A Survey on Routing Protocols in SDN - VANET
Rohini Balaji Augastiny Ramesh Dr. P. Samundiswary
Department of Electronics Engineering Department of Electronics Engineering Department of Electronics Engineering
Pondicherry University Pondicherry University Pondicherry University
Puducherry, India Puducherry, India Puducherry, India
rohinibalaji399@[Link] sweetaugas@[Link] [Link]@[Link]
Abstract—Vehicular Ad-hoc Networks (VANETs) hold
immense potential for Intelligent Transport Systems (ITS) by
facilitating communication between vehicles and roadside
infrastructure. The vehicular network has many advantages
that can revolutionize transportation systems, such as
improved road safety, enhanced traffic management,
cooperative driving applications, efficient emergency response,
infotainment services, etc. However, like other communication
networks VANETs also face challenges in areas of routing,
mobility and security. This survey mainly focuses on the
routing aspects of VANETs which is essential for reliable
communication. But as the VANET network is dynamic there
can be frequent disruptions in connections which may lead to
unreliable data delivery. To tackle such problems in VANET
routing, Software Defined Networking (SDN) can be integrated
with it. SDN is a centralized approach that offers advantages
such as dynamic routing, improved efficiency and reduced FIGURE 1. VANET Architecture
latency over traditional VANET routing protocols. Thus this
study is about SDN assisted VANET routing protocols.
Routing protocols act as the invisible traffic directors
within a VANET, ensuring that information reaches its
Keywords—VANET, Routing Protocols, Software Defined intended recipient efficiently and swiftly. [1] Routing
Networking, Road Safety, Efficiency. protocols in VANET can be broadly classified into five
categories - topology based, position based, broadcast based,
I. INTRODUCTION clustering based and geocast based routing protocols. Among
VANET is a network where vehicles communicate with these topology based such as Ad hoc On-Demand Distance
each other and with the Road Side Units (RSU). Fig.1 depicts Vector (AODV), Optimized Link State Routing (OLSR) and
the architecture of VANET. There are two main types of position based routing protocols [2] like Greedy Perimeter
communication in VANETs. Vehicle to Vehicle (V2V) Stateless Routing (GPSR), Greedy Perimeter Coordiator
communication where the vehicles share information with Routing (GPCR) are widely used routing protocols in
each other like traffic information, accident and safety VANET. Thus these classifications provide different routing
information etc. Another type of communication is Vehicle protocols in VANET but also have their trade-offs. Unlike
to Infrastructure (V2I) communication in which the vehicles traditional networks, VANETs present unique challenges for
communicate with RSU. VANET relies on an architecture routing protocols due to their characteristics such as high
that facilitates communication between the vehicles and the mobility, limited transmission range, scalability and limited
roadside infrastructure that consists of an On Board Unit resources. These challenges necessitate the development of
(OBU), RSU and Application Unit (AU). The OBU is specialized routing protocols for VANETs.
installed in the vehicles and is responsible for collecting data To overcome such challenges, a centralized controller
such as vehicle speed, location, direction etc. This collected like SDN can be integrated with VANET. [10] SDN is
information is processed and transmitted as messages to controller whose data plane and control plane are separated.
other vehicles and RSUs. The RSUs are deployed along the The rules are created in the controller before sending the data
roadsides that act as a gateway between the vehicles and the to the vehicles or RSUs in the network. SDN uses Open Flow
network. The AU is equipped within the vehicle which acts (OF) protocol which is utilized in communication interface
as a bridge between applications and communication system. between the data and control planes. Fig. 2 shows the basic
The need for VANET stems from the challenges faced by architecture of SDN [10]. It consists of three layers -
the modern transportation systems such as traffic congestion, application layer, control layer and the infrastructure layer.
accidents, and a lack of real-time information that creates The application layer is the top most layer of SDN which is
frustrating and often dangerous environment for drivers. made up of software programs that define the network
VANET addresses these issues by providing enhanced road behaviour and control the underlying hardware through the
safety, improved traffic efficiency, cooperative driving, SDN controller. The control layer is the brain of SDN which
emergency response optimization and more. Though acts as a central point of communication between the
VANETs have the potential to improve the efficiency of application layer and the data plane. The control layer
ITS, it still faces challenges mainly in the area of routing, receives instructions from the upper layer. Based on the
security and mobility. instructions the network devices are programmed.
1
reception of packets using Network Simulator 3 (NS3)
software. End results show that AODV outperforms OLSR
and DSDV protocols.
Ankita Srivatsava et al [3] compared location based
routing protocols GPSR and GPCR. Location based routing
has an advantage over other methods which is, it does not
rely on route entries. Also, these protocols face challenges
and issues such as imprecise location selection, choosing a
best forwarder and managing overheads.
Houmer et al [4] proposed a protocol called Maxduration
Minangle GPSR to over come the drawback of traditional
GPSR of selecting the next hop neighbour. MM - GPSR is
implemented to both greedy forwarding and perimeter
forwarding of GPSR. The performance evalution takes into
consideration the following metrics: Packet loss ratio, end to
end delay and throughput. The proposed MM - GPSR
protocol works better but still its performance can be
FIGURE 2. SDN Architecture optimized.
Also the control layer maintains the network states of the An Enhancement for GPSR in VANET is proposed by Y
connected devices and communicates with the data plane. ang et al [5] called E-GPSR. Traditional GPSR protocol has
The infrastructure layer lso known as the data plane of SDN two modes: Greedy mode and Perimeter mode. In E - GPSR,
consists of physical network devices such as switches and instead of perimeter mode, it makes use of the distance
routers which are responsible for actual data forwarding between the source node and the intended destination node
according to instructions received from the control layer. and the distance between the source node and next hop for
SDN integrated with VANET offers numerous selecting the next hop node. The proposed method shows
advantages such as dynamic route optimization, improved improved QOS metrics than traditional GPSR protocol.
security, flexibility and programmability and more. Yang et al [6] focused on multi-objective optimization of
Traditional VANET routing protocols struggle with OLSR protocol in VANET. Traditional OLSR protocol is not
scalability. By the use of SDN’s centalized control, it can suitable for VANET due to its high mobility. Optimization of
handle large number of vehicles and manage the network OLSR protocol with Particle Swarm Optimization technique
effectively. The programmability of SDN allows for the is done. Results show that the optimized OLSR protocol
creation of routing protocols that are more efficient in terms performs better than traditional OLSR protocol in terms of
of bandwidth usage and processing power required on QOS metrics such as PLR, throughput, end to end delay and
individual vehicles. This is crucial for VANETs, where delivery cost.
vehicles have limited battery life and computational
resources. A bio inspired routing optimization is proposed by
Azzoug et al [7] which gives a overview of challenges faced
Overall SDN is a powerful approach that can be used to in VANET routing and a detailed view on bio- inspired
address the limitations of traditional VANET routing optimization techniques. An overview of existing and
protocols. The centralized control, programmability, and emerging optimization techniques inspired from nature such
scalability of SDN can significantly improve routing as PSO, GA and more are discussed. With the use of
efficiency, enhance network security, and pave the way for optimization techniques like these, integrated with existing
innovative applications in the future of connected and routing algorithms the overall performance of the protocols
autonomous vehicles. in VANET can be improved.
This study is focused mainly on VANET routing protocols Singh et al [8] proposed an hybrid Genetic Firefly (GF)
and the rest of the paper is organized as follows. Section II is based routing Protocol for VANET. Traditional protocols
the literature review on SDN assisted VANET routing such as AODV, OLSR, and DSDV are can be utilized for
protocols. Section III shows the performance comparison of generic mobile nodes but are not suitable for VANET due to
different protocols and section V concludes the paper. their dynamic nature. Hence this hybrid approach is
proposed. This type of optimization technique proves to be
more efficient when used in routing but increases the
II. LITERATURE REVIEW complexity of the protocol.
K. Singh et al [1] has done the survey on VANET routing A survey on QoS routing protocols in VANET is done by
protocols which gives detailed view on different routing Belamri et al [9]. This paper is a survey on the improvement
protocols for vanet. Routing protocols in VANET are of QOS metrics of routing protocols in [Link] the
broadly classified into 5 types based on topology, position, routing protocols are divided into QoS based protocols by
broadcast, clustering and geocast. VANET is a network with meta-heurestics approach. The most used meta heurestics
distinct characteristics like high mobility, unstable methods are ACO, PSO, GF and GA which can be combined
connectivity etc, hence efficient routing protocols are with the routing protocols for optimized performance.
required.
Al-Heety et al [10] elaborated the benefits, services
Kumar et al [2] performed an analysis on different provided, issues and challenges and how SDN-VANET can
routing protocols in VANET such as AODV, OLSR and be made use of. This study is about integrating SDN with
DSDV. The analysis is made in terms of QOS parameters VANET. In SDN, the data and the control plane are
such as delivery ratio of the packet, delay and successful separated. By integrating SDN with VANET enhanced
2
security, reduced latency, improved routing protocols and used in VANET as it combines the capabilities of cognition
more are achieved. together with SDN. It consists of two phases: the register
phase and the route prediction phase. Experimental findings
K. Mershad [11] explored SDN based routing protocols.
showed better performance in terms of respected key metrics.
To improve the stability of VANET routing protocols SDN is
introduced here Two algorithms are implemented in this Rajashekar et al [19] discussed the next generation
paper: SURFER - I and SURFER - II. The use SDN in Road technologies empowered for future IoV. Next-generation
side units (RSU) shows improved performance than existing technologies such as 5th Generation (5G) and SDN
protocols. technologies can significantly address the challenges of
vehicular networks. These technologies can provide potential
Zhu et al [12] initiated a SDN based routing framework
solutions for efficient management, enhance the performance
for message transmission with a routing protocol called
of IoV. Simulation results show that 5G and SDN have lower
Centralized Routing Protocol (CRP). SDN is a developing
transmission delay and higher throughput.
technology where the data and the control plane are
separated. All the switches and are connected to a central Traffic alert system based on SDN for smart cities is
controller. Compared with existing protocols CRP performed proposed by Raja et al [20]. In this paper the RSUs are
well. placed at the intersection and considered to have a smooth
functioning. Road users can get traffic information through
X. Ji et al [13] introduced a SDN-based geographic
SDN. If an accident occurs, alert messages are transmitted to
routing protocol for VANET based on location of the nodes,
vehicles and RSUs in the perimeter. Thus this system
density of vehicles and digital map. An SDN controller is
highlights the decrease in traffic congestion and travel
implemented in which optimum path forwarding algorithm
duration.
is used. The SDGR scheme is analysed in NS2 and compared
with existing routing protocols (AODV and GPSR) in Alouache et al [21] proposed an hybrid SDN architecture
VANET. The results show that SDGR achieves better for V2V communication. The proposed hybrid SDN
performance in terms of both PDR and delivery delay time. architecture aims in enhancing the communication reliability
and availability which in turn reduces the link failure and
SDN Architecture for ITS is proposed by Sadio et al [14].
reducing resolution cost while maintaining a good end to end
A traffic flow model called Kraus model is used to predict
delay. The results show that this hybrid SDN outperforms
the path of the vehicles and the information has to be sent to
other protocols.
the SDN controller periodically. The SDN controller
generates flow tables based on the predicted topology. Increased use of fog computing in IoV is suggested by
Python script is used for analysis and the evaluation shows Kadhim et al [22]. A new system merging fog computing and
that the proposed architecture outperforms OLSR and GPSR SDN tackles tasks from IoV sensors locally, reducing
in terms of PDR and overhead. reliance on the cloud and minimizing response times. This
approach optimizes bandwidth between fog and cloud while
Noorani et al [15] proposed a routing method that makes
also incorporating load balancing to minimize idle fog
use of the capabilities of SDN and fog computing. A routing
servers. This makes it ideal for bandwidth-limited networks
method called SFIR is introduced. It is based on the road
and applications requiring low latency.
intersections which leverages the potential of both SDN and
fog. Here the shortest path is selected for message Hai Tao et al [23] advocated a SDN-assisted technique
transmission. In turn SFIR can reduce the challenges faced for control of traffic and executing information in VANET.
by traditional systems such as encountering a local The concept of this paper is to leverage SDN’s capability to
maximum, dispersion and low vehicle density. The results monitor the infrastructure of the network and redirecting the
indicate improvement in PDR, PLR, and delay. traffic on the metrics such as number of nodes, bandwidth
and overall node utilization.
The integration of SDN with Internet of Vehicles (IoV) is
initiated by Sarra et al [16]. To address the issues of An SDN controller implementing dynamic shortest path
traditional routing protocols for VANET, latest technologies routing called Pox Plus is put forward by Alshammari et al
such as: IoV, Internet of Things (IoT) and SDN can be [24] . It is a SDN controller built upon the widely used POX
employed. Combining SDN and IoV will improve future controller which introduces a novel routing method called
ITS. Low latency is crucial in this combination. SDN-IoV [Link] calculates and manages paths between the nodes
ensures different services, like security and streaming, get the within the SDN infrastructure. Findings indicate that DR-
level of speed and reliability they need. APSP notably enhances the efficiency of SDN routing.
An SDN-Enabled Spectral Clustering-Based Optimized Gopi et al [25] analysed SDN and conventional networks
Routing Using Deep Learning in VANET Environment is using routing protocols. In traditional networks, when a link
introduced by Nahar et al [17]. Here spectral clustering is or node fails, it takes longer for the routing to stabilize as the
responsible for formulating stable clusters. In the first phase, network gets bigger. This is because information about the
clustering is done and in the second phase, SDN is used to failure spreads through the network by flooding, which
obtain an efficient route to the destination vehicle in a means it takes more time to update all the routers' routing
topological scenario. This approach shows improved tables. Analysis shows that flooding takes more time as the
throughput and low latency than conventional protocols. network grows, it slows down the routing stabilization
process.
Integration of SDN with Cognitive Radio (CR) is
discussed by Ghafoor et al [18]. This protocol is unique to be
III. PERFORMANCE COMPARISON
3
Ref. No Compared Protocols Inference Performance Metrics Tools used
PDR Delay Throughput
Use of SDN shows
[11] SURFER improvement in PDR 85% 1.25 s - NS3, SUMO
and has low delay.
Delay is reduced in
CRP, AODV, OLSR, CRP compared to
[12] - 1.8 s - NS3, SUMO
GPSR traditional routing
protocols.
SDN based protocol
SDGR shows better
[13] SDGR, AODV, GPSR 90% 1.2 s - NS2, SUMO
performance in PDR
and delay.
87% PDR is achieved
[14] SDN, GPSR, OLSR 87% - - Python
by using SDN.
SFIR an SDN based
SFIR,AODV,GPSR,T routing has shown
[15] 60% 2.5 s - NS2, OSM
ORA improved performance
in PDR and delay.
SeScr has an end to
OMNET++,
[17] SeScr end delay of 40 ms to - 40 ms 700 kb/s
SUMO
700 kb/s.
Hybrid SDN shows
improved PDR and the
[18] HSDV delay is less 70% 2.0 s - NS2
SDN based routing has
HSDN, HA-DSDV, less delay compare to NS2,
[21] - 2.5 s -
DSDV traditional routing VanetMobiSim
methods
This method achieves
[23] SDN more packet delivery 90% - - NS3, SUMO
ratio by using SDN
VANETs, characterized by high mobility and frequent link
disruptions. SDN-based routing protocols offer several
advantages over traditional approaches. Centralized control
IV . CONCLUSION by the SDN controller enables optimal routing decisions
based on real-time traffic information. The programmability
In this survey the potential of SDN for routing protocols
of SDN allows for flexible and customized routing protocols
in VANETs is explored. Traditional routing protocols have
tailored to specific VANET applications and network
limitations in handling the highly dynamic nature of
4
conditions. Dynamic adaptation capabilities of SDN
controllers allow for swift adjustments to routing strategies
in response to frequent topology changes caused by vehicle
movement.
SDN-based routing protocols offer a promising solution
for overcoming the limitations of traditional routing
approaches in VANETs. While challenges remain, further
research and development efforts can pave the way for a
more efficient, reliable, and secure communication
framework for a smarter and safer transportation future.