Int. J. Ad Hoc and Ubiquitous Computing, Vol. 40, Nos.

1/2/3, 2022 159

Architecture and routing protocols for internet of

vehicles: a review

Farhana Ajaz and Mohd Naseem*

Department of Computer Sciences,
Baba Ghulam Shah Badshah University,
Rajouri, Jammu and Kashmir – 185234, India
Email: [email protected]
Email: [email protected]
*Corresponding author

Sparsh Sharma
Department of Computer Science and Engineering,
National Institute of Technology Srinagar,
Jammu and Kashmir – 190006, India
Email: [email protected]

Gaurav Dhiman
Department of Computer Science,
Government Bikram College of Commerce,
Patiala, Panjab – 147001, India
Email: [email protected]

Mohammad Shabaz
Department of Computer Science Engineering,
Lovely Professional University,
Pharwara, Jalandhar, Panjab – 144001, India
Email: [email protected]

S. Vimal
Department of Computer Science and Engineering,
Ramco Institute of Technology,
Rajapalayam, Tamil Nadu – 626117, India
Email: [email protected]

Abstract: Modern vehicles should be able to commute a tremendous amount of data and
information within their neighbourhood. To incorporate the requirements of modern vehicles,
the conventional vehicular ad hoc network (VANETs) are emerging to the internet of vehicles
(IoV). IoV keeps all the smart vehicles connected with the help of sensors, GPS, entertainment
system, brakes and throttles. These devices send and store their data with the help of cloud. This
paper intends to contribute to the review of IoV, its challenges, characteristics and application.
A detail discussion on architectures and routing protocols along with its classification is also
discussed. This paper ought to guide and motivate researchers working in the area of IoV to
develop scalable and efficient routing protocols.

Keywords: internet of vehicles; IoV; internet of things; IoT; routing protocols; architecture;
vehicular ad hoc networks; VANETs; mobile ad hoc networks; MANETs; cloud computing; fog
computing.

Reference to this paper should be made as follows: Ajaz, F., Naseem, M., Sharma, S.,
Dhiman, G., Shabaz, M. and Vimal, S. (2022) ‘Architecture and routing protocols for internet
of vehicles: a review’, Int. J. Ad Hoc and Ubiquitous Computing, Vol. 40, Nos. 1/2/3,
pp.159–175.

Biographical notes: Farhana Ajaz is working as a Research Scholar at the Department of

Computer Sciences, Baba Gulam Shah Badshah University, Rajouri, J&K, India. Her research
interests include VANET’s, internet of vehicles and internet of things.
Mohd Naseem is working as an Assistant Professor at the Department of Computer Sciences,
BGSB University. His research interests include QoS, load balancing and routing in MANETs,
sensor networks, VANETs and IoV.

Copyright © 2022 Inderscience Enterprises Ltd.

160 F. Ajaz et al.

Sparsh Sharma is working as an Assistant Professor in Department of Computer Science and

Engineering, NIT Srinagar, J&K. His area of interest is cyber security, digital forensics, ethical
hacking and blockchains.

Gaurav Dhiman is working as an Assistant Professor at the Department of Computer Science,

Government Bikram College of Commerce, Patiala. His research interest includes single-,
multi-, many-objective optimisation soft computing power systems, and change detection using
remotely sensed high-resolution satellite data.

Mohammad Shabaz is working as an Assistant Professor at Lovely Professional University,

India. His area of interest includes network analysis, OFC, social networks, machine learning
and algorithms design but not limited to it.

S. Vimal is working at the Department of Computer Science and Engineering, Ramco Institute
of Technology, Rajapalayam, Tamil Nadu, India. His areas of interest include game modelling,
artificial intelligence, cognitive radio networks, network security, machine learning and big data
analytics.

1 Introduction Table 1 Abbreviation

VANETs Vehicular ad hoc networks

In VANETs, all the autonomous vehicles are connected
MANETs Mobile ad hoc networks
using the ad hoc network in which each vehicle serves as a
RSUs Road side units
node and the ad hoc network in this particular coverage area
is called VANETs. Each participating vehicles can either act V2V Vehicle to vehicle
as a router or host. The VANETs reduced road menace to V2I Vehicle to infrastructure
a greater extent. The road safety is maintained by V2V or V2S Vehicle to sensor
V2I communication as shown in Figure 1. V2H Vehicle to humans
The VANETs is specially adapted in autonomous V2P Vehicle-to-pedestrian
vehicles where a car can have breaking controls or can DGPS Differential global positioning system
change lanes, turning, etc. A vehicle can act as an initiator GPS Global positioning system
of the packet, a receiver or a router. For this V2V or V2I IoV Internet of vehicles
type of communication, the vehicles should be equipped IoT Internet of things
with DGPS, GPS, etc. The biggest problem with VANETs QoS Quality of service
is their limited range. It has a short coverage area. Besides DSDV Destination sequenced distance vector
it is less flexible, scalable, it has connectivity problems and DSR Dynamic source routing
intelligence is not adequate. Because of VANETs temporary AODV Ad hoc on-demand distance vector
network, there were frequent topology changes and the DTP Delay tolerant protocol
vehicles were sometimes in the coverage area of some GPCR Greedy perimeter coordinator routing
ad hoc network and sometimes not a lot of valuable GPSR Greedy perimeter stateless routing
information, in the form of packets, was lost (Li et al., VADD Vehicle assisted data delivery
2019). IoV overcomes all the drawbacks of VANETs. IoV is CAR Connectivity aware routing
mainly concerned with communicating information among ACAR Adaptive connectivity aware routing
humans, roadside units and vehicles. With the increase ECAR Enhanced connectivity aware routing
in the number of vehicles need for management of these DIR Diagonal intersection routing
vehicles are used for avoiding accidents and road menace. RIVER Reliable inter vehicular routing
Table 2 shows the comparison of VANETs and IoV. ZRP Zone-based routing protocol
So, IoV came into existence. IoV is a moving network DHCP Dynamic host contribution protocol
made up of IoT enabled cars using modern electronics and GSR Geographic source routing
integrated information. IoT is a collection of related digital SDMA Space division multiple accesses
X68 machines, computing devices, mechanical devices, RREQ Route request packet
objects and other things that has some unique identification RREP Route reply packet
and are able to transfer data through a network without RERR Route error
human intervention. IGRP Intersection-based geographic routing protocol
WiMax Worldwide interoperability for microwave access
CBDRP Cluster-based directional routing protocol
 Architecture and routing protocols for internet of vehicles: a review 161

Figure 1 Generic scenario of VANETs (see online version for colours)

Table 2 Comparison between VANETs and IoV is a future interpretation of the Internet that recognises
machine-to-machine learning. Everyone can be connected
VANETs IoV to everything. The basic idea of IoT is that it provides
VANETs has simple network It has complex network connectivity for anyone at any time and place to anything
Communication is Communication is V2V, V2I, at any time and place. IoT allows autonomous and
vehicle-to-vehicle, V2S, and V2H. secure connection and exchange of information between
vehicle-to-infrastructure the existing real-world devices. There is an enormous
The network in VANETs is It has a large network increase in the number of devices that are connected to
limited the Internet these devices may include personal computers,
The main actors that we have The main actors we have in smartphones, PDAs, etc. (Sheu et al., 2019). Figure 2
in VANETs are vehicles and IoV are humans, vehicles and explains the generic scenario of IoT.
RSUs RSUs. IoV keeps all the smart vehicles connected where
It has flat platform It has hierarchical platform each vehicle serves as a node. Electronics used in IoV
The network here is The network is heterogeneous include sensors, GPS, entertainment systems, brakes and
homogeneous
throttles. IoV is an advancement of V2V (vehicle to
vehicle communication). IoV is the communication between
vehicles over a network. The vehicle could be a sensor
Table 3 Similarities between IoT and IoV
platform, collecting information from the driver, utilising it
IoT IoV for secure navigation, pollution control and management of
traffic (Poongodi et al., 2019). Figure 3 explains the generic
For processing and storing of For processing and storing of
big data we use cloud big data we use cloud.
view of IoV.
There are evident five features of smart vehicles mobile
It has complex network It has complex network
application, electric vehicle, self-driving, safety driving,
It has a large network It has a large network
and social driving (Maglaras et al., 2016). Nowadays,
The network is heterogeneous The network is heterogeneous
transportation has stretched beyond levels as the number of
individuals using those keeps on rising. In the recent report,
Table 4 Comparison between IoT and IoV it was seen that the quantity of vehicles utilised overall is
somewhat is higher than 1 billion and is relied upon to
IoT IoV arrive at 2 billion by 2035 (Kaiwartya et al., 2016). Every
IoT is a super set of IoV IoV is a sub set of IoT year, approximately 1.2 million people die because of road
IoT is the information IoV is an information accidents further 7 million individuals are harmed in around
interaction between different interaction among vehicles, 8 million accidents. This rapid increase in the number of
smart things. roadside units and humans. vehicles causes traffic blockage and fatalities that happen
Automation of things. Automation of vehicles. because of accidents on the road. The biggest challenge
It connects heterogeneous It connects heterogeneous in IoV is that of big data to be stored and processed.
object sensors vehicles sensors. Cloud computing and mobile cloud play a significant part
in handling big data (Dhaliwal et al., 2018; Singh et al.,
The communication we have seen till now is human-human 2016). Similarities and comparisons between IoT and IoV
or human-device but IoT promises communication between are shown in Tables 3 and 4, respectively.
machine-machine. Internet is evolving continuously. IoT
162 F. Ajaz et al.

Figure 2 Generic scenario of IoT

Figure 3 Generic scenario of IoV (see online version for colours)

Rest of the paper is organised as follows. Section 2 2 Challenges

describes Challenges of modern communication.
Characteristics of IoV are explained in Section 3. In • An economical technique like IP addressing can be
Section 4, we have explained the application of IoV. used by researchers to replace digital maps.
Detailed discussion along with comparison of various Researchers can use cheaper techniques to replace
architectures of IoV are analysed in Section 5. Section 6 digital maps such as IP Addressing. If we give every
describes different types of routing protocols for IoV. vehicle an IP address, that will be far cheaper than
Finally, Section 7 concludes the paper and provide the making use of digital maps. To incorporate IP
future direction in this research area. addressing in the vehicles we can make use of DHCP
which is a protocol of the application layer.
 Architecture and routing protocols for internet of vehicles: a review 163

• Design the protocols that are scalable. The number of the triggered event. The message can be sent from a
vehicles is increasing in an extensive way so, the single source to a single destination (unicasting) or
protocols that are scalable should be proposed, i.e., sending messages from a single source to a particular
they should perform well under increased load. The group by using multi-hop communication
protocols should perform equally well under heavy (multicasting) or vehicles can also send messages to
loads. Their performance should not get hampered. all other vehicles by broadcasting.
• Rural and highway areas should be the centre of • Handling big data: Huge amount of data is generated
focus for researchers in the future. Researchers should because large numbers of vehicles are connected, So
specifically concentrate their future researches on the IoV uses cloud computing for processing and storing
rural and highway areas. Rural areas should not be the big data.
isolated. What this means is that these areas require a
• Internet facility: IoV has a unique characteristic of
great deal of improvement as everything so far has
having access to the Internet. The vehicles that are
been more inclined towards the urban areas. This
connected can take benefit from this huge network
strictly calls for improvement for the rural areas, so
(Senouci et al., 2019).
that we make them as good as urban areas.
• Protocols should consider the density of the network
that is badly balanced. Researchers should take into 4 Application
consideration the density of vehicles as its poorly
balanced. The areas that have higher vehicular density The vehicle could be a sensor platform, collecting
tend to have lower delays because the nodes are information from the driver, utilising it for secure
closer and the packet transfer is easier. However, this navigation, pollution control and management of traffic that
is not the case if the vehicular density is lower, i.e., are given below:
these have higher delays and the packet transfer is not • drivers can notice the traffic plans and fix their
guaranteed. So, we should design such protocols that driving routes accordingly
perform equally well with the areas of lower
vehicular densities also. • minimal fuel consumption
• Efficient methods should be used to irradiate and • avoiding accidents
diminish the network fragmentation problem. Network • mitigating the problem of traffic jams
fragmentation can be reduced by sending periodic
re-transmission. Network fragmentation is reduced to • fleet management.
a great extent in geocast routing protocols. • pollution control.
• Network overload should be reduced. One way to do
Figure 4 shows some of the applications of IoV.
that is to avoid redundant packets and unnecessary
packets which cause a great load to the network. We • Drivers can know the traffic plans of road and can
can also overcome this problem by making use of the adjust their driving routes accordingly: As the vehicle
cloud computing technique because in cloud has the capability to absorb the important information
computing we tend to get data on demand. Hence, due to the embedded sensors. Drivers can know the
there will be fewer loads on the network. traffic plans of the road and can adjust their driving
routes accordingly. Traffic plans will help the drivers
to choose a route with less traffic. This will help the
3 Characteristics drivers to reach from source to destination in less
time.
Numerous characteristics differentiate between ad hoc
• Fuel consumption: The vehicles consume fuels such
networks and IoV:
as petrol, diesels, etc. IoV helps in choosing the
• Information collection: The data is gathered by optimal paths using the shortest path algorithms to the
employing various sensors (cameras, road sensors, destination. So, the vehicles can consume less fuel
rain sensors, etc.) that let drivers have adequate and can be more economical.
information and react to the environment changes in • Avoiding accidents: Accidents are increasing
an adequate and efficient way (Senouci et al., 2019). worldwide and millions of people die every year due
• Highly mobile: Vehicular networks have highly to accidents. So, another big advantage of these
mobile nodes, related to the speed of cars, so that vehicles is that they can help in reducing the number
node can enter and leave in a short time, thus lead to of accidents. As smart vehicles are embedded with
frequent topology changes. sensors, they can sense the movement of traffic or
people around them. This will give the vehicle proper
• Type of information transported: The messages are time to put on brakes and thus it will lead to a lesser
transported according to the degree of involvement in number of accidents.
164 F. Ajaz et al.

Figure 4 Applications of IoV

• Mitigating the problem of traffic jams: Traffic jam is 5.1 The three-layer architecture
a big problem nowadays when we travel from one
place to another. These smart vehicles will have the On the basis of communication among different
ability to choose the route that has less traffic with technologies in the IoV environment, researchers have
the help of GPS and their sensors. Thus avoids traffic proposed three-level architecture (Liu, 2011).
jams which will be very beneficial as it will save the The IoV consists of three layers as shown in Figure 6.
total time that is required for the journey.
• Perception layer: The first layer contains all the
• Fleet management: Another advantage of these smart sensors within vehicles that gather information about
vehicles is that these vehicles will also help us in the environment and detect various events like vehicle
fleet management by choosing a route that has less situations, driving patterns and environmental
fleet movement. These will help the pedestrians to pollution.
move with ease from one place to another.
• Network layer: It is the interactive layer that supports
• Pollution control: Pollution is the biggest problem distinct wireless communication modes like V2I,
when it comes to using vehicles for a journey. As we V2S, V2V, V2P. It ensures connectivity with other
know that these vehicles are embedded with sensors communication networks, e.g., GSM, WiMax,
will have the ability to choose the shortest path to the WLAN, Bluetooth.
journey, which will mean that the vehicle will burn
less fuel and that will lead to less air pollution. • Application layer: This layer encloses the statistical
Another big advantage to these smart vehicles is that tools, processing information and storage support. It
they will use the latest emission standards in the also helps in decision making such as risk situation,
engine of the vehicle which means that the engine bad weather, traffic congestion, etc. It is accountable
will produce less CO2 . When we combine the shortest for processing, analysis, storage, decision making
journey with the latest emission standards our vehicle about different risk situations. The architecture of IoV
will help in reducing air pollution significantly. is the same as that of VANETs with few
developments and improvements. Kaiwartya et al.
(2016) gave layered architecture that has five layers
5 Architecture of IoV that include business, application, artificial
intelligence (AI), coordination, and perception.
In Section 5, the IoV connects billions of vehicles by using
the heterogeneous network, so a basic need for flexible
architecture aroused. Different architectures are proposed
by different researchers and these are presented in the
following section: A brief overview is shown in Figure 5.
 Architecture and routing protocols for internet of vehicles: a review 165

Figure 5 Architectures of the IoV in literature

Figure 6 Three-layer architecture Figure 7 Four-layer architecture

5.2 Four-layer architecture

5.3 Five-layer architecture
Four-layer IoV architecture has been proposed by CISCO
(Flavio, 2013). The endpoint layer encloses the vehicles, It is proposed by Kaiwartya et al. (2016). The five layer
software and V2V communication that takes place through architecture is shown in Figure 8. Each layer has some
802.11p. The infrastructure layer includes technologies specific functionality which are explained as follows:
that allow connection among all the actors of IoV. The
operational layer monitors flow-based management and • business layer: this layer is used for application
policy enforcement. Lastly, the service layer defines the pricing, preparation of budget, and aggregation of data
services which cloud offers to drivers based on on-demand,
• application layer: this layer provides a friendly user
subscription shown in Figure 7.
interface
After analysing the above architectures several
weaknesses were identified: • artificial intelligent layer: this layer is responsible for
processing, storing and data analysis
1 the above proposed architectures do not take security
into consideration • coordination layer: this layer is responsible for the
security of information
2 communication among passengers and drivers is
hampered • perception layer: this layer contains all the sensors
and gather information from outside.
3 all the compiled information is disseminated without
processing that often leads to congestion. Table 5 compares the five layers with respect to its
representation and its functionalities.
166 F. Ajaz et al.

Table 5 Five-layer architecture

S. no. Layers Representation Functionalities

1 Business Graphics, tables, diagrams, flowcharts Usage of resource and application pricing, preparing
budget and aggregation of data.
2 Application Smart vehicle applications Intelligent service to end users.
3 Artificial intelligence Expert systems, analysis of big data, cloud Processing, storing and data analysis.
computing
4 Coordination LTE, Wave, Wi-Fi Secure information transportation, unified structure
transformation, interoperability provisions.
5 Perception Personal devices, sensors and actuators. Information gathering, energy optimisation at lower
level (Sharma and Kumar, 2019, 2017).

Figure 8 Five-layer architecture • Management and control layer: It applies various

functions and policies such as traffic engineering,
packet inspection, traffic management to better
manage the information. It also manages different
service providers.
• Processing layer: It processes the huge data using
different cloud computing techniques then the
processed data is used by various service providers
for further service enhancement.
• Security layer: It is accountable for all the functions
related to security (like integrity, confidentiality,
non-repudiation, data authentication access control,
and availability) among others (Wang et al., 2019).

Figure 9 Seven-layer architecture of IoV

5.4 Seven-layer architecture

Contreras-Castillo et al. (2017) proposed an architecture

that has seven layers for IoV which allows transparent
interconnection of the network segments, the spread of
information into IoV environment. A security layer was
provided for managing authentication, authorisation among
IoV entities. The seven layers architecture is shown in
Figure 9.

• User vehicle interface layer: It provides interface to

coordinate with notifications for all drivers and picks
the best element of display to help reduce drivers
distractions (Contreras-Castillo et al., 2017).
• Data acquisition layer: This layer is responsible for
gathering all the information from various sources.
• Pre-processing and data filtering layer: It filters the
compiled information to prevent the spread of
un-important information.
• Communication layer: It selects the optimum network
to transmit information, which is used for several Various architectures for IoV available in literature
selection parameters like QoS, congestion levels in proposed by various researchers are compared in Table 6.
various networks available.
 Architecture and routing protocols for internet of vehicles: a review 167

Table 6 Comparison between different layers architecture

S. no. Three-layer architecture Four-layer architecture Five-layer architecture Seven-layer architecture

1 It is proposed by Liu (2011) It is proposed by CISCO It is proposed by Kaiwartya This architecture is proposed by
(Flavio, 2013) et al. (2016) Contreras-Castillo et al. (2017)
2 It consists of three layers: It consists of four layers: It consists of five layers: It consists of six layers:
1 Perceptron layer 1 Endpoint layer 1 Business layer 1 User vehicle interface layer
2 Network layer 2 Infrastructural layer 2 Application layer 2 Data acquisition layer
3 Application layer 3 Operational layer 3 Artificial intelligent 3 Pre-processing and data
4 Service layer 4 Coordination layer filtering layer
5 Perception layer 4 Communication layer
5 Management and control layer
6 Processing layer
7 Security layer
3 Data is not secured. This architecture does not Communication of data is It consists of the security layer
take security into secured to some extent by which is accountable for all the
consideration using a coordination layer. functions related to security (like
integrity, confidentiality,
non-repudiation, data
authentication and access control,
availability) among others.
4 Network congestion because No concept of No pre-processing layer Network congestion is reduced
of the lack of pre-processing pre-processing. used. because of the implementation of
techniques. the processing layer in IoV
architecture.

6 Routing protocols • Hybrid routing protocol: The hybrid routing

amalgamate the features of the proactive and reactive
Routing protocols define how the routers communicate with protocols. In fact, the advantages of proactive routing
one another to disseminate information that can be used to are merged with the advantages of reactive routing.
select the best route between the two nodes. The routing The hybrid routing protocol is used to find out the
protocol first shares the routing information among the optimal network destination routes.
immediate neighbours then throughout the network. The
purpose of configuring routing protocols on routers is to
share and exchange routing information. Various routing 6.1 Information-based routing protocol
protocols are shown in Figure 10. Routing protocols are
It distributes the routing information by either sending the
divided into three broad categories based on their update
smaller incremental updates or sending the entire dumps
mechanism: proactive routing protocol, reactive routing
infrequently. Examples of routing protocols are proposed by
protocol and hybrid routing protocol (Ajaz et al., 2021).
Hager et al. (2015), Yokoyama et al. (2014) and Carpenter
• Proactive routing protocol: Proactive protocols are (2014).
also called table driven routing protocol. In this
approach, routing information of all the nodes is 6.2 Topology-based routing protocol
maintained, routing updates are distributed among all
the nodes periodically for adding new routes or Destination-based distance vector (DSDV): it makes use of
updating routes. Routes to the destination are already the Bellman-Ford algorithm. For hop count a cost metric
established and can be used at any time. The is used, that is the number of hops it takes to reach
disadvantage of this routing protocol is that the the destination. DSDV is a table-driven routing scheme.
routing table can grow in size (hence the network In the routing table, each entry has a sequence number;
density) than the number of routes that are actually the sequence numbers are even if the link is present
needed (Naseem and Kumar, 2015). otherwise odd. It uses sequence number in routing tables for
• Reactive routing protocol: Reactive routing protocol eradicating loops. A node will update its table if it receives
is also called an on-demand routing protocol. In this its updated route to the destination (i.e., if a node receives
protocol, routes are built only whenever needed. Thus a route with a higher sequence number) the information
these routing protocols do not maintain routes for of routing is disseminated within nodes by transmitting
every node and there is no need for frequent updates smaller incremental updates more frequently and sending
of routing tables. One of the disadvantages of the full dumps infrequently (Naseem and Kumar, 2013).
reactive routing protocol is the delay in finding the
routes (Naseem and Kumar, 2017).
168 F. Ajaz et al.

Figure 10 Routing protocols of IoV

In DSDV, each node has a routing table that contains is broadcasted to make them update their cached route.
a list of all the destinations available, next-hop to each RERR packet is sent to the transmitting node and the
destination, distance to destination, sequence number which process repeats. The comparison between AODV and DSR
is produced by making use of target node as shown in is in the way the route information is stored. In AODV, the
Figure 11. By using these routing tables, the packets are route information is kept at the intermediate nodes while in
transferred between the nodes. DSR it is stored at the sender node and in the header of
the transferred packets. DSR protocol is specially designed
Figure 11 Routing table entry of DSDV (see online version for multi-hop wireless ad hoc networks. DSR network is
for colours) self-organised and self-configured, without the need for
any existing network infrastructure. DSR has two phases:
route discovery and route maintenance. Both of these are
performed on-demand (Johnson and Maltz, 1996).
DSDV routing table entry includes: destination, next hop, • Route discovery: Is the process in which the node
distance, sequence number. which wants to send a packet to the destination node
obtains a source route to the destination. The route is
Problems: not predefined. Figure 12 illustrates route discovery.
• Routes with newer sequence numbers are always • Route maintenance: Route maintenance is used when
preferred; routes that have older sequence numbers the source node is transmitting packets to the target.
are discarded. It checks whether the route to the destination is
broken or network topology has changed because the
• If a route has a sequence number equal to the existing
link no longer works at that time source uses another
route then it is preferred. If it has some better metric,
route to the destination.
then the existing route is dropped.
In Figure 12, a node X attempts to find a route to V.
DSR: it is a reactive routing protocol. The first phase is
X broadcasts a RREQ packet that is received by all the
route discovery, node discovers the route by creating a
nodes in the transmission range of X. Each RREQ packet
RREQ packet that is flooded in the network. As RREQ
identifies the initiator and contains a unique request-id.
message propagates through the network packet header of
Each RREQ also has the address of the intermediate nodes
RREQ which consists of the list of hops is incrementally
through which the RREQ has been forwarded. If any other
added into the packet header of RREQ. Once the RREQ
node receives the Route request and if it is the destination
reaches the destination RREP is sent back along the
node then a RREP packet is sent to the initiator of the route
reverse path. The source node will cache the route for
discovery along with the copy of the route record. If the
the destination node. In the future, if the source wants
node is not the destination node it appends its own address
to send a packet to the same destination it will get a
in the RREQ packet and transmits it.
route from the cache. If any link is broken information
 Architecture and routing protocols for internet of vehicles: a review 169

Figure 12 Route discovery (see online version for colours)

AODV: it is a routing protocol that is reactive where also called geo routing. A position-based routing protocol
on-demand routes are formed whenever needed (Perkins is one of the most reliable routing protocol for vehicular
and Royer, 1999). The AODV routing protocol is mostly networks it is because of the actuality that nodes can get the
used in ad hoc networks. Route discovery in AODV is information about the position from on-board GPS receivers
the same as DSR. The network remains silent until a and obtains from the onboard digital map the information
connection is required. Whenever a node wants to identify related to the road layout. Position-based routing eliminates
a path to some target node, it broadcasts a route request some of the limitations of topology-based protocols such
packet (RREQ) other node receives it, forwards it, and as maintaining routes. One of the most familiar techniques
records the node from where they have received it. When that exploit location information for routing packets is the
the destination node receives the RREQ it replies by greedy approach (Liu et al., 2016). Using this approach
transmitting the RREP packet through the temporary route the neighbouring node which is closest to the target node
to the node that is requesting it. The source then begins is eligible to advance the data packet into the network.
using the route. A route error packet (RERR) is transmitted The geographic protocols are divided into three categories
to inform all nodes about the failure of the link. One of that are progress-based, randomised-based, and face-based
the disadvantages of DSR is that the routing information (Bujari et al., 2018a, 2018b).
is stored in the header of the data packet which is to be
transmitted. If the network size grows the length of the • Progress-based: In this routing protocol, the node that
route path also increases and thus the data packets header currently holds the packet forwards it to a node that
increases, so most of the bandwidth will be consumed by is making maximum progress towards the target node.
the route path. It is hence not the efficient utilisation of • Randomised-based approach: The node that is
network bandwidth. The routing table of DSDV consists of currently holding the data packet forwards it by
destination, next hop, number of hops, destination sequence choosing the next node randomly.
number, and active neighbours of the route and expiry time
of the routing table entry. Routing request: the route request • Face-based: Face-based approach uses an algorithm
message is used when there is no route available; RREQ is that forwards the packet within faces and always
flooded throughout the network. It has the following fields guarantees packet delivery in 2D network.
shown in Figure 13:
In 3D space, geographic routing is very difficult than
• Route error: When any link is broken or the node is in 2D topology. The greedy forwarding approach in 3D
no longer available, RERR is generated to inform all reaches more local minima than in 2D. Moreover, many
the other nodes about the loss of the link. 2D position-based routing protocols cannot be extended to
3D. Huang et al. (2016) stress the inability of the routing
• Advantage of AODV: AODV is reactive it does not protocols to assure delivery in 3D ad hoc networks.
require a routing table. So, it does not require GPCR: this is proposed by Lochert et al. (2005). The
additional traffic to maintain routing tables. main goal of this protocol is to improve the performance of
• Disadvantage of AODV: It takes more time for GPSR. GPCR has two sections restricted greedy forwarding
establishing routes as compared to DSDV. (RGF) procedure and repairing strategy. It completely
depends on the junctions and the real world street topology.
Figure 13 Route request packet (see online version for colours) All the packets are forwarded to a junction node to find the
next hop. This protocol can not completely entirely resolve
the local maximum problem in a way that when forwarding
the packet it may arrive at a situation where its distance
to the destination to target is closer than the neighbour’s
distance to the destination (Lochert et al., 2005).
6.3 Position-based routing protocol GPSR: it is given by Karp and Kung (2000), it makes
use of a position of neighbouring nodes, the position of
Another approach of routing is position based that uses the destination packet and its current position for making
the location of the node to determine the next hop. It is decisions about forwarding the packets. It is an efficient
170 F. Ajaz et al.

protocol for mobile wireless networks. It makes the packet network. However, one of the big problems that occur in
forwarding decisions by exploiting the position of the heterogeneous networks including IoV is hand-off. Table 7
nodes. It makes use of the greedy forwarding nodes that compare some of the routing protocols.
make the most progress towards the destination.
IGRP: this protocol was proposed by Kihl et al. (2008)
in order to resolve the QoS problem in VANETs in 6.6 Map-based routing
metropolitan areas. A genetic algorithm is used to identify
GSR: it uses a forwarding path to forward packets,
the best solution in order to satisfy QoS constraints on
based on the location of the vehicle and its arrangement
performance metrics of building backbone routes based on
on the map the forwarding path is calculated, but it
adjacent and intermediate road intersections towards the
fails when vehicular density is low on roads (Zhongyi
Internet gateway.
et al., 2009; Xiang et al., 2013), has proposed a stateless
Bana and Varaiya (2001) proposed a robust MANETs
geographic VANETs routing (Geo-SVR) which merges the
architecture called SDMA. It gives access to the users to
position of a node with the digital map, it also uses an
the communication channels on the basis of their location.
improvised forwarding algorithm for addressing unstable
Users get a communication medium that is free from the
wireless channel problems. Experiments and simulations
collision. Enhanced SDMA is used in order to improve the
have evaluated its performance. Geo-SVR has a higher
bandwidth efficiency of SDMA.
packet delivery ratio (PDR) with latency equal to that of
GPSRs and AODV.
6.4 Delay-based routing protocols

This type of routing protocols are classified into two types: 6.7 Path-based routing
delay sensitive protocol and delay tolerant protocols.
Zhao and Cao (2008) presented the (VADD) vehicle
Delay sensitive protocol: they cannot tolerate delay and
assisted data delivery protocol. VADD uses a carry forward
requires transmission of road’s information, rapidly. Delay
mechanism to route data from sender to destination. The
sensitive protocols need an exchange of the packets as
problem of frequent disconnections and high mobility are
soon as possible. DSP should reduce the probability of
dealt with using this protocol. VADD is not that efficient
congestion of routing packets and enable a high delivery
when dealing with sparsely connected networks. VADD
ratio with low latency.
minimises delay by choosing the highly-dense path where
Delay tolerant protocol: DTP use a carry forward
the network density is high. VADD calculates the best path
mechanism. They can manage the failures during carry
with the lowest data delivery delay.
forward mechanism (Li and Boukhatem, 2013; Khanna
et al., 2013; Lee et al., 2008). It deals with cases of
occasional connectivity problems. Their representative is 6.8 Transmission-based routing strategy:
GeoSpray (Soares et al., 2014). GeoSpray improves the
probability of delivering packets and also reduces the delay Routing protocols are also based on transmission strategies
in delivering packets. One of the limitations of DTN routing and there are three modes of transferring information from
protocol is that it limits the network transmission while the source to the destination. Unicasting is the sending of
achieving a high degree of network coverage. the packets from a single source to a single destination,
while multicasting or geo casting is sending of the packets
from a single source to a particular group of nodes that
6.5 Target network routing protocols are located at some particular position of interest. The
transmission-based routing protocol is shown in Figure 14:
Homogeneous network: it refers to the access type of
technology that a vehicle could use. The classical routing
protocols in VANETs concentrate on transmitting the packet 6.8.1 Unicast routing
over short distances using wireless technologies (Ho et al.,
2010; Dorge et al., 2012; Shafiee et al., 2011) like ‘wireless It allows the transfer of data from one sender to only one
access in vehicular environment’ WAVE standard (Ho et al., receiver. It is performed using two methods named as carry
2010). forward technique and multi-hop transmission. In wireless,
Heterogeneous network: heterogeneous characteristic multi-hop central nodes that are in the path of the routing
of IoV includes diverse radio access technologies. IoV is are used to route data from sender to receiver (Naseem and
complicated than VANETs because IoV includes different Kumar, 2014). The other technique is carry forward. In this
radio access technologies, hence it leads to a heterogeneous technique, the sender node carries data as long as possible.
 Architecture and routing protocols for internet of vehicles: a review 171

Table 7 Comparison between different types of routing protocols

Information-based TBR (topology-based PBR (position-based DBR (delay-based TNR (target network
routing protocol routing protocol) routing protocol) routing protocol) routing protocols)
Routing information is These routing protocols It makes use of the Delay-based routing It refers to a type of
distributed by either are based on the position of the node to protocols are access technology that
sending the smaller topology, i.e., whether determine the next hop. delay-sensitive routing can be used by the
incremental updates or the routing protocol is It is also called geo protocols (cannot tolerate vehicle: homogeneous
sending the entire dumps reactive (on-demand) like routing. delay) and delay-tolerant network and
infrequently. DSR, AODV, proactive routing protocols (can heterogeneous network.
(table-driven) like DSDV tolerate delay)
or hybrid.

Figure 14 Transmission-based routing protocols

Next, we have provided a description of some already source and destination. In case of a primary path failure,
existing protocols. the backup path is used by the source. Whenever the source
Naumov and Gross (2007) proposed connectivity aware node gets a reply packet, it transmits the data by using the
protocol (CAR). In this protocol, first, the searching packet primary path and in case some complication arises at the
is sent, each forwarding packet records the ID and number time of this data transfer, then the source node would resend
of hops in the searching packet. After the packet reaches the the data packets using the backup path. The advantage of
destination using some chosen routing path, the target sends ECAR is it is an efficient route maintenance mechanism.
a reply to the initiator and then the reply traverses through Chen et al. (2011) proposed diagonal intersection-based
the set of intersection points called anchor points. The routing protocol (DIR). DIR is used to improve CAR. The
source then starts sending data from the chosen intersection process of the DIR protocol is the same as that of the CAR
points through the highly-dense path using the ‘greedy protocol with some modifications. Firstly the source node
forwarding technique’. search packets are broadcasted to locate the destination
Yang et al. (2010) proposed adaptive connectivity aware node. RREP shows series of diagonal intersections. It builds
routing protocol (ACAR). In this protocol, the source series of diagonal intersections between the initiator and
locates the destination by using the GPS system. It target vehicles. The source node that wants to send the
uses statistical data for selecting an optimal path to the data will have to compulsorily follow a series of diagonal
destination. After the request is received from the source intersections. At each intersection there are two sub-paths,
node, the destination node will make a comparison between the optimal sub-path will be chosen for the delivery of the
the real and statistical density collected. If their difference is data packet.
huge, then the destination node will notify the source node Bernsen and Manivannan (2012) presented reliable
to choose another path. inter-vehicular routing protocol for VANETs. The data
Maidorawa (2013) proposed enhanced connectivity packets are forwarded using the greedy forwarding strategy.
aware routing protocol (ECAR). It is an extension of CAR The reliable routes can be selected by using traffic
protocol. Control broadcast is used over the request path monitoring components, especially in rural areas.
and reply path in order to build a backup-path between
172 F. Ajaz et al.

Figure 15 Multicast and geocast routing (see online version for colours)

Figure 16 Broadcast routing (see online version for colours)

Table 8 Comparison between type of transmission-based routing protocols

S. no. Unicast routing protocol Multicast routing protocol Broadcast routing protocol
1 It allows transfer of data from single It transmits the data from single source This protocol allows source to
source to only one receiver. It is to a specific group. It can be transmit packet to all the nodes or
performed using two methods. ‘Carry categorised into two categories: other vehicles that are in the network.
forward technique’ and ‘multi-hop geocast-based routing protocol and
transmission’. cluster-based routing protocol.
2 In carry forward technique, the data is Geocast-based routing protocols: we Broadcast routing protocols are
stored and then forwarded from node to can say geo cast is location-based frequently used for sharing traffic,
node when both the nodes meet until multicast routing. The source node can weather, and road condition
all the packets are sent to receiver node deliver packets to a group of nodes that information among all vehicles
from the initiator node. Whereas, are placed in some specified geographic
multi-hop transmission makes use of region. Whereas, in cluster-based
multiple wireless hops to transfer routing protocols, network is divided
information from the initiator to the into sets known as clusters. Every
receiver cluster has one cluster head for
managing both the intra and inter
cluster communications.

6.8.2 Multi-cast routing protocol relevance can receive the packet while the nodes that are
present in the zone of forwarding region must forward a
It transmits the data from a single source to a geo-cast message.
specific group. It can be categorised into two categories: Bachir and Benslimane (2003) proposed a protocol
geocast-based routing protocol and cluster-based routing called as a multi-cast protocol in ad hoc networks
protocol. Figure 15 shows multi-casting of the packets. inter-vehicle geocast, It is used to inform vehicles about the
risk situation such as accidents etc. The multi-cast group
6.8.3 Geocast routing protocol is defined and the IVG notifies about the location of the
risk area to all the vehicles. Defected vehicle advances the
We can say geocast is location-based multicast routing. The packets to the multicast group. All the vehicles that are in
source node can deliver packets to a group of nodes that are the risk belt transfer packets to the other vehicles in the risk
placed in some specified geographic region called as zone belt in order to notify every vehicle about the risk factor.
of relevance.
Joshi et al. (2007) presented a protocol called distributed
robust geocast (DRG). This protocol avoids network 6.8.4 Cluster-based routing
fragmentation and the network load is reduced by sending In this technique, the network is divided into sets known as
periodic re-transmissions. The distributed approach of this clusters. Every cluster has one cluster head for managing
protocol ensures the delivery of packets. The nodes that both the intra and inter-cluster communications.
are present in the specified geographical region or Zone of
 Architecture and routing protocols for internet of vehicles: a review 173

Song et al. (2010) proposed CBDRP. To set up the Step 3 If the destination node is not available within the
communication link, the source sends the data to its cluster zone of the source node, then the source sends a
head and then data is forwarded to the cluster head of RREQ packet to its peripheral nodes.
another cluster. This process continues until the data is
Step 4 If a peripheral node receives the RREQ packet,
transmitted to the cluster head of the destination cluster.
it checks whether the destination is available
The data is then forwarded to the destination.
within its zone or not. If the peripheral node has
Girinath and Selvan (2010) presented a protocol
the destination in its zone, then it sends the
for cluster-based algorithm for hybrid mobility model.
RREP packet to the source node and if not then
This protocol makes use of both dynamic clustering
it sends the RREQ to its peripheral nodes and
techniques and static clustering techniques. The layering
the process continues.
environment is produced by this clustering technique. The
first layer constitutes of static cluster head, the second layer Step 5 If multiple route replies are received, the optimal
constitutes a dynamic cluster head, last layer or the third one is chosen.
layer constitutes of common node. The static cluster head
collects the important information for the creation of the Various nature inspired algorithms are also brought in use
cluster. The role of the dynamic cluster head of the second by researchers for solving different optimisation problems
layer is accountable for maintenance, i.e., systematisation related to routing such as finding the optimal low cost route.
or coordination of nodes that are within the cluster and Example of such algorithms are proposed by Dehghani
between the cluster. This protocol shows good performance et al. (2020a, 2020b), Dhiman et al. (2020), Dhiman et al.
for data transmission at the cluster level. (2021), Dhiman and Kaur (2019) and Dhiman (2019).

6.8.5 Broadcast routing protocol 7 Conclusions

This protocol allows the source to transmit the packet to In recent time, IoV is emerged as a disruptive
all the nodes or other vehicles that are in the network technology which can be widely used for communication
as shown in Figure 16. Broadcast routing protocols are between vehicles over a network. There are number of
frequently used for sharing traffic, weather, and road communication problems that can be solved using this
condition information among all vehicles. technology such as minimal fuel consumption, avoiding
Durresi et al. (2005) proposed a protocol. This accidents, pollution control, mitigating the problem of
protocol is used as an emergency broadcast protocol called traffic jams etc. This paper provides a extensive details
BROADCOMM. The sensors that are present in modern of IoV, its applications, characteristic and challenges. A
cars gather all the information and in case of emergency, detail analysis of various architectures of IoV along with
they trigger an immediate broadcast. This protocol gives an comparison has also been explained. Along with these,
improved QoS. this paper also provides a complete taxonomy of various
Tonguz et al. (2006) presented a protocol known as existing routing schemes with their relative advantages and
a distributed vehicular broadcast protocol. It is used for disadvantages. In the future, we would like to propose
multi-hop broadcast in VANETs. It can be used both for a new routing protocols based upon the discussion and
sparse network densities as well as for dense network analysis provided in this paper.
densities on highways. It also reduces the broadcasting
overhead.
Table 8 compare different type of transmission-based 7.1 Future work
routing protocols:
While designing protocols of routing for IoV many
challenges are encountered. Some of the challenges are
6.9 Hybrid routing protocol the fragmentation of the network, high mobility and quick
topology changes, use of digital maps that are very
Zone routing protocol (ZRP): it uses both the intra-zone expansive. An economical technique like IP addressing
routing protocol and inter-zone routing protocol. Intra zone can be used by researchers to replace digital maps. The
routing protocol is used for maintaining routes in the zone protocols that are scalable should be proposed time ahead.
using the proactive protocol. Inter-zone routing protocol Rural and highway areas should be the center of focus for
finds the routes to nodes outside the zone using reactive researchers in the future. Rural areas should not be isolated
routing protocols. Working of ZRP is explained as follows: that is these areas require a great deal of improvement as
everything so far has been more inclined towards the urban
Step 1 Source node will check whether the destination areas. This strictly calls for improvement for the rural areas,
is available within its zone or not. so that we make them as good as urban areas, another major
Step 2 If the destination is available within the zone of threat in IoV is security, and therefore it should be strictly
the source node, then packets can be directly imposed for safe navigation of data. Above all the serious
sent to the destination node. issues are addressed and these should be dealt with in near
future.
174 F. Ajaz et al.

References Dhiman, G., Oliva, D., Kaur, A., Singh, K.K., Vimal, S., Sharma, A.
and Cengiz, K. (2021) ‘BEPO: a novel binary emperor penguin
Ajaz, F., Naseem, M., Ahamad, G., Khan, Q.R., Sharma, S. and optimizer for automatic feature selection’, Knowledge-Based
Abbasi, E. (2021) ‘Routing protocols for internet of vehicles: Systems, Vol. 211, p.106560.
a review’, in Malik, H., Fatema, N. and Alzubi, J.A. (Eds.): Dorge, P.D., Dorle, S.S., Chakole, M.B. and Thote, D.K.
AI and Machine Learning Paradigms for Health Monitoring (2012) ‘Improvement of QoS in VANET with different
System. Studies in Big Data, Vol. 86, pp.95–103, Springer, mobility patterns’, 2012 International Conference on Radar,
Singapore. Communication and Computing (ICRCC), pp.206–209, IEEE.
Bachir, A. and Benslimane, A. (2003) ‘A multicast protocol in ad Durresi, M., Durresi, A. and Barolli, L. (2005) ‘Emergency broadcast
hoc networks inter-vehicle geocast’, The 57th IEEE Semiannual protocol for inter-vehicle communications’, 11th International
Vehicular Technology Conference, 2003. VTC 2003-Spring, Conference on Parallel and Distributed Systems (ICPADS’05),
Vol. 4, pp.2456–2460, IEEE. Vol. 2, pp.402–406, IEEE.
Bana, S.V. and Varaiya, P. (2001) ‘Space division multiple access Flavio, B. (2013) The Smart and Connected Vehicle and the Internet
(SDMA) for robust ad hoc vehicle communication networks’, of Things, WSTS, San José, CA.
ITSC 2001. 2001 IEEE Intelligent Transportation Systems. Girinath, D.R. and Selvan, S. (2010) ‘A novel cluster based routing
Proceedings (Cat. No. 01TH8585), pp.962–967, IEEE. algorithm for hybrid mobility model in VANET’, International
Bernsen, J. and Manivannan, D. (2012) ‘River: a reliable Journal of Computer Applications, Vol. 1, No. 15, pp.35–42.
inter-vehicular routing protocol for vehicular ad hoc networks’, Hager, M., Wernecke, L., Schneider, C. and Seitz, J. (2015)
Computer Networks, Vol. 56, No. 17, pp.3795–3807. ‘Vehicular ad hoc networks: multi-hop information
Bujari, A., Gaggi, O., Palazzi, C.E. and Ronzani, D. (2018a) ‘Would dissemination in an urban scenario’, 2015 38th International
current ad-hoc routing protocols be adequate for the internet Conference on Telecommunications and Signal Processing
of vehicles? A comparative study’, IEEE Internet of Things (TSP), pp.65–70, IEEE.
Journal, Vol. 5, No. 5, pp.3683–3691. Ho, K-Y., Kang, P-C., Hsu, C-H. and Lin, C-H. (2010)
Bujari, A., Palazzi, C.E. and Ronzani, D. (2018b) ‘A comparison of ‘Implementation of WAVE/DSRC devices for vehicular
stateless position-based packet routing algorithms for FANETs’, communications’, 2010 International Symposium on Computer,
IEEE Transactions on Mobile Computing, Vol. 17, No. 11, Communication, Control and Automation (3CA), Vol. 2,
pp.2468–2482. pp.522–525, IEEE.
Carpenter, S.E. (2014) ‘Obstacle shadowing influences in VANET Huang, H., Yin, H., Luo, Y., Zhang, X., Min, G. and Fan, Q. (2016)
safety’, 2014 IEEE 22nd International Conference on Network ‘Three-dimensional geographic routing in wireless mobile ad
Protocols, pp.480–482, IEEE. hoc and sensor networks’, IEEE Network, Vol. 30, No. 2,
Chen, Y-S., Lin, Y-W. and Pan, C-Y. (2011) ‘DIR: pp.82–90.
diagonal-intersection-based routing protocol for vehicular ad Johnson, D.B. and Maltz, D.A. (1996) ‘Dynamic source routing
hoc networks’, Telecommunication systems, Vol. 46, No. 4, in ad hoc wireless networks’, Mobile Computing, pp.153–181,
pp.299–316. Springer, Boston, MA.
Contreras-Castillo, J., Zeadally, S. and Guerrero-Ibañez, J.A. (2017) Joshi, H.P. et al. (2007) ‘Distributed robust geocast: a multicast
‘Internet of vehicles: architecture, protocols, and security’, IEEE protocol for inter-vehicle communication’, in Proceedings of
Internet of Things Journal, Vol. 5, No. 5, pp.3701–3709. WEIRD Workshop on WiMax, Wireless and Mobility, Vol. 3,
Dehghani, M., Montazeri, Z., Dhiman, G., Malik, O., pp.27–30.
Morales-Menendez, R., Ramirez-Mendoza, R.A., Dehghani, A., Kaiwartya, O., Abdullah, A.H., Cao, Y., Altameem, A., Prasad, M.,
Guerrero, J.M. and Parra-Arroyo, L. (2020a) ‘A spring search Lin, C-T. and Liu, X. (2016) ‘Internet of vehicles: motivation,
algorithm applied to engineering optimization problems’, layered architecture, network model, challenges, and future
Applied Sciences, Vol. 10, No. 18, p.6173. aspects’, IEEE Access, Vol. 4, pp.5356–5373.
Dehghani, M., Montazeri, Z., Givi, H., Guerrero, J.M. and Karp, B. and Kung, H-T. (2000) ‘GPSR: Greedy perimeter stateless
Dhiman, G. (2020b) ‘Darts game optimizer: a new optimization routing for wireless networks’, Proceedings of the 6th Annual
technique based on darts game’, Int. J. Intell. Eng. Syst., International Conference on Mobile Computing and Networking,
Vol. 13, No. 6, pp.286–294. pp.243–254.
Dhiman, G. (2019) ‘ESA: a hybrid bio-inspired metaheuristic Khanna, B., Redi, J., Basu, P. and Ramanathan, R. (2013) ‘Disrupted
optimization approach for engineering problems’, Engineering adaptive routing: gossip-based routing in delay-tolerant
with Computers, Vol. 37, No. 1, pp.323–353. networks’, MILCOM 2013-2013 IEEE Military Communications
Dhaliwal, J.K., Naseem, M., Lawaye, A.A. and Abbasi, E.H. Conference, pp.1099–1104, IEEE.
(2018) ‘Fibonacci series based virtual machine selection for Kihl, M., Sichitiu, M. and Joshi, H.P. (2008) ‘Design and evaluation
load balancing in cloud computing’, International Journal of of two geocast protocols for vehicular ad-hoc networks’,
Engineering & Technology, Vol. 7, No. 3.12, pp.1071–1077. Journal of Internet Engineering, Vol. 2, No. 1, pp.127–135.
Dhiman, G. and Kaur, A. (2019) ‘STOA: a bio-inspired based Lee, U., Oh, S.Y., Lee, K-W. and Gerla, M. (2008) ‘Relaycast:
optimization algorithm for industrial engineering problems’, scalable multicast routing in delay tolerant networks’,
Engineering Applications of Artificial Intelligence, Vol. 82, 2008 IEEE International Conference on Network Protocols,
pp.148–174. pp.218–227, IEEE.
Dhiman, G., Garg, M., Nagar, A., Kumar, V. and Dehghani, M. Li, G. and Boukhatem, L. (2013) ‘A delay-sensitive vehicular
(2020) ‘A novel algorithm for global optimization: rat swarm routing protocol using ant colony optimization’, 2013
optimizer’, Journal of Ambient Intelligence and Humanized 12th Annual Mediterranean Ad Hoc Networking Workshop
Computing, pp.1–26. (MED-HOC-NET), pp.49–54, IEEE.
 Architecture and routing protocols for internet of vehicles: a review 175

Li, X., Xi, J., Cai, Z., Yang, T. and Cheang, C.F. (2019) Sharma, A. and Kumar, R. (2017) ‘An optimal routing scheme
‘Analyzing the structure and connectivity of continent-level for critical healthcare HTH services – an IoT perspective’,
internet topology’, Computers, Materials & Continua, Vol. 59, 2017 Fourth International Conference on Image Information
No. 3, pp.955–964. Processing (ICIIP), pp.1–5, IEEE.
Liu, J., Wan, J., Wang, Q., Deng, P., Zhou, K. and Qiao, Y. Sharma, A. and Kumar, R. (2019) ‘Service-level agreement – energy
(2016) ‘A survey on position-based routing for vehicular ad cooperative quickest ambulance routing for critical healthcare
hoc networks’, Telecommunication Systems, Vol. 62, No. 1, services’, Arabian Journal for Science and Engineering, Vol. 44,
pp.15–30. No. 4, pp.3831–3848.
Liu, N. (2011) ‘Internet of vehicles: your next connection’, Huawei Sheu, J-S., Chen, I-C. and Lio, Y-S. (2019) ‘Realization of internet
WinWin, Vol. 11, pp.23–28. of things smart appliances’, Intelligent Automation & Soft
Lochert, C., Mauve, M., Füßler, H. and Hartenstein, H. (2005) Computing, Vol. 25, No. 2, pp.395–404.
‘Geographic routing in city scenarios’, ACM SIGMOBILE Singh, S., Jeong, Y-S. and Park, J.H. (2016) ‘A survey on cloud
Mobile Computing and Communications Review, Vol. 9, No. 1, computing security: issues, threats, and solutions’, Journal of
pp.69–72. Network and Computer Applications, Vol. 75, pp.200–222.
Maglaras, L.A., Al-Bayatti, A.H., He, Y., Wagner, I. and Janicke, H. Soares, V.N., Rodrigues, J.J. and Farahmand, F. (2014) ‘Geospray:
(2016) ‘Social internet of vehicles for smart cities’, Journal of a geographic routing protocol for vehicular delay-tolerant
Sensor and Actuator Networks, Vol. 5, No. 1, p.3. networks’, Information Fusion, Vol. 15, pp.102–113.
Maidorawa, A. (2013) An Enhanced Connectivity Aware Routing Song, T., Xia, W., Song, T. and Shen, L. (2010) ‘A cluster-based
Protocol for Vehicular Ad Hoc Networks, PhD thesis, Universiti directional routing protocol in VANET’, 2010 IEEE 12th
Teknologi Malaysia. International Conference on Communication Technology,
Naseem, M. and Kumar, C. (2013) ‘EDSDV: efficient DSDV routing pp.1172–1175, IEEE.
protocol for MANET’, 2013 IEEE International Conference on Tonguz, O.K., Wisitpongphan, N., Parikh, J.S., Bai, F., Mudalige, P.
Computational Intelligence and Computing Research, pp.1–4, and Sadekar, V.K. (2006) ‘On the broadcast storm problem in ad
IEEE. hoc wireless networks’, 2006 3rd International Conference on
Naseem, M. and Kumar, C. (2014) ‘Distributed bandwidth guarantee Broadband Communications, Networks and Systems, pp.1–11,
using multi-path routing protocol in mobile ad hoc network’, IEEE.
2014 International Conference on Science Engineering and Wang, X., Jiang, J., Zhao, S. and Bai, L. (2019) ‘A fair blind
Management Research (ICSEMR), pp.1–5, IEEE. signature scheme to revoke malicious vehicles in VANETs’,
Naseem, M. and Kumar, C. (2015) ‘Congestion-aware Fibonacci Comput. Mater. Continua, Vol. 58, No. 1, pp.249–262.
sequence based multipath load balancing routing protocol for Xiang, Y., Liu, Z., Liu, R., Sun, W. and Wang, W. (2013) ‘GeoSVR:
MANETs’, Wireless Personal Communications, Vol. 84, No. 4, a map-based stateless vanet routing’, Ad Hoc Networks, Vol. 11,
pp.2955–2974. No. 7, pp.2125–2135.
Naseem, M. and Kumar, C. (2017) ‘Queue-based multiple path load Yang, Q., Lim, A., Li, S., Fang, J. and Agrawal, P. (2010) ‘ACAR:
balancing routing protocol for MANETs’, International Journal adaptive connectivity aware routing for vehicular ad hoc
of Communication Systems, Vol. 30, No. 6, p.e3141. networks in city scenarios’, Mobile Networks and Applications,
Naumov, V. and Gross, T.R. (2007) ‘Connectivity-aware routing Vol. 15, No. 1, pp.36–60.
(CAR) in vehicular ad-hoc networks’, IEEE INFOCOM Yokoyama, R.S., Kimura, B.Y., Jaimes, L.M. and Moreira, E.D.
2007-26th IEEE International Conference on Computer (2014) ‘A beaconing-based opportunistic service discovery
Communications, pp.1919–1927, IEEE. protocol for vehicular networks’, 2014 28th International
Perkins, C.E. and Royer, E.M. (1999) ‘Ad-hoc on-demand Conference on Advanced Information Networking and
distance vector routing’, Proceedings WMCSA’99. Second IEEE Applications Workshops, pp.498–503, IEEE.
Workshop on Mobile Computing Systems and Applications, Zhao, J. and Cao, G. (2008) ‘VADD: vehicle-assisted data delivery
pp.90–100, IEEE. in vehicular ad hoc networks’, IEEE Transactions on Vehicular
Poongodi, M., Hamdi, M., Sharma, A., Ma, M. and Singh, P.K. Technology, Vol. 57, No. 3, pp.1910–1922.
(2019) ‘DDoS detection mechanism using trust-based evaluation Zhongyi, L., Tong, Z., Wei, Y. and Xiaoming, L. (2009) ‘GOSR:
system in VANET’, IEEE Access, Vol. 7, pp.183532–183544. geographical opportunistic source routing for VANETs’, ACM
Senouci, O., Aliouat, Z. and Harous, S. (2019) ‘A review of routing SIGMOBILE Mobile Computing and Communications Review,
protocols in internet of vehicles and their challenges’, Sensor Vol. 13, No. 1, pp.48–51.
Review, Vol. 39, No. 1, pp.58–70.
Shafiee, K., Attar, A. and Leung, V.C. (2011) ‘Optimal distributed
vertical handoff strategies in vehicular heterogeneous networks’,
IEEE Journal on Selected Areas in Communications, Vol. 29,
No. 3, pp.534–544.