networking devices like switch and router. The data plane for-
wards the packets to the appropriate destination.6 SDN is an architecture that not only controls the network devices but
also controls the entire network.
SDN architecture is dynamic, sensible, and cost-effective. It manages the whole system centrally, and hence, there is no
need for touching individual devices for any change in the network. The OpenFlow standard is the primary protocol for the
development of SDN-based applications.4 There are three components in the SDN model which includes SDN controller,
southbound API, and northbound API. SDN controllers are OpenFlow-enabled controller offering a centralized view of
the whole network. SDN uses a southbound API for exchanging the information from the controller to network hardware
devices like switches and router using OpenFlow standard. The northbound API enables communication between various
applications like the load balancer, firewall, etc and the controller. With SDN, it is possible to oversee the entire system
through insightful coordination and give a framework that empowers the virtualized network and provisions of the assets
according to the request. 7
SDN is the new network framework that can provide an effective and efficient solution for the complex network traffic.4
The applicability of an SDN paradigm offers a flexible way to control the network, and hence it can change the con-
ventional method of VANET operations. The flexibility of SDN to satisfy the various VANETs requirements makes it an
attractive approach. As the traditional VANETs are unable to exploit the full fledge advantages of vehicular technologies,
a new concept of integrating SDN into VANETs have arisen. It motivates us to compile the recent advancements of SDN in
vehicular networks. In past years, many researchers have implemented the SDN functionalities on the top of the VANET
to make the vehicular networks pervasive.

1.2 Scope
The scope of this work is to present a comprehensive survey of all existing SDVN systems. This review covers the trends,
use-cases, and challenges of an SDVN system as a whole. Table 2 shows the past few years progresses of SDVN along
with their limitations. The current survey focuses on the use of SDN paradigm in VANET along with opportunities and
challenges for traffic engineering, support for heterogeneity. The proposed review provides the coverage of recent advance-
ments in SDVNs along with their open issues. We have selected the brief taxonomy for the study of this paper; however,
the content of SDN in VANETs are not necessarily be limited to this work.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 of 22 BHATIA ET AL.

FIGURE 1 Organization of the paper

1.3 Organization
The rest of the paper is organized as follows. Our subsequent sections present the bigger picture of SDVN systems. Section
2 briefly discusses the SDN and VANET along with their convergence into a single system. The new architecture for the
software-defined vehicular network (SDVN) is presented late in that section. Section 3 represents the SDVN system on a
large scale and taxonomy of SDVN. Section 4 and section 5 present use cases with case study and open issues/research
challenges respectively. Finally, Section 6 concluded with future research directions. Figure 1 shows the outline of the
paper.

2 CO NVERGENCE O F SDN IN VANET

2.1 SDN enabled VANET

The SDN paradigm is widely accepted as an encouraging way of revamping the vehicular networks. SDN has arisen as a
stable approach to govern the network in an organized way. An OpenFlow is used as the most promising SDN protocol
for communication between the control plane and data plane. The flexibility of SDN makes it an appealing perspective to
satisfy the requirements of VANET scenarios. The centralized control, flexibility, and programmability which are missing
in today's ad hoc wireless networks can be availed by applying SDN principles to VANETs. SDN concepts in VANET will
simplify the network management and also introduces the new V2V and V2I services. It provides wireless resource opti-
mization such as channel allocation, interference avoidance, packet routing in multi-hop multi-path scenarios, efficient
mobility, and network heterogeneity management. Being just a soft concept, rather than a rigid structure (architecture),
integrating SDN into VANET is a considerable solution. Hence, the revamped architecture of VANET with the SDN
will reduce the impact of disconnectivity caused by vehicles' mobility and enhance the reliability of communication in
VANETs, promoting the development of smart mobility. Following are the three major aspects of the SDN integrated
VANET.
• Appropriate path: Extending SDN into VANET provides freedom of taking more detailed routing decisions. Some-
times, data traffic can be congested or excavated in VANET scenarios. It happens when all the nodes try to use the
shortest path which will make some node very congested. When the controller detects scenarios mentioned above,
then it can stop the currently executing process and start it again to increase utilization of the network and decreases
the network blockages.
• Channel/frequency: Integrating SDN in VANET made available various accessible wireless interfaces or config-
urable radios (eg, cognitive radios). For instance, the controller will decide dynamically about the radio interface.
It enables adaptive radio frequency selection for different sort of traffic at the run time. The major aim of this is to
conserve the frequency or channel for emergency services.
• Transmission power: It is an essential aspect for VANET to choose the proper energy level of wireless interfaces
and its transmission range. In VANET scenario, the controller collects the neighbour information from vehicles
(wireless nodes). From this gathered information, it can be known when roads are empty or vehicles are farther
away from each other. The controller adaptively assists every node to tune the power their level for the efficient
packet delivery.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 5 of 22

FIGURE 2 Software-defined vehicular network (SDVN) architecture

Through many state-of-the-art functionalities such as network function virtualization, centralized controls, and so on,
the convergence of SDN and VANET emerges to provide a pleasing solution to implement the smart VANET, which is
widely adopted as SDVN. Figure 2 shows generic architecture for SDVNs, which is compatible with both V2I and V2V.
Vehicles are connected under a particular RSU, and all the RSUs are OpenFlow-enabled. These RSUs are equipped with
short-range radio transceivers and can communicate with the available central global controller through cellular base
stations. An application plane is connected with the global controller via the northbound API.
Traditional networks are application-aware as they are tightly coupled systems and work as single integrated systems.
But, SDVN systems generally have three logical groups of components, namely, infrastructure plane, control plane, and
application plane. Figure 3 shows the components involved in the software-defined vehicular network with some sig-
nificant functions implemented in it. The SDVN functioning follows both the bottom-up and top-down paradigm. The
vehicular information flows from bottom to up, whereas aggregated decisions and instructions propagated from top to
down. In a software-defined vehicular network, the application layer is considered as a boss, whereas the control layer and
infrastructure layer are supposed to be a brain and a body, respectively. The following subsection describes the component
view of three-layered SDVN architecture.

2.1.1 Component view of SDVN

This section discusses the different possible components falling at each layer with their respective functionality to exploit
the full fledge benefits of SDVN.
Infrastructure plane: The infrastructure plane spans across the physical infrastructures and network devices. It
includes the various types of network devices such as mobile vehicles, OBUs, and RSUs that are responsible for generating
vehicular information such as routes, speed, positions, etc. It also comprises several forwarding devices such as routers,
switches, etc. The heterogeneity of current vehicular networks is eliminated by introducing the overlay network in the
infrastructure plane. The SDN switches provide the abstraction of all the vehicles and RSUs.
SDN wireless node: It refers to the mobile network device at data plane known as vehicle or OBU that receive con-
trol messages from SDN controller via RSU or cellular base station and are also responsible for collecting the various
network data.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 of 22 BHATIA ET AL.

FIGURE 3 Component diagram for

software-defined vehicular network (SDVN)

SDN RSU: These are the stationary network devices on the edge of the data layer deployed along the road segments or
with the cellular base station. Mostly, RSUs are open flow–enabled and its communication interface is compatible with
wired, wireless, and LTE standards.
Cellular Base Station (BS): These are stationery network devices on the edge of the control layer. The cellular base
station also termed as BS or NodeB or eNodeB. They work under the control of the SDN controller, and BS facilitates
wireless connectivity between SDN controller and RSUs or OBUs using wireless communication technologies, ie, 3G, 4G
LTE, Wi-Fi, WiMAX, 5G, CDMA.
Two types of approaches are favoured to make infrastructure plane layer adequate for sending and retrieving the data.
• Vehicle-to-infrastructure (V2I): Under this vehicular communication environment, vehicles can communicate
with deployed RSUs and are not concerned with the information of other vehicles. As all the devices along the path
to vehicles are static, also known as fixed infrastructure plane.
• Vehicle-to-vehicle (V2V): This approach is suitable when all the traditional mobile vehicles are equipped with the
special on-board units that support SDN functionalities (also called a mobile infrastructure plane). In V2V approach,
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 7 of 22

each SDN-compatible vehicle monitors the vehicle attributes, ie, id, speed, direction, etc. The concept of clustering
is introduced in this approach. This information is periodically forwarded to the cluster head. The cluster head is
one of the vehicle playing the role of local controller and maintains the flow table.

Control plane: It acts as a core part of software-defined VANET consisting of SDN controller and other modules such
as firewall module, failure control, protocols, system status, database, flow tables, etc. It takes care of centralization by
providing the global view of the entire network.
SDN controller: SDN controller is a logical entity that collects information from infrastructure plane as per the require-
ment of the application plane. It leverages the architecture with storage and calculation capabilities. The controller also
provides an abstract view of the network including statistics and events.
The control plane also has two types that are fixed control plane and mobile control plane.
• Fixed control plane: It has all the controllers fixed on the infrastructure. The global controller (GC) is set at one
place or in the cloud server. Local controllers are configured on RSUs or cellular base stations and are used to collect
and filter out the received data. All controllers have several interfaces for communicating up and down.
• Mobile control plane: It has its global controller located at the central unit or in the cloud. The local controllers
(LC) are deployed in the mobile vehicle itself. Tasks are divided between GC and LCs. The global controllers respond
to requests that require complex and powerful computing resources. It also serves the bandwidth sensitive demands
or asks for the global view of the network. The LCs take care of the normal running requests and become active only
when that vehicle is chosen as the cluster head. LC comes with limited memory and computational power, so it acts
as a slave of the global controllers.

Application plane: This plane consists of various applications that communicate with control layer plane and pro-
vides insights of the network by fetching the information from the SDN controller. There are the plethora of applications
that can be executed such as monitoring, QoS, analytics, recovery, security, routing, load balancing, network virtualiza-
tion and management, traffic engineering, cloud computing,14,15 or business applications used. The prediction model can
be trained, which predicts the location of vehicles based on historical information. Another example is an analytic appli-
cation that can be incorporated to recognize suspicious network activity. Application layer provides an open interface for
optimizing network function.
The advantages of integrating SDN into VANETs are due to the decoupling of the three layers which are mentioned
above. APIs such as northbound, southbound, eastbound and westbound are implemented to connect this decoupled
layers with each other. The northbound API is implemented between the controller (control plane) and applications
(application plane), while the southbound API is implemented between the controller and the physical networking
hardware (data plane layer or infrastructure plane).

3 SDVN: A BIG PICTURE

Recently, SDN-enabled vehicular networks (SDVNs) have witnessed substantial upliftments from technical and archi-
tectural aspects. Many researchers have implemented and proposed SDVN in different ways. This work has tried to
encompass the most recent developments in SDVN. The intuition behind this is to provide the perspective knowledge and
insights of existing SDVNs. This section highlights some of the significant work done till date in the field of SDVN which
is classified into mainly three categories such as routing protocols, architectural, and integration with new technologies.
The previous section highlighted the architecture and components for SDVN in which almost all the SDVN comprises
three planes. The existing SDVNs do not differ in these planes, but they vary based on how these planes are implemented
in the real scenarios, the set of protocols, and the architectural strategies of SDVN. This section focuses on the existing
SDVNs based on key attributes as per the taxonomy mentioned in Figure 4.

3.1 Architectural modes and access technologies

Table 3 shows the operating details of SDVN where access technologies represent the communication technologies used in
the SDVN system. The architecture attribute represents the mode in which the SDVN system operates. The SDVN system
supporting the fallback mechanism confirms that even if the links do not present between vehicle and infrastructure, the
network will remain active using the pre-functioned stack of protocols and the absence of it, will force the network to be
crashed.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 of 22 BHATIA ET AL.

FIGURE 4 Taxonomy of software-defined vehicular network (SDVN)

TABLE 3 Architectural mode of software-defined vehicular network (SDVN) systems

Author Year Objective Access Architecture Fallback

Technologies Mechanism
Ku et al8 2014 Implementation of backup 4G (LTE), Hybrid, V2X Yes
controller when links are broken Wi-Fi
for global controller.
Sudheera et al16 2016 Reducing transmission delay for Cellular Centralized, V2X No
VANET (3G), DSRC,
802.3x
Kazmi et al17 2016 DeCentralization of SDVN to 3G/LTE, Wi- Decentralized No
manage scalability Fi, WAVE
Li et al18 2016 Balancing latency requirement and 5G, 802.11 Centralized, V2X Yes
reducing cost of network through
southbound modifications.
Duan et al19 2017 Implementation of cluster head for 5G Centra lized, V2X No
5G enabled SDVN.
Alioua et al20 2017 Deployment of multiple controllers 4G (LTE), Hierarchically Yes
for heterogeneous SDVN DSRC, Wi-Fi centralized (Semi-
(802.11p) centralized),
V2X
Din et al21 2018 Implementation of sensing layer, 5G, Wi-Fi Hierarchically No
relay layer, and core network layer centralized, V2X
in SDVN.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 9 of 22

Duan et al19 propose SDN-enabled 5G-VANET to resolve the raising traffic conditions by promoting the HetNet (Het-
erogeneous Network). In 5G-VANET, neighbouring vehicles are put under the same cluster, depending on the real-time
road situations using SDNs global information collection and network control capabilities. Because of the segregation of
information plane and control plane, 5G-VANET handles and encourages centralized control over HetNets by giving a
comprehensive network view. The road traffic is predicted to achieve the adaptive clustering mechanism in the vehicu-
lar network. From each vehicle cluster, a cluster head is elected for a particular group of vehicles. Then, a dual cluster
head system is proposed for robustness and seamless trunk link communications. Moreover, 5G-VANET uses an adap-
tive beam formed trunk link transmission scheme for better communication capacity, quality, and decreasing the traffic
distribution latency. In this work, the author has conducted MATLAB simulations for SDN-adaptive clustering scheme
over the conventional scheme of transmission in VANET. Normally, SDN-adaptive clustering provides a low bit error rate
when SNR is increased and provides better communication.
Alioua et al20 presents an adaptable SDN-based HetVNet. It is a semi-centralized architecture with hierarchical
multi-controllers available at the core of the network and a superior fall-back recovery mechanism. This architecture
captures both covered and uncovered areas. The aim is to provide feasibility, scalability, and efficiency of SDVN in both
infrastructure and infrastructure-less area. The architecture is cluster-based and maintains the hierarchy of multiple
controllers, eg, level 2 (global controllers), level 1 (central controllers), and level 3 (local controllers).
Sudheera et al16 suggests a new architecture for the delay-efficient software-defined VANET. The major requirement
of the new network is the low setup time using wireless technologies, ie, DSRC. The core approach is to select a few
prime RSUs as local SDN controllers which can provide a specified localized global view, rather than considering all. This
approach brings downs the control layer to the lowest possible level to reduce the transmission delay in VANET. Both
mathematical and practical results show that this proposed system uniquely performs better than the conventional SDVN
in case of packet delivery ratio (PDR).
Kazmi et al17 propose a decentralized SDN concept in VANET. In DeVANET, the author has exploited the control plane
by distributing the central intelligence into LCs (also named as the domain controller) and GC (also designated as core
controller). Their power of performing operations organizes the GC and LC. The domain controllers (DC) provide services
to the domain-oriented network, which processes the request. The primary goal of this architecture is to increase the
scalability of the network.
Li et al18 proposes a bandwidth modification for the southbound interface in SDVN with 5G cellular networks. Although
cellular 5G networks consume more energy and are not budget friendly, they provide far better performance compared
with ad hoc networks. The author improved a bandwidth diminishing approach to balance the cost to sustain the perfor-
mance in southbound communication. The simulation shows that the hybrid southbound communication mode achieves
the balance between network cost and performance in the software-defined VANET.
Ku et al8 proposes an architecture based on the integration of SDN with VANET. This system behaves like a hybrid
SDVN. All the nodes in VANET are connected with RSUs. But, it is not easy that all nodes are strongly connected to the
data plane through RSUs. So, the SDN module is embedded with local SDN agent. In a normal scenario, when connection
to controllers is good, then these local agents have minimal intelligence. Contrary to this, when the link between the
controller and the wireless node is broken, then the embedded local SDN agent can work as the backup controller. The
local SDN agents replace the SDN intelligence to take charge of the SDN network and return to ad hoc network mode.
The agents support traditional ad hoc routing protocols like GPSR, AODV, DSDV, and OLSR as fall-back mechanisms.
Hence, a single SDN wireless node is a data propagator as well as the data consumer. The results of the simulation show
that the PDR is higher in SDVN.

3.2 Implemented protocols

Table 4 shows the detailing of the implemented protocols in existing SDVNs, along with their support for resource
management and multiple controllers for the respective protocol.
Qi et al30 proposes SDN-enabled social-aware clustering (SESAC) algorithm for the SDN enabled the 5G-VANET system
to empower the adaptive and efficient clustering for communication and information sharing between the base stations.
The network congestion and packet loss are reduced by integrating cluster-based VANETs with 5G cellular networks. The
core logic is to build a cluster of nodes that follow the similar social match, which can have possibly the same future routes.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 of 22 BHATIA ET AL.

TABLE 4 Implemented protocols

Author Year Objective Multiple Resource
Controller Management
Ku et al8 2014 Integration of SDN into VANETs. No Moderate
Liu et al22 2015 Geobroadcast in SDVN for region wise broadcast. No Low
Kazmi et al17 2016 DeCentralization of SDVN to manage scalability. Yes Moderate
He et al23 2016 PRETTI protocol for multicasting in SDVN. No Low
Dong et al24 2016 SDN-based on-demand routing protocol (SVAO). Yes Low
Liu et al25 2016 Co-operative data scheduling in SDVN using central of RSUs. No Moderate
Ji et al26 2016 SDN-based geographic routing (SDGR) protocol in VANET. No Low
Zheng et al27 2016 Implementing SDN into heterogeneous VANET by the help of Yes Moderate
cloud radio access networks (Cloud-RAN).
Correia et al28 2017 Clustering concept for broken links in hierarchical Yes High
software-defined VANET.
Venkatramana et al29 2017 SDN-empowered connectivity-aware geographical routing No Low
protocol (SCGRP)for SDVN.
Qi et al30 2018 Social aware clustering of vehicles SESAC algorithm for 5G No High
SDVN
Luo et al31,32 2018 MAC-based SDVN and roadside-OpenFlow-switch (ROFS) No High
Huang et al33,34 2018 Data offloading for using LifeTime-based network state No Low
routing (LT-NSR)
Al et al35 2018 Lodging of mobile edge computing(MEC) in SDVN for efficient Yes Low
content delivery.
Lin et al36 2018 Improved genetic algorithm (IGA) for SDVN to balance DSRC No Low
usage.
Rayeni et al37 2018 To find optimal path in SDVN using ORUR scheme and WAVE No Moderate
connectivity.
Ghafoor et al38 2018 To find stable route in cognitive radio software defined vehicular Yes Low
networks.

Then, the cluster head is chosen depending on some attributes like distance-matrix, relative speed, vehicle specifications,
etc. The simulation shows that SESAC-based clustering performs better than other clustering-based algorithms regarding
choosing the cluster's reliability and lifespan.
Luo et al31 defines a novel SDN-influenced MAC-based protocol for providing the cooperative safety and transportation
management in VANETs. The objective is to handle the rapidly changing densities and mobility of vehicles in VANETs.
To attain the safety in VANETs, vehicles (OBUs) must broadcast cooperative awareness messages (CAM) to road-side
OpenFlow switch (ROFS) periodically,39,40 The CAM contains safety-related information, ie, id, position, and speed.
This is achieved by replacing TDMA-based MAC protocol (eg, STDMA and VeMAC), with ROFS-implemented sdnMAC.
The ROFSes cooperatively share the slot information through the data plane, detect impending collisions, and manage
vehicles within its communication range.32 The management of ROFSes (MA-ROFS) schedules the cooperative sharing
of time slot information among ROFSes based on vehicle densities, and the management of vehicles (MA-VEH) uses the
slots allocated by MA-ROFS. 41 The simulation results yield the shortest access time, lowest packet loss rate and access
collision.
Huang et al33,34 proposes SDN with mobile edge computing (MEC) architecture for data offloading between two vehicles
in V2V infrastructure. The LifeTime-based network state routing (LT-NSR) algorithm has been developed and executed
in the SDNi-MEC server to find the true V2V routing path that will last for the maximum amount of time in the existing
network topology. The core idea is to recover a broken V2V using LifeTime-based path recovery (LTPR). The simulation
result shows that the proposed LT-NSR algorithm performs better than the GD-NSR scheme for the moderate vehicle
density situation. And for low and high vehicle density situations, LT-NSR and GD-NSR performs the same.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 11 of 22

Al et al35 proposes a software-defined edge computing infrastructure for VANET to support the efficient content delivery
between connected vehicles and to mitigate the WAN-latency problem. The aim is to deliver the services with minimum
latency. In this work, roadside units or base stations are suggested to lodge mobile edge computing (MEC) services nearer
to the vehicles. The V2V communication uses vehicle-level caching, and V2I communications use a cooperative search
strategy. The prototype is implemented in Mininet WiFi emulator. The simulation results show the efficacy of the proposed
addition in VANET.
Lin et al36 proposes an improved genetic algorithm (IGA) to optimize the dynamic network changes. The IGA includes
the solution for adjusting the dynamic change often occurring in the network and also give the surety about solution
diversity. The main idea is to balance the use of DSRC, ie, V2V and V2I traffic for minimizing the latency. Through the
simulation, the number of packets transmitted through cellular lines is positively correlated with the rebate ratio and the
other parameters. Moreover, the IGA implementation can potentially improve the performance in case of searching the
optimal solutions.
To serve the need of rapidly changing the dynamic topology of VANET, Rayeni et al37 proposes a new routing scheme
named the optimal resource utilization routing (ORUR). It is a cloud-based routing approach aiming at balancing the
load among the communication paths and improving the QoS related to channel busy ratio.
Correia et al28 proposes a solution named as the hierarchical software-defined VANET (HSDV) for the situations in
which connection is broken between SDN controller and vehicles. The main logic is to build clusters of vehicles that create
independent local domains. The HSDV makes clusters in infrastructure, which enables the network to sustain a working
state regardless of central command given by the controller. The maximum rate of delivery with minimum overhead and
delays is achieved leading to the higher network throughput. The simulation result shows that HSDV provides flexibility
and programmability and outperforms the conventional routing protocols (like AODV, DSDV, GPSR) by improving the
overall performance.
Venkatraman et al29 proposes the SDN-empowered connectivity-aware geographical routing protocol (SCGRP) in
which the routing calculation is shifted to the global controller to calculate real-time routing measurements for benefi-
ciary vehicles using cellular technology. The SCGRP reduces link breakages and transmission delay and gains more PDR.
It also takes care of heterogeneous traffic conditions. In this work, the author has made a comparison between SCGRP
and centralised routing protocol (CRP), both theoretically and practically. The simulation result clearly states that SCGRP
is better than CRP in terms of packet delivery ratio, average hop count, and end-to-end delays.
He et al23 proposes a protocol named Pretti, which aims to reduce the network management overhead and commu-
nication cost. The network management overhead is reduced by simplifying the network topology information and also
by predicting the network topology of the near future. It minimizes the frequent collection of the vehicular statistical
information. The main idea is to apply the centralized multicast scheduling decision over the SDN-enabled heteroge-
neous VANET. The simulation result shows that the proposed scheduling algorithm outperforms the traditional multicast
protocol AODV known as MAODV in terms of successful delivery ratio, delivery cost, and mean latency.
Liu et al25 proposes a novel model for the cooperative data scheduling (CDS) under V2I and V2V communication
environment in SDVN. The RSUs in this model is equipped with the central scheduler. Each vehicle sends its current
neighbouring vehicle's list and the identifiers of the retrieved and newly requested data to the RSUs. Then, RSU chooses
sender and recipients with related information for V2V communication, while at the same time it sends a data item acting
as a command to vehicles for utilizing the I2V channel. The cooperative data dissemination (CDD) algorithm has been
used for performing CDS. For performance evaluation, the author has implemented three algorithms that are first come
first serve (FCFS), most requested first (MRF), and CDD.42 From the simulation result, it shows that service ratio in CDD
is more than FCFS and MRF. The CDD also performed significantly well in heavy traffic conditions.
Ji et al26 proposes an SDN-based geographic routing (SDGR) protocol for VANET, which depends on vehicle density
and location. In SDGR, two algorithms are defined out of which first is the optimal forwarding path algorithm for finding
the shortest forwarding path with shorter roadway length and higher vehicles density, utilizing the global view of SDN
controller. The second is the packet forwarding algorithm focusing on a congestion detection mechanism to address the
load balancing. The packet forwarding algorithm also chooses the optimal vehicle for the next hop in V2V infrastructure.
The simulation results show that SDGR achieves better performance in packet delivery ratio and delivery delay time
compared with AODV and GPSR.
Zheng et al27 proposes a novel framework named SERVICE (Soft-dEfined heteRogeneous VehICular nEtwork). The
author implemented SDN into heterogeneous VANET with the help of cloud radio access networks (Cloud-RAN). The
infrastructure plane is enabled with remote radio head (RRH), which helps in minimizing the communication cost.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 of 22 BHATIA ET AL.

TABLE 5 Opportunities for trending technologies in SDVN

Author Year Objective Integrated Technology
Truong et al44 2015 SDN enabled VANET with fog computing Fog computing
(FSDN)
Salahuddin et al46 2015 Development of RSU as CRM and micro Cloud computing
data centres for IoV.
Sahoo et al47 2016 Ferrying vehicular data in clouds-vehicle-clouds Cloud computing
through SDVN
Zhao et al48 2016 SVM-based routing scheme for VANET using Machine learning
greedy forwarding algorithm
Xiao et al49 2017 Minimizing delay constraint and fault tolerant Fog computing
mechanism in SDVN
Soua et al45 2018 Offloading of computing services in 5G SDVN. Cloud computing and fog computing.

The SERVICE has hierarchical control plane equipped with the primary and secondary level of controllers and provides
network virtualization function to its application plane.
Liu et al22 implements the SDN-enabled VANET to broadcast the information in a specific region automatically. With
the use of IP addresses, geobroadcast can be implemented in the current ITS. The controller will use the topological and
geographical information to select the routing paths for the packets to be sent to next (or last) RSUs, by creating appropri-
ate flow entries on the related RSUs and mediator switches. The geobroadcast mechanism shows a significant reduction in
latency, controller overhead, and network bandwidth consumption. The simulation results state that controller overhead
is reduced to 84% and network bandwidth consumption is reduced to 60% in ITS using this mechanism.

3.3 Opportunities for trending technologies

Table 5 furnishes the list of trending technologies which are integrated with SDVN to extends its capabilities and enhanc-
ing the reliability. Nobre et al43 proposes the design principle for fog-enabled vehicular software-defined networking
(VSDN). The aim is to increase efficiency in every aspect of vehicular networks. Truong et al44 proposes a novel SDVN
architecture called FSDN that integrates two emerging technologies, which are SDN and fog computing into VANETs. The
proposed concept leverages all the SDN advantages like scalability, flexibility, programmability, and fog-computing offers
delay-sensitive, context-sensitive, and location-awareness services which could satisfy the demands of future VANETs
scenarios. Contrary to the above two, Soua et al45 proposes vehicles compatible with fog computing aiming to improve
the QoS of vehicular applications and reducing the computation and communication complexities in VANETs.

3.4 Implementation
Table 6 provides the brief information about various SDVN scenarios and simulation tools used. Over half-decade, there
are lots of tools and technologies used by researchers to implement the SDN concepts along with VANET system. Many
researchers in previous studies50–53 have shown their efforts to implement SDN capabilities into VANET using various
technologies such as SUMO,54 OMNET++,55 VEINS,56 Mininet-WiFi,57 Network simulator 3,58 Floodlight,59 ONOS,60
POX61 controller, etc. Fontes et al50 have used Mininet modules for node car architecture, OpenFlow-enabled controller
is implemented to manage cars, RSUs, and eNodeBS. The node cars have root spine switch installed for LTE and Wi-Fi
communication. Here, node car captures a short duration video for safety purpose, video surveillance, or traffic analysis is
done using the onboard camera. Then, the SDN controller chose proper routing path and communication technology for
that captured video to send over the central repository. Berk et al52 uses a Raspberry Pi module as an OpenFlow-enabled
RSU. The open daylight SDN controller is connected with all the RSUs through wired LAN. Furthermore, RSUs are
connected with wireless internet access points (AP) through Wi-Fi. Here, instead of cars, smartphones are considered as
mobility nodes. The smartphone connects to the wireless internet with which RSUs are linked.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 13 of 22

TABLE 6 Simulation details of various SDVN systems

Author Year Simulated Scenario Simulation Tool(s) Controller Module
Ku et al8 2014 SDN-enabled VANET. SUMO & NS-3 NS-3
Liu et al22 2015 SDN-based geobroadcast in VANET. OpenNet62 (Mininet- Floodlight
WiFi & NS-3)
He et al23 2016 Multicast protocol for SDVN. SUMO, Python & NS-3
NS-3
Dong et al24 2016 On demand routing for SDVN. SUMO & NS-3 NS-3
Ji et al26 2016 Geographic routing for SDVN. SUMO & NS-2 NS-2
He et al10 2016 Heterogeneous vehicular communication SUMO & NS-3 POX
in SDVN.
Li et al63 2016 Network virtualization optimization for SUMO, Veins & OMNET++
SDVN. OMNET++
Correia et al28 2017 Hierarchical SDVN. SUMO & NS-3 NS-3
Alioua et al20 2017 SDN-based HetVNet. SUMO & NS-3 NS-3
Duan et al19 2017 Adaptive vehicle clustering for 5G-SDVN. MATLAB MATLAB
Secinti et al52 2017 Wireless access management (WAM) in Open vSwitch, OpenDayLight
SDVN Rasberry Pi
Hans et al53 2017 Wireless SDN mobile ad hoc network. Open vSwitch, ONOS
Rasberry Pi 3B+
Venkatramana et al29 2017 Geographical routing for SDVN. SUMO & Mininet- OpenDayLight
WiFi
Al et al35 2018 SDN-based HetVNet. SUMO & Mininet- Mininet-WiFi
WiFi
Luo et al31 2018 SDN-based MAC protocol for VANETs. SUMO & NS-3 NS-3
Rayeni et al37 2018 ORUR for heterogeneous SDVN. SUMO, Veins & OMNET++
OMNET++
Din et al21 2018 5G-based SDVN for ITS. C language Bravo 180 S
Atwal et al64 2018 SDN-based mobility management and FlowVisor, Mininet- ONOS
QoS support for VANETs. WiFi, Open vSwitch
& NS-3

4 USE CASES

Use cases of SDVN systems range from safety services to data dissemination and network virtualization. The versatility,
real-time, highly responsiveness are some of the reasons behind its extensive set of applications. Adequate implementation
of SDVN system can usher the industrial revolution in its field. ITS is mainly powered by vehicular networks. The ITS
alone consists of so many concise applications that it can change the automobile industry. Figure 5 demonstrates already
various implemented use cases of SDVN.
• Efficient data dissemination: Highly customized data dissemination of SDVNs is one of the salient features. The
data dissemination gives a soar to many services such as emergency broadcast services, adaptive broadcast interval
time, security of users, etc. Nguyen et al65 develop an adaptive beacon-based data dissemination (ABDDis) protocol
which disseminates the warning messages to their nearby vehicles as soon as a dangerous situation occurs without
throwing an extra load on the channel. Hasrouny et al66 proposed a novel application that handles trustworthi-
ness evaluation, misbehaviour detection, and revocation process. It is based on the secure and modular public key
infrastructure (PKI), vehicular groups, hybrid trust model, and CRL management system.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
14 of 22 BHATIA ET AL.

FIGURE 5 Various use cases of software-defined vehicular network (SDVN)

• Safety services: The functionalities of SDN in VANET enhances road safety using V2V and V2I communications.
The SDN can secure or restrict particular frequencies or channels for later use. Emergency traffic or other privileged
activities can utilize that reserved frequencies.8 The main advantage of SDN-enabled VANET over conventional
emergency channels is that the allotment of frequency can be done dynamically. The SDN controller can reserve
channels for emergency services based on traffic conditions and application necessities. Similarly, these chan-
nels can be utilized to accommodate various application or services. The emergency messages such as cooperative
awareness messages (CAM) can avail the benefit of this reserved frequencies.31 Various privacy schemes can also be
implemented to improve two safety services, ie, lane change warning and forward collision.67
• On demand surveillance service: On-demand surveillance is another service which is made available for the
authorized client of an SDVN. This service can be used in emergency or investigation scenario. To use this service,
an authorized node (e.g., a police car) must send a surveillance request. The controller services this request. In reply
to this request, the controller embeds the flow rules for the surveillance data to reach the requesters. When there are
more than one requests for the identical information, such as many requester nodes request for video surveillance
feed, the controller creates entries of rules in flow table so that the same copy is forwarded to many requesters.8 This
is also applicable to infotainment services like video streaming, video conferencing, media sharing, etc.42
• Virtualization of wireless network: Network virtualization services expects to give abstract and informative
virtual networks over shared physical network assets. The data centres are utilizing SDN to provide network
virtualization services, and such an idea can also be applied to SDN-enabled VANETs. When individual network
has used distinct frequencies, then individual networks traffic are separated from each other. Author has tried to
effectively chop the network into the network fragments and created virtual wireless networks. By dynamically
grouping wireless node for RSUs, overhead can be reduced. Now, each RSU only propagates traffic from its group.
The encouragement behind this idea is to support the time slicing. An SDN controller governs the network and time
period of its respective radio interface/frequency, which allocates the network traffic in a programmable manner.
Furthermore, the controller has a facility of setting input filters over wireless nodes, and those nodes can only accept
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 15 of 22

those traffic as an input for which they are allowed. The main application of input filtering is to stop interference
between the two services.63
• Dynamic air quality monitoring: With the fast-growing automobile industries, pollution in the air by the emis-
sion of gases is increasing day by day. The Air Quality Index is the measurement of the polluted air. In some countries,
government or independent bodies are measuring AQI for various purposes. The vehicular sensor network (VSN)
is implemented to monitor the air quality. 68 To make it more solution-oriented, deployment of IoT-based vehicular
pollution monitoring system is trending69 but it still lacks the global view and some state-of-the-art functionalities.
This can be overcome by deploying SDN-enabled RSU in VANET. An SDN-enabled RSU is used to provide central
access to raw data collected by all the RSUs. All the RSU can work as commanded by the central controller. The raw
data gathered can further be processed using various data analytics techniques. The output from processed data can
benefit the health sector, environment sector, infrastructure sector, etc.
• Smart parking: It is one of the trending technology under the hood of smart city applications. The smart parking
is the combination of IoT devices, access technologies, and an ad hoc network. The conventional network will fail
when the vehicle with IoT devices are overnumbered, and scalability leads to network crash and will not allow
various services to deploy simultaneously. The VANET with SDN will be able to eliminate all problems by enabling
network flexibility.42 The Nokia has released a new network based on SDN framework and provides different services
such as software-defined wide area networking (SD-WAN), virtualized cloud services (VCS), and virtualized security
services (VSS).
• Heterogeneous support: In the modern communications system, heterogeneity is the common thing. But, tra-
ditional vehicular networks use IEEE 802.11p (Wi-Fi) and cellular communication depend on either V2V or V2I.
However, novel VANET applications require more efficient and cost-effective communication technologies to reduce
communication cost. One of the known and cost-effective technology is DSRC but, it is not suitable for V2I
communications as its wireless range is very short. On the other hand, for long-range communications, 4G/5G cel-
lular networks are used but, they are costly. Other than this, Wi-MAX, WAVE, and UMTS are also available. In
SDN-enabled VANET, researchers are free to choose more than one access technologies, for the different layers of
SDVN, which enable heterogeneous access and improve the usage of bandwidth and channel.20,42
• Latency-based routing: SDN-based VANET effectively considers the latency when congestion on particular routes
increases. In such cases, all the other traditional routing protocols, ie, border gateway protocol (BGP), interior gate-
way routing protocol (IGRP), and routing information protocol (RIP) becomes useless as they are not able to give
routing decision effectively and quickly to the changing topology. This new SDN paradigm dynamically responds to
latency and queuing delay. In the case of voice traffic, a safety-related or real-time application, which cannot tolerate
delays and jitter are supported using minimum latency path.42
• Lane change assistance: Without safety applications like lane change warning, changing of the lane may result
in deadly accidents. An existing lane changing system has to pose many challenges such as direction and speed of
vehicles, road conditions,70 traffic densities, etc. Since these are unable to work multi-lane change task. The con-
trollers of SDN can address this problem because of its global view over the road network. The controller commands
RSU to gather information such as speed of vehicles, density, map, and lane trajectory. When the lane change is
requested by vehicle, the controller takes a decision depending on the information gathered by RSU. The DSRC is
also an essential technology for application of lane change assistance.42,71
• Traffic accident detection: Dogru et al72 analyses a real-time traffic accident detection method and combined a few
of the machine learning techniques with V2V communication. On the basis of speeds and coordinates of vehicles,
the data is collected from VANETs. Since the available data is according to speed and coordinates of every vehicle,
machine learning algorithms are used to analyse the vehicle's behaviour and to predict the accident if it is bound
to happen. The artificial neural networks, support vector machine, and random forest are the algorithms that are
employed on the data set obtained to distinguish the normal case from the accident-prone case. The Random Forest
algorithm turned out to be more efficient than the other two with the accuracy of 91.56%.
• Bandwidth management: Drastic increment in the applications of SDVN increases the demand of bandwidth
requirement. The existing network's architecture and features do not support the dynamic bandwidth demand. SDN
paradigm makes available the cost-effective bandwidth-on-demand (BWoD) in the highly changing environment
of VANETs. An OpenFlow enabled SDN allows consumers to avail the bandwidth-on-demand services allowing
vehicles to span and shrink their bandwidth requirements dynamically. This adaptiveness can be implemented using
network devices such as RSUs, routers, switches, and bandwidth scheduler installed on SDN controller. Though,
efficient allocation of bandwidth is also a challenge.42,73,74
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
16 of 22 BHATIA ET AL.

4.1 Case study: Overlay virtual network for traffic management

When it comes to the handling of big VANETs, there is a functionality available in SDN called network function virtu-
alization (NFV), which can provide robustness to big VANETs. The NFV can avail the maximum benefits from limited
system resources of VANET. Basically, traffic management in VANET depends on how efficiently the data can be dis-
tributed and forwarded into the large VANET. By leveraging the control plane in SDN, the system can effectively collect
and maintain individual vehicle states in a logically centralized way. In actual, the flow table of control plane stores each
vehicle's information and based on that controller optimizes V2V/V2I multi-hop routing/switching decisions. The RSUs
can serve as fog devices so that when routing information is requested by the vehicles, the information is made available
to them.42,75 An SDN makes the behaviour of the network devices (such as routers/ switches) programmable and allows
them to be controlled by a central element, thus offering advanced customizability of network control and forwarding
behaviours. When SDN is incepted into highly dynamic VANET to serve the various demanding needs of ITS, the new
concept called Internet of Vehicles (IoV) arises. An IoV is the evolved form of VANET and it inherit many characteristics
from SDVN including architecture. However, there can be add-ons to the application layer as per the implementation of
IoV. IoV is more powerful, yet more difficult to implement. The centralized SDN controllers manages the data forward-
ing in IoV. If the SDN controllers takes control of the resource management in IoV then it can serve a large number of
vehicles. There will be flexibility, easy implementation of network function virtualization (NFV), light-weight onboard
hardware, and service-oriented resource allocation. The general objective of the case study is to decide how data is for-
warded among the vehicular nodes. An SDN can provide dynamic connection management to IoV. For managing the
connections, there is the control process defined and it executes in two phases. The first phase is involved in controlling
the messages in which vehicles report status (ie, location and request) to RSU in sequence. Next, RSU will contact the
controller and the controller provides the decision to RSU. Then, RSU disseminates the decision to all the vehicles under
the coverage. The second part involves the data transmission. In this, the actual data transmission takes place between
V2R, V2V, RSU to fog, and RSU to cloud. SDN provides the overlay vehicular network (OVN) on the top of the underlay
vehicular network through its network virtualization. The underlay physical network comprises of vehicles and RSUs
and a subset of underlay nodes forms the OVN. Dedicated resources are allocated to the OVN and one OVN is created
per service. An OVNs should not interfere with each other. Figure 6 shows the concept of overlay networks on the top of
the underlying physical network. The overlapping OVNs operate on different channels. There are topology restrictions
for OVNs. The permitted topologies are single-hop unicast (sender-receiver), single-hop-multicast (sender- receivers),
dual-hop unicast (sender-helper-receiver). Thus, OVNs provide multiple concurrent request support, flexible topologies,
and efficient resource utilization. Moreover, OVN concept can be used with some other technologies such as fog com-
puting, cloud computing, and machine learning. There are many situations that can lead to congestion of vehicles in
VANET. The severe weather condition is one of the situations that cause traffic in VANET. Koesdwiady et al76 proposes
an approach that applies to the traffic flow based on weather conditions and data fusion technique to enhance the accu-
racy of traffic flow prediction. Using machine learning, Cheng et al77 proposed vehicle grouping for traffic scheduling
instead of traffic lights. In a smart transportation system, efficient and flexible traffic scheduling is a must. For grouping
and schedule, the neuro-fuzzy network is proposed where the network is trained using a reinforcement learning tech-
nique. The results show that fuzzy group learning reduces average waiting time (AWT) and waiting time variance (WTV)
compared with traffic light algorithms.

FIGURE 6 Overlay virtual networks

 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 17 of 22

FIGURE 7 Challenges for the futuristic

software-defined vehicle network (SDVN)
system

5 OPEN ISSUES A ND RESEARC H CHALLENGES

Though SDVN is rapidly evolving, still there are lots of open issues that are questioning about the efficiency, scalability,
and reliability (trustfulness) of it. These challenges may decide the future direction in the development of the SDVN.
Figure 7 shows the enlisted challenges for the existing set of SDVN systems.
• Management of rapidly changing SDVNs: High mobility of nodes (vehicles) causes rapid and unpredictable
changes to the network topology of SDVN. It is arduous for SDN controllers or RSUs to control the vehicles and
unstable communication channels in real time. As DSRC or WAVE connectivity are not able to tolerate significant
fluctuations between speeds in V2V infrastructure, there are more chances of links broken. However, this process
can be handled by some efficient routing algorithm and modifications in the infrastructure but, it becomes costly.
• Security of SDVN: Security is still the major fear that is preventing the SDVN from becoming widely acceptable.
Many of the SDVN applications follow a centralized approach where the SDN controller is the main unit that is
responsible for the functioning of the whole network.12 It is easy to strike on the single controller to make the whole
network down. A malicious user can intermeddle into the system and can take decisions instead of the controller.
Through such incursions, networks can be manipulated or it can put users' safety into risk. Tanwar et al78 proposed
a survey of a detailed taxonomy based upon the types attacks. Furthermore, DDoS attacks can overflow the flow
table with fake entries, which makes the network down and unreachable for real users. So, it is essential to protect
controller.79 Hasrouny et al80 listed many types of attacks in four main categories. First is threats to wireless interface
such as location tracking, DDoS, Sybil attack, malware, spam, man-in-the-middle attack, brute force attack, and
black hole attack. Second, to hardware and software components of the architecture, ie, injection of wrong messages
(bogus info), message suppression or alteration, usurpation of the identity of a node (spoofing or impersonation or
masquerade), tampering hardware, routing network, GPS spoofing, tunneling attack, and grey hole attack. The third
is a menace to the hardware, such as the sensors and transceivers, of the vehicle. Illusion attack and jamming attack
are in the list of the third category. Fourth and last is threats to infrastructure, ie, session hijacking, repudiation (loss
of event traceability), and unauthorized access.
• Latency control in SDVN: In SDVN, latency is a very unpredictable constraint. As it is wireless service, one cannot
ensure that after what time data will be received. Nonetheless, latency control can be achieved by optimizing some
or all the other aspects of network performance by resource management. The latency control is directly dependable
on optimization of resources. Nowadays, cloud computing approaches are trending as they are more optimized. But,
the cost of using cloud computing in VANET gets expensive as the number of vehicles starts growing exponentially.
The reasons behind the cost are operations such as passing user information to cloud database, collecting user infor-
mation regarding channel conditions, computational workload in the cloud, QoS requirements, location tracking,
providing the optimal resource access to users, etc. These all introduces latency in the network. Therefore, latency
control should be a majorly focused topic for the futuristic networks.81
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
18 of 22 BHATIA ET AL.

• Scalability of architecture: Scalability of existing SDVNs is the key element as the automobile industry is over-
growing. The chances of unanticipated obstructions and sudden changes are unpredictable while traveling. There
are many factors such as technical up-gradation, complex road topologies, infrastructure damage, etc, which can
affect the performance of SDVN. The increasing number of vehicles followed by an increased number of communi-
cations can also degrade the performance of low scalable SDVNs. Furthermore, architecture should be able to resolve
the demanded services and diverse weather conditions. This results in an inefficient resource management50,82 and
degradable network performances.83 One more consideration is disaster management. When some broken infras-
tructures because of the disaster are not capable of transmitting packets, then the SDN controller should intelligently
assign other functionalities to help, to uphold the QoS parameters.
• Heterogeneous network: Another challenging task for SDN-enabled VANET is interworking gaps among
heterogeneous networks as the future of SDVN is not limited to communication between vehicles. It will extend to
new technologies and devices having multiple features and are from different manufacturers. Hence, the problem
of mutual exclusiveness occurs, and communication between vehicles can fail. The standardization of technolo-
gies is a solution to overcome this issue. Because of a large number of vehicles, some common problems such
as collision, long delays, increased packet loss rates, interference, and noises can occur.83 Therefore, existing
heterogeneous V2X networks requires efficient interworking mechanisms. As per the real world scenarios, the
design of SDVN architecture can be decided based upon the heterogeneity of RSUs. For example, in the urban
environment where aggressive traffic is common, SDVN architecture requires many novelizations of the network
topology. Therefore, the generating capacity of network topologies must be improved. For the highway application
of SDVN implementations, topology prediction is not critical tasks for RSUs as the network topology does not differ
a lot. In the city side, there are fewer vehicles, and hence, the RSUs can lessen their ability of computing multi-hop
transmissions.
• Trustworthiness evaluation, misbehaviour detection, and revocation process: Trustworthiness evaluation of
the participating nodes in VANET is an open problem. Any wrong step in evaluating vehicle can put the lives of users
in danger. Hence, concrete parameters for deciding whether a vehicle is trusty or not are still lacking. Researchers
have proposed a method for misbehaviour detection,84 but no approaches are present for utilizing this information.
The penalties or fine are not defined for malicious vehicles. There is still no work implemented for the revocation
process to be undertaken if any node is misbehaving in the network. The certificate revocation list (CRL)-based
solutions85 are still undergoing the development phase. Using the limited time certificates for CRL and certificates
modification, strategies are not yet defined and are still vulnerable under no infrastructure for CRL. The certificate
verification and revocation are longer processed. So, alternatives should be found. Many researchers in previous
studies86,87 have proposed approaches to replace CRL.
• Demarcation of SDN into VANET: The procedure of defining the boundaries to the extent in which the one can
integrate the SDN into VANET, is yet not defined. Only changing the wired part is not true integration. On the con-
trary, converting VANET into an SDN-based network is not a wise option. There are higher chances of vulnerability
when any SDVN configured with all the three layers of SDN. Such converged systems can introduce single point of
failure and software vulnerabilities. Generally, infrastructure plane in VANET contains large number of vehicles. So,
there will be more number request from infrastructure plane to control plane and these situation can cause perfor-
mance problem. Because of separation between the infrastructure and control plane in SDN, the process of designing
a good response strategy is being possible. This strategy can create more decisions to resolve the disturbed and upset
traffic flow and propose adaptable path selection policy. SDN can offer a channel allocation and cognitive radio pol-
icy which permits high-bandwidth and low-latency communication. Implementation of southbound interface (SBI)
should be standardized and vender-agnostic because it is important how infrastructure plane communicates with
control plane.44 A full-fledged assessment of vehicular network can identify which component of the control layer
can be removed from the infrastructure plane to take the full advantages of SDVN. However, the lacking of such
real-world implementations and simulation tools impede such evaluation. In the future, serious attention must be
given to the development of real-world implementations and simulation tools for the performance evaluation of
vehicles under SDN to determine which component of control layer should be decoupled from the infrastructure
plane.12

Furthermore, there are some challenges such as full-featured OpenFlow protocol stacks, standardization of protocols,
physical-virtual integration, core services, lacking eastbound/westbound APIs, adaptive network reconfiguration, overall
reproducibility, practical, mature implementation, deployable code, etc.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 19 of 22

6 CO N C LU S I O N

In this paper, we provide insights to the readers about the future of the SDVNs in the 5G era. The survey is divided into four
parts. The first part discussed the convergence of SDN and VANETs in detail. SDN convergence with VANET is analysed
based on year-wise improvements and innovation in 5G networks. Next, the architecture components such as infras-
tructure plane, control plane, and the application plane are analysed. The second part discussed the architectural modes
and access technologies of SDVN, implementation protocol, opportunities for trending technologies, and various imple-
mentation scenario for SDVN. The comparative analysis of the existing state-of-the-art approaches is performed based
on problem formulation and technology used to handle the smart applications through SDVN using 5G networks. The
third part demonstrates the suitability of SDN for vehicular applications through various use cases of SDVN. Finally, open
issues and challenges of SDVN are highlighted in the fourth part. The open issues and challenges covered in this paper can
provide the future directions to researchers. Real-Time monitoring of road traffic and the communication infrastructure
applications managed by SDVN can be motivation for the future direction.

ORCID
Jitendra Bhatia https://orcid.org/0000-0002-2375-5057
Yash Modi https://orcid.org/0000-0003-0036-1570
Sudeep Tanwar https://orcid.org/0000-0002-1776-4651
Madhuri Bhavsar https://orcid.org/0000-0003-3576-9947

REFERENCES
1. Jiang T, Alfadhl Y, Chai KK. Efficient dynamic scheduling scheme between vehicles and roadside units based on IEEE 802.11 p/wave
communication standard. In: 2011 11th International Conference on Its Telecommunications (ITST). St. Petersburg, Russia IEEE;
2011:120-125.
2. Bilgin B, Gungor V. Performance comparison of IEEE 802.11 p and IEEE 802.11 b for vehicle-to-vehicle communications in highway, rural,
and urban areas. Int J Veh Technol. 2013;2013. Research Article (10 pages), Article ID 971684 (http://dx.doi.org/10.1155/2013/971684)
3. Margaret PP. What is software-defined networking (sdn)? - definition from whatis.com. http://searchsdn.techtarget.com/definition/
software-defined-networking-SDN
4. Govindarajan K, Meng KC, Ong H. A literature review on software-defined networking (SDN) research topics, challenges and solutions.
In: 2013 Fifth International Conference on Advanced Computing (ICoAC); 2013; Chennai, India:293-299.
5. Rawat DB, Reddy SR. Software defined networking architecture, security and energy efficiency: A survey. IEEE Commun Surv Tutorials.
2017;19(1):325-346.
6. Virtualization N. What the definition of software defined networking? https://www.sdxcentral.com/sdn/definitions/what-the-definition-
of-software-defined-networking-sdn/
7. Akram Hakiri PB. Software-defined networking: Challenges and research opportunities for future internet. http://www.dre.vanderbilt.
edu/~gokhale/WWW/papers/COMNET14_SDN.pdf
8. Ku I, Lu Y, Gerla M, Gomes RL, Ongaro F, Cerqueira E. Towards software-defined vanet: Architecture and services. In: 2014 13th Annual
Mediterranean ad Hoc Networking Workshop (Med-Hoc-Net). Piran, Slovenia: IEEE; 2014:103-110.
9. Zhu M, Cai Z, Xu M, Cao J. Software-defined vehicular networks: oppor-tunities and challenges. In: Energy Science and Applied Tech-
nology: Proceedings of the 2nd International Conference on Energy Science and Applied Technology (ESAT 2015). Wuhan, China: CRC
Press; 2015:247.
10. He Z, Cao J, Liu X. Sdvn: Enabling rapid network innovation for heterogeneous vehicular communication. IEEE Netw. 2016;30(4):10-15.
11. Mendiola A, Astorga J, Jacob E, Higuero M. A survey on the contributions of software-defined networking to traffic engineering. IEEE
Commun Surv Tutorials. 2017;19(2):918-953.
12. Yaqoob I, Ahmad I, Ahmed E, Gani A, Imran M, Guizani N. Overcoming the key challenges to establishing vehicular communication: is
SDN the answer? IEEE Commun Mag. 2017;55(7):128-134.
13. Zekri A, Jia W. Heterogeneous vehicular communications: a comprehensive study. Ad Hoc Netw. 2018;75:52-79.
14. Jiang B, Huacón CF. Cloud-based smart device for environment monitoring. In: 2017 IEEE Conference on Technologies for Sustainability
(SUSTECH). Phoenix, AZ, USA: IEEE; 2017:1-6.
15. Hayes B. Cloud computing. Commun ACM. 2008;51(7):9-11.
16. Sudheera KK, Ma M, Ali GMN, Chong PHJ. Delay efficient software defined networking based architecture for vehicular networks. In:
2016 IEEE International Conference on Communication Systems (ICCS). Shenzhen, China: IEEE; 2016:1-6.
17. Kazmi A, Khan MA, Akram MU. Devanet: Decentralized software-defined vanet architecture. In: 2016 IEEE International Conference on
Cloud Engineering Workshop (IC2EW). Berlin, Germany: IEEE; 2016:42-47.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
20 of 22 BHATIA ET AL.

18. Li H, Dong M, Ota K. Control plane optimization in software-defined vehicular ad hoc networks. IEEE Trans Veh Technol.
2016;65(10):7895-7904.
19. Duan X, Liu Y, Wang X. SDN enabled 5G-VANET: adaptive vehicle clustering and beamformed transmission for aggregated traffic. IEEE
Commun Mag. 2017;55(7):120-127.
20. Alioua A, Senouci S-M, Moussaoui S, Sedjelmaci H, Boualouache A. Software-defined heterogeneous vehicular networks: taxonomy and
architecture. In: 2017 Global Information Infrastructure and Networking Symposium (GIIS). St. Pierre, France: IEEE; 2017:50-55.
21. Din S, Paul A, Ahmad A, Ahmed SH, Jeon G, Rawat DB. Hierarchical architecture for 5g based software-defined intelligent transportation
system. In: IEEE Infocom 2018-IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). Honolulu, HI, USA:
IEEE; 2018:462-467.
22. Liu Y-C, Chen C, Chakraborty S. A software defined network architecture for geobroadcast in vanets. In: 2015 IEEE International
Conference on Communications (ICC). London, UK: IEEE; 2015:6559-6564.
23. He Z, Zhang D, Zhu S, Cao J, Liu X. Sdn enabled high performance multicast in vehicular networks. In: 2016 IEEE 84th Vehicular
Technology Conference (VTC-FALL). Montreal, QC, Canada: IEEE; 2016:1-5.
24. Dong B, Wu W, Yang Z, Li J. Software defined networking based on-demand routing protocol in vehicle ad hoc networks. In: 2016 12th
International Conference on Mobile Ad-Hoc and Sensor Networks (MSN). Hefei, China: IEEE; 2016:207-213.
25. Liu K, Ng JK, Lee V, Son SH, Stojmenovic I. Cooperative data scheduling in hybrid vehicular ad hoc networks: VANET as a software
defined network. IEEE/ACM Trans Netw (TON). 2016;24(3):1759-1773.
26. Ji X, Yu H, Fan G, Fu W. SDGR: An SDN-based geographic routing protocol for VANET. In: IEEE; 2016; Chengdu, China:276-281.
27. Zheng K, Hou L, Meng H, Zheng Q, Lu N, Lei L. Soft-defined heterogeneous vehicular network: architecture and challenges. IEEE Netw.
2016;30(4):72-80.
28. Correia S, Boukerche A, Meneguette RI. An architecture for hierarchical software-defined vehicular networks. IEEE Commun Mag.
2017;55(7):80-86.
29. Venkatramana DKN, Srikantaiah SB, Moodabidri J. SCGRP: SDN-enabled connectivity-aware geographical routing protocol of VANETs
for urban environment. IET Netw. 2017;6(5):102-111.
30. Qi W, Song Q, Wang X, Guo L, Ning Z. SDN-enabled social-aware clustering in 5G-VANET systems. IEEE Access. 2018;6:28213-28224.
31. Luo G, Li J, Zhang L, Yuan Q, Liu Z, Yang F. sdnMAC: a software-defined network inspired MAC protocol for cooperative safety in
VANETs. IEEE Trans Intell Transp Syst. 2018;19(6):2011-2024.
32. Luo G, Jia S, Liu Z, Zhu K, Zhang L. sdnMAC: a software defined networking based MAC protocol in VANETs. In: 2016 IEEE/ACM 24th
International Symposium on Quality of Service (IWQOS). Beijing, China: IEEE; 2016:1-2.
33. Huang C-M, Chiang M-S, Dao D-T, Su W-L, Xu S, Zhou H. V2V data offloading for cellular network based on the software defined network
(SDN) inside mobile edge computing (MEC) architecture. IEEE Access. 2018;6:17741-17755.
34. Huang C-M, Chiang M-S, Su W-L, Dao D-T. SDN-based V2V offloading for cellular network using the lifetime-based network state routing
(lt-nsr). In: 2018 International Conference on Information Networking (ICOIN). Chiang Mai, Thailand: IEEE; 2018:274-279.
35. Al-Badarneh J, Jararweh Y, Al-Ayyoub M, Fontes R, Al-Smadi M, Rothenberg C. Cooperative mobile edge computing system for
vanet-based software-defined content delivery. Comput Electr Eng. 2018;71:388-397.
36. Lin C-C, Chin H-H, Chen W-B. Balancing latency and cost in software-defined vehicular networks using genetic algorithm. J Netw Comput
Appl. 2018;116:35-41.
37. Rayeni MS, Hafid A. Routing in heterogeneous vehicular networks using an adapted software defined networking approach. In: 2018 Fifth
International Conference on Software Defined Systems (SDS). Barcelona, Spain: IEEE; 2018:25-31.
38. Ghafoor H, Koo I. CR-SDVN: a cognitive routing protocol for software-defined vehicular networks. IEEE Sens J. 2018;18(4):1761-1772.
39. ETSI E. 302 637-2 V1. 3.1-Intelligent Transport Systems (ITS); Vehicular Communications; Basic Set of Applications; Part 2: Specification
of Cooperative Awareness Basic Service. ETSI, Sept. 2014.
40. Fogue M, Martinez FJ, Garrido P, et al. Securing warning message dissemination in VANETs using cooperative neighbor position
verification. IEEE Trans Veh Technol. 2015;64(6):2538-2550.
41. Luo G, Yuan Q, Zhou H, et al. Cooperative vehicular content distribution in edge computing assisted 5G-VANET. China Commun.
2018;15(7):1-17.
42. Chahal M, Harit S, Mishra KK, Sangaiah AK, Zheng Z. A survey on software-defined networking in vehicular ad hoc networks: challenges,
applications and use cases. Sustain Cities Soc. 2017;35:830-840.
43. Nobre JC, de Souza AM, Rosário D, et al. Vehicular software-defined networking and fog computing: integration and design principles.
Ad Hoc Netw. 2019;82:172-181.
44. Truong NB, Lee GM, Ghamri-Doudane Y. Software defined networking-based vehicular adhoc network with fog computing. In: 2015
IFIP/IEEE International Symposium on Integrated Network Management (IM); 2015; Ottawa, ON, Canada:1202-1207.
45. Soua A, Tohme S. Multi-level SDN with vehicles as fog computing infrastructures: a new integrated architecture for 5G-VANEts. In: 2018
21st Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN). Paris, France: IEEE; 2018:1-8.
46. Salahuddin MA, Al-Fuqaha A, Guizani M. Software-defined networking for rsu clouds in support of the internet of vehicles. IEEE Internet
Things J. 2015;2(2):133-144.
47. Sahoo PK, Yunhasnawa Y. Ferrying vehicular data in cloud through software defined networking. In: 2016 IEEE 12th International
Conference on Wireless and Mobile Computing, Networking and Communications (WIMOB). New York, NY, USA: IEEE; 2016:1-8.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BHATIA ET AL. 21 of 22

48. Zhao L, Li Y, Meng C, Gong C, Tang X. A SVM based routing scheme in VANETs. In: 2016 16th International Symposium on
Communications and Information Technologies (ISCIT). Qingdao, China: IEEE; 2016:380-383.
49. Xiao Y, Ren Z, Zhang H, Chen C, Shi C. A novel task allocation for maximizing reliability considering fault-tolerant in VANET real
time systems. In: 2017 IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC).
Montreal, QC, Canada: IEEE; 2017:1-7.
50. Fontes RDR, Campolo C, Rothenberg CE, Molinaro A. From theory to experimental evaluation: resource management in software-defined
vehicular networks. IEEE Access. 2017;5:3069-3076.
51. Skiles J, Mahgoub I. A geographical hybrid solution for inter-vehicular communication in VANET. In: 2016 International Wireless
Communications and Mobile Computing Conference (IWCMC). Paphos, Cyprus: IEEE; 2016: 250-255.
52. Secinti G, Canberk B, Duong TQ, Shu L. Software defined architecture for VANET: a testbed implementation with wireless access
management. IEEE Commun Mag. 2017;55(7):135-141.
53. Hans CY, Quer G, Rao RR. Wireless SDN mobile ad hoc network: from theory to practice. In: 2017 IEEE International Conference on
Communications (ICC). Paris, France: IEEE; 2017:1-7.
54. SUMO. Urban mobility simulator. http://sumo.dlr.de/
55. Varga A. Omnet++ discrete event simulation system version 3.2 user manual. URL: http://www.omnetpp.org/doc/manual/usman.html;
2005.
56. LTE V. A simulator for heterogeneous vehicular networks. http://veins-lte.car2x.org/
57. Fontes RR, Afzal S, Brito SH, Santos MA, Rothenberg CE. Mininet-wifi: emulating software-defined wireless networks. In: 2015 11th
International Conference on Network and Service Management (CNSM). Barcelona, Spain: IEEE; 2015:384-389.
58. Network-Simulator3. A discrete-event network simulator for internet systems. https://www.nsnam.org/
59. Floodlight. Open sdn controller is an enterprise-class, apache-licensed, java-based openflow controller. http://floodlight.openflowhub.
org/
60. ONOS. Open network operating system. http://onosproject.org/
61. POX. An open source development platform for python-based software-defined networking (SDN) control applications. http://www.
noxrepo.org/pox/about-pox/
62. Chan M-C, Chen C, Huang J-X, Kuo T, Yen L-H, Tseng C-C. Opennet: a simulator for software-defined wireless local area network. In:
2014 IEEE Wireless communications and networking conference (WCNC). Istanbul, Turkey: IEEE; 2014:3332-3336.
63. Li H, Ota K, Dong M. Network virtualization optimization in software defined vehicular ad-hoc networks. In: 2016 IEEE 84th Vehicular
Technology Conference (VTC-FALL). Montreal, QC, Canada: IEEE; 2016:1-5.
64. Atwal KS, Guleria A, Bassiouni M. SDN-based mobility management and QoS support for vehicular ad-hoc networks. In: 2018 Interna-
tional Conference on Computing, Networking and Communications (ICNC). Maui, HI, USA: IEEE; 2018:659-664.
65. Nguyen TD, Huynh Q-B, Pham H-A. An adaptive beacon-based scheme for warning messages dissemination in vehicular ad-hoc networks.
In: 2017 International Conference on Advanced Computing and Applications (ACOMP). Ho Chi Minh City, Vietnam: IEEE; 2017:47-53.
66. Hasrouny H, Samhat AE, Bassil C, Laouiti A. A security solution for V2V communication within vanets VANETs.
In: 2018 Wireless Days (WD). Dubai, United Arab: IEEE; 2018:181-183.
67. Emara K. Safety-aware location privacy in VANET: evaluation and comparison. IEEE Trans Veh Technol. 2017;66(12):10718-10731.
68. Wang Y-C, Chen G-W. Efficient data gathering and estimation for metropolitan air quality monitoring by using vehicular sensor networks.
IEEE Trans Veh Technol. 2017;66(8):7234-7248.
69. Rushikesh R, Sivappagari CMR. Development of IoT based vehicular pollution monitoring system. In: 2015 International Conference on
Green Computing and Internet of Things (ICGCIOT). Noida, India: IEEE; 2015:779-783.
70. Mednis A, Elsts A, Selavo L. Embedded solution for road condition monitoring using vehicular sensor networks. In: The 6th IEEE
International Conference on Application of Information and Communication Technologies (AICT-12). Tbilisi, Georgia: IEEE; 2012:1-5.
71. Guo J, Balon N. Vehicular Ad Hoc Networks and Dedicated Short-Range Communication. Dearborn, Michigan: University of Michigan;
2006.
72. Dogru N, Subasi A. Traffic accident detection using random forest classifier. In: 2018 15th Learning and Technology Conference (L&T).
Jeddah, Saudi Arabia: IEEE; 2018:40-45.
73. Burde A, Pingat S. Next generation vehicular traffic management using vanet technology. Int J Comput Sci Inf Technol. 2013;4:715-720.
74. Haouari N, Moussaoui S, Senouci S-M, Boualouache A, Messous MA. An efficient management of the control channel bandwidth in
vanets. In: 2017 IEEE International Conference on Communications (ICC). Paris, France: IEEE; 2017:1-6.
75. Rego A, Garcia L, Sendra S, Lloret J. Software defined network-based control system for an efficient traffic management for emergency
situations in smart cities. Futur Gener Comput Syst. 2018;88:243-253.
76. Koesdwiady A, Soua R, Karray F. Improving traffic flow prediction with weather information in connected cars: A deep learning approach.
IEEE Trans Veh Technol. 2016;65(12):9508-9517.
77. Cheng J, Wu W, Cao J, Li K. Fuzzy group-based intersection control via vehicular networks for smart transportations. IEEE Trans Ind Inf .
2017;13(2):751-758.
78. Tanwar S, Vora J, Tyagi S, Kumar N, Obaidat MS. A systematic review on security issues in vehicular ad hoc network. Secur Priv.
2018;1(5):e39.
79. Akhunzada A, Ahmed E, Gani A, Khan MK, Imran M, Guizani S. Securing software defined networks: taxonomy, requirements, and open
issues. IEEE Commun Mag. 2015;53(4):36-44.
 10991131, 2019, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/dac.4005 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
22 of 22 BHATIA ET AL.

80. Hasrouny H, Samhat AE, Bassil C, Laouiti A. VANET security challenges and solutions: a survey. Veh Commun. 2017;7:7-20.
81. Deng D-J, Lien S-Y, Lin C-C, Hung S-C, Chen W-B. Latency control in software-defined mobile-edge vehicular networking. IEEE Commun
Mag. 2017;55(8):87-93.
82. Jitendra Bhatia PP, Bhavsar M. Effective resouce utilization using sdn based load balancer for minimizing request service time. Far East
J Electron Commun. 2016;2016. 3rd International Conference on Computing Sciences., At LPU- Punjab.
83. Mekki T, Jabri I, Rachedi A, ben Jemaa M. Vehicular cloud networks: challenges, architectures, and future directions. Veh Commun.
2017;9:268-280.
84. Ghaleb FA, Zainal A, Rassam MA, Mohammed F. An effective misbehavior detection model using artificial neural network for vehic-
ular ad hoc network applications. In: 2017 IEEE Conference on Application, Information and Network Security (AINS); 2017; Miri,
Malaysia:13-18.
85. Singh JR, Kumar A, Singh D, Dewang RK. A single-hop based fast certificate revocation protocol in VANET. In: International Conference
on Computational Intelligence and Networks (CINE). Bhubaneswar, India: IEEE; 2016:23-28.
86. Alrawais A, Alhothaily A, Mei B, Song T, Cheng X. An efficient revocation scheme for vehicular ad-hoc networks. Proc Comput Sci.
2018;129:312-318.
87. Jiang S, Zhu X, Wang L. An efficient anonymous batch authentication scheme based on HMAC for VANETs. IEEE Trans Intell Transp
Syst. 2016;17(8):2193-2204.

How to cite this article: Bhatia J, Modi Y, Tanwar S, Bhavsar M. Software defined vehicular networks:
A comprehensive review. Int J Commun Syst. 2019;32:e4005. https://doi.org/10.1002/dac.4005