Received 12 December 2024, accepted 26 January 2025, date of publication 3 February 2025, date of current version 7 February 2025.

Digital Object Identifier 10.1109/ACCESS.2025.3537800

Navigating the Challenges and Opportunities of

Securing Internet of Autonomous Vehicles With
Lightweight Authentication
HAZEM ISMAIL ALI 1 , HARRISON KURUNATHAN 2 , MOHAMED HAMDY ELDEFRAWY 1,

FLAVIUS GRUIAN 3 , AND MAGNUS JONSSON 1 , (Senior Member, IEEE)

1 Department of Computing and Electronics for Real-Time and Embedded Systems (CERES), School of Information Technology (ITE), Halmstad University,
301 18 Halmstad, Sweden
2 CISTER/ISEP, Polytechnic Institute of Porto, 4200-465 Porto, Portugal
3 Department of Computer Science, Lund University, 221 00 Lund, Sweden

Corresponding author: Hazem Ismail Ali (hazem.ali@hh.se)

This work was supported in part by Halmstad University (Halmstad i Högskolan) Research Base Funding; in part by NordForsk
86220 through the Proposal HI2OT: Nordic University Hub on Industrial Internet of Things; in part by the Research Centre in Real-Time
and Embedded Computing Systems (CISTER) Research Unit financed by National Funds through Fundacao para a Ciencia e a Tecnologia
(FCT)/Ministry of Science, Technology and Higher Education (MCTES) (Portuguese Foundation for Science and Technology) under Grant
UIDP/UIDB/04234/2020; and in part by the Project Route 25 funded by the EU/Next Generation within Call 02/C05-i01/2022 of the
Recovery and Resilience Plan (RRP) and Hardware Abstraction Layer for a European Software Defined Vehicle Approach (HAL4SDV)
Funding within the Chips Joint Undertaking (Chips JU)—The Public-Private Partnership for Research, Development and Innovation under
Horizon Europe—and National Authorities under Grant TRB/2022/00061-C645463824-00000063 and Grant 101139789.

ABSTRACT The Internet of Things (IoT) can be defined as the network of physical objects, or ‘‘things,’’
embedded with sensors and software for processing and exchanging data with other devices and ecosystems
using the Internet as a medium. With its rapid growth over the past decade, it has permeated several
application domains, including intelligent vehicular systems. The Internet of Autonomous Vehicles (IoAV)
is a subset of IoT that envisions dynamic autonomous driving without human intervention. The dynamic
nature of the environment in which autonomous vehicles operate introduces significant challenges, such
as real-time communication and security vulnerabilities. These challenges cannot be directly addressed by
standard cybersecurity solutions designed primarily for static IoT environments. In this work, we outline
the various vulnerabilities of the IoAV systems, and we delve into the critical importance of adopting
lightweight security protocols. These protocols are crucial to ensure robust protection while at the same
time not jeopardizing the performance of the IoAV system. We also highlight the fast lightweight security
protocols implemented on heterogeneous embedded, low-power, high-performance computing platforms as
a viable solution to address these challenges.

INDEX TERMS Lightweight authentication, Internet of Autonomous Vehicles (IoAV), embedded systems.

I. INTRODUCTION The Internet of Autonomous Vehicles (IoAV) enables

The concept of autonomous vehicles was envisioned to real-time communication between autonomous vehicles
become a reality by the late 20th century [1]. Human and their supporting infrastructures. This communication
error, vehicle design and infrastructure were considered encompasses crucial information on the surroundings, traf-
to be the main causes of accidents, fuelling the idea fic conditions, and security data that are vital to ensur-
of machine-substitution of the human driver [2]. This ing the safe and efficient functioning of autonomous
led to technologies such as cruise control and eventually vehicles.
autonomous-driven vehicles enabled through the internet. As depicted in Figure 1, IoAV can be compartmentalized
into three types of communication, namely vehicle-to-vehicle
The associate editor coordinating the review of this manuscript and (V2V), vehicle-to-infrastructure (V2I), and vehicle-to-cloud
approving it for publication was Mohamad Afendee Mohamed . (V2C) communication [3].
2025 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
VOLUME 13, 2025 For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/ 24207
 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

FIGURE 1. IoAV encompassing V2V, V2I and V2C real-time communications.

Vehicle-to-vehicle (V2V) communication enables imposing a significant burden on the system’s CPU or
autonomous vehicles to exchange critical data, such as memory. In the context of V2V and V2I communications,
real-time information on precise location, speed, intended we identify the security threats and offer insights into the
path, and security data. This creates situational awareness suitability of these protocols for deployment in AVs.
among autonomous vehicles allowing them to adapt to An Internet of Autonomous Vehicles (IoAV) system is a
dynamic traffic conditions. V2V communication enables networked ecosystem in which autonomous vehicles (AVs)
multiple autonomous vehicle applications such as lane are interconnected through the internet, enabling them to
changes, braking manoeuvres, and speed adaptation based communicate with each other, with infrastructure, and with
on road conditions. With coordinated communication, road external networks. This connectivity enables data transfer,
congestion and bottlenecks at intersections can be avoided. subsequently enhancing the operational efficiency, safety,
Furthermore, V2V also facilitates driving methods such and overall functionality of autonomous vehicles. With this
as platooning [4], which reduces aerodynamic drag and interconnectivity, such an ecosystem is highly vulnerable
improves fuel efficiency. to security threats [12]. In this work, we aim to survey
Vehicle-to-Infrastructure (V2I) communication enables the the key cybersecurity threats impacting road traffic safety.
vehicles to communicate with roadside units such as traffic Authentication is a critical process of IoAV as it is used for
lights and traffic management systems. Autonomous vehicles identity verification by coordinating communication between
can use V2I to share real-time traffic data and coordinate with on-board vehicles, infrastructure, and remote servers.
other vehicles connected to the infrastructure [5]. Through Over the past decade, traditional [8], [13] and deep learning
V2I communication, the vehicles can get an idea of traffic methodologies have been used for the detection of vehicles,
conditions such as road closures, traffic lights, construction pedestrians, and road lanes, and the field of autonomous
zones, weather, and speed zones in advance to improve vehicles has consistently evolved. Recently, machine learning
overall navigation and safety. Furthermore, this V2I can and deep learning methods have been used extensively in
also reduce the bottlenecks at road intersections, enabling authentication mechanisms within the IoAV ecosystem due
smoother traffic flow and reduced fuel consumption [6], [7]. to their ability to detect patterns, analyze complex data, and
Finally, Vehicle-to-Cloud (V2C) communication allows adapt to new threats. For example, researchers in [14] propose
vehicles to constantly upload and access information on a novel federated learning collaborative authentication-based
navigation data, sensors, and diagnostics [8], [9]. Akin to protocol to secure communication in IoAV vehicles. This
V2I communication, V2C can also aid in traffic data coor- protocol reduces the number of vehicle certifications for
dination. Connectivity through the cloud enables over-the- each dynamic Roadside Unit (RSU) while ensuring the
air software updates and remote diagnostics for autonomous safety of vehicle upload training parameters in the feder-
vehicles. Combined with the scope of machine learning, V2C ated architecture. They also include an anonymous mutual
communication can enable predictive analysis of vehicular authentication and key agreement. In another work [15], they
health monitoring for autonomous vehicles [10], [11]. used novel deep learning models that include algorithms like
In this work, we aim to explore lightweight authentication the Deep Sparse Stacked Autoencoder Network (DS2AN)
techniques employed in V2V and V2I communications, algorithm to ensure the authentication and security of
specifically focusing on minimizing computational load the vehicles. Furthermore, in [16], they developed trust
while maintaining adept security. We place a particular management approaches for software-defined vehicles. This
emphasis on authentication protocols which require fewer novel architecture works with a centralized software-defined
computational resources (i.e., lightweight) to execute without network controller that serves as a learning agent and obtains

24208 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

the optimal communication link policy using a deep learning associated state-of-the-art defences. Table 1 presents some of
approach. Therefore, with the ever-growing complexity and the significant attack types, targeted vehicle parts, and their
connectivity of these automotive systems, there is a mandate respective countermeasures.
for robust authentication to ensure the integrity and security
of data. lightweight authentication protocols are crucial to A. DENIAL-OF-SERVICE (DoS) ATTACKS
maintaining the performance of IoAV without burdening the Denial-of-Service (DoS) attacks target communication chan-
limited computational resources. nels by overwhelming them with an excessive amount of
Traditionally, intense cryptographic operations used in traffic requests (Figure 2). These traffic requests lead to
authentication mechanisms strain limited processing power degradation and disruption of services. This deterioration
and memory [17]. It is essential to look into methods such as results in significant operational risks for autonomous
hardware acceleration, protocol optimization, precomputed vehicles on the road [19].
authentication, and dynamic resource allocation to reduce
the computational strains due to intensive cryptographic
operations. This survey aims to provide an in-depth analysis
of security threats and their respective mitigation strategies.
We identify the most critical security issues surrounding V2V
and V2I communication in IoAV and explore state-of-the-
art solutions to achieve efficient and reliable authentication
protocols.
The contributions of this work are as follows:
• We outline the various vulnerabilities of the IoAV
systems and discuss several cybersecurity attacks and
their targeted regions of the IoAV networks.
• We stress the crucial importance of adopting lightweight FIGURE 2. Denial-of-Service (DoS) attacks target communication
security protocols by showcasing several real-life channels and attempt to disrupt their functionality by overwhelming
them with an excessive amount of traffic requests.
examples and discussing the concepts of lightweight
protocols.
DoS attacks can occur through flooding V2X commu-
• We highlight fast lightweight security protocols imple-
nication channels resulting in congestion that leads to
mented on heterogeneous embedded, low-power, high-
excessive critical delay. This can result in catastrophic
performance computing platforms as a viable solution
events such as the loss of critical messages, preventing
and present several research directions in line with
vehicles from receiving essential safety information and
system design, blockchain solutions and cryptographic
coordination signals. Attackers can also specifically attack
functions.
the crucial computing resources within IoAV components
The rest of this article is organized as follows: in Section II, and vehicular internal networks. These targets encompass
we analyze some of the significant IoAV threats in the the onboard processing units of autonomous vehicles or the
literature, then we discuss some real-life use cases of Controller Area Networks (CAN) responsible for internal
authentication-based threats in Section III. Protocols enabling communication.
vehicular communication and their key security issues are The authors of [20] highlight a noteworthy example
elaborated in Section IV. In Section V, we discuss some of such an attack. They present a selective DoS attack
of the state-of-the-art lightweight authentication protocols, that specifically targets the CAN standard. This method
their advantages, their features and security issues. Then in is notable because it does not necessitate the exchange of
Section VI, we discuss the challenges and provide the aspects complete frames for execution, rendering it undetectable
of designing a secure authentication protocol in line with via frame-level analysis. This stealthy nature of the attack
IoAV systems. makes it incredibly challenging to identify and counteract,
exacerbating the potential risks and consequences for IoAV
II. IoAV SECURITY THREATS systems.
Security threats in IoAV are a significant risk to the safety,
privacy, and functionality of the interconnected vehicles [18]. B. SYBIL ATTACKS
It is vital to understand these threats in order to design The Sybil attack (Figure 3) is a significant security concern
defence mechanisms that enhance the safety of vehicles. in the IoAV context [3]. This attack derives its name from
To mitigate the security risks introduced by the current ‘‘Sybil,’’ a book by Flora Rheta Schreiber [21], presenting
growth in communication required for IoAV, a number of a case study of a woman with dissociative identity disorder
countermeasures have been proposed, including encryption, who exhibited multiple personalities. In the context of
authentication, filtering and intrusion detection. In the IoAV, a Sybil attack occurs when a malicious entity creates
following, we explore different security threats along with multiple fake identities (known as Sybil nodes) to deceive the

VOLUME 13, 2025 24209

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

TABLE 1. Different attack types on IoAV and their respective countermeasures.

IoAV system. These counterfeit identities, masquerading as autonomous vehicles or between vehicles and infrastructure
legitimate vehicles or infrastructure, deceive the system into components (Figure 4). Then, the adversary re-exchanges
trusting them. these copied packets, as they are or with tiny changes, back
into the network. Replayed packets, which appear genuine
to the IoAV system, can carry critical safety instructions or
control commands in which autonomous vehicles may act in
unsafe manoeuvres, leading to collisions on the road.
To mitigate replay attacks, IoAV can employ cryptographic
primitives and security measures such as including times-
tamps and sequence numbers in data packets. Timestamps
and sequence numbers can help detect and discard replayed
packets with outdated or duplicate values. In the same
context, moulding a time-to-live parameter for packets can
limit their validity period, reducing the possibility window
for such attacks [24], [25]. It can also be mitigated by
using digital signatures or message authentication codes
(MACs) that can verify the authenticity and integrity of data
FIGURE 3. Sybil attack occurs when a malicious entity creates multiple packets, making it harder for attackers to forge or replay
fake identities (known as Sybil nodes) to deceive the IoAV system. These
result in the legitimate vehicles taking wrong actions due to false them. Another security measure is the usage of cryptographic
information. nonces or challenge-response methods that can ensure each
communication session is unique, preventing attackers from
The Sybil nodes use these deceptive identities to gain reusing intercepted data.
control over the IoAV network, compromise its integrity,
and manipulate its operations for malicious purposes. For
example, Sybil nodes can inject false data into the IoAV D. BLACKHOLE ATTACKS
network to misguide autonomous vehicles, leading to poten- A Blackhole attack [3] is a security threat where malicious
tially dangerous decisions and movements on the road. Also, entities, called blackhole nodes, drop data packets instead
by generating many Sybil nodes, attackers can overwhelm the of forwarding them. This can have dangerous consequences
IoAV system resources, leading to DoS attacks. like missing traffic accident information. In such an attack,
Several defence mechanisms, such as solid authentication the malicious actor deceives other nodes in the network to
and digital signature methods, can mitigate Sybil attacks transfer their network traffic data to blackhole nodes by
in IoAV [22]. Also, AI-based approaches, such as anomaly falsely claiming that these nodes have certain advantages in
detection algorithms and behaviour analysis, can detect the network, such as the shortest path to their final destina-
and separate suspicious activities associated with Sybil tions. Blackhole attacks can have serious consequences, such
nodes. Furthermore, location-based detection using a direc- as dropping all incoming communication packets, leading
tional antenna with beam-forming techniques can also be to a failure in communication and coordination within the
employed [22]. IoAV components that increases the risk of accidents [26].
Blackhole attacks in IoAV can be prevented using [27]:
C. REPLAY ATTACKS (i) Intrusion Detection Systems (IDS) to detect abnormal
Replay attacks [3], [23] are a security concern in IoAV. The traffic patterns and identify blackhole nodes based on their
adversary copies legitimate data packets exchanged between dropping behaviour. (ii) Secure and authenticated routing

24210 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

FIGURE 4. The adversary copies legitimate data packets exchanged between autonomous
vehicles or between vehicles and infrastructure components. Then, the adversary re-exchanges
these copied packets, resulting in catastrophes like collision.

protocols can be used to ensure that data packets are Unfortunately, when the telematics system of the vehicle lost
routed through authorized nodes, avoiding possible blackhole its cell signal, the control unit of the car became inaccessible.
nodes. (iii) Implementing low-cost lightweight encryption In this instance, there is a possibility of a potential attacker
and authentication protocols can protect data packets from manipulating the controls using the fob’s shortened distance
being tampered with or read by blackhole nodes. signal [31]. This is a possibility because the ‘‘sender’’ (the
car) challenges the ‘‘prover’’ (the key fob), and then the
E. IMPERSONATION AND PASSWORD ATTACKS ‘‘sender’’ can read how long it has taken to receive an
In IoAV, impersonation attacks [27] involve malicious answer since the initial challenge was sent. Lightweight
nodes pretending to be legitimate vehicles. These attacks authentication protocols, specifically RFID-based Distance
utilize authentication violations to control weaknesses in the Bounding Protocols [32], can effectively mitigate this threat.
security mechanisms by forging the malicious node identity These protocols use cryptographic primitives to measure the
to achieve unauthorized access to manipulate the IoAV time it takes for a signal to travel between the key fob and the
system. IoAV is susceptible to secret key disclosure attacks car, ensuring that only devices within a specified distance can
where the intruders focus on compromising the encryption authenticate. This prevents relay attacks by ensuring that the
keys to impersonate themselves as authorized nodes. Offline authenticating device is physically close to the vehicle.
guessing attacks target authentication mechanisms and cryp-
tographic keys to help attackers guess passwords, keys, B. TESLA PHISHING VULNERABILITIES
or other sensitive information offline without interacting A case study on credential theft done by [33] demonstrates
directly with the system [28]. Security measures like the vulnerabilities in vehicular systems through infotainment
solid authentication methods, digital signatures, secure key systems in IoAV-enabled cars. Several companies, like Berla,
management, public key infrastructure (PKI) with digital offer sophisticated tools for data extraction from major
certificates, and secure Over-the-Air (OTA) [29] updates can vehicle brands, like Tesla. Such data extraction methods can
help to stop or mitigate such attacks in IoAV. be violated to perform credential theft of critical information
from state-of-the-art vehicles with modern assistant systems
III. VEHICLE CYBERSECURITY THREATS: REAL-LIFE CASE like the Tesla autopilot.
STUDIES Using phishing attacks in IoAV, client-level access can be
In this section, we explore some real-life instances of security stolen from the vehicle’s business platform. This can even
vulnerabilities in vehicular networks due to weaknesses in lead to the theft of the vehicle. Researchers from Mysk
authentication systems. Inc [34]. demonstrated how connected autonomous vehicles
could be susceptible to theft through a man-in-the-middle
A. OAKLAND FOB MANIPULATION INCIDENT (MITM) phishing attack. This phishing attack, as depicted in
Key fobs are wireless devices that allow vehicle owners to Figure 5 can enable malicious actors to create and use new
lock, unlock, start and even park their vehicles remotely. digital keys to unlock Tesla cars and access their systems.
Fob manipulation refers to unauthorized control of vehicles The attack starts with the attacker issuing a counterfeit
using key fobs. A technical journalist purchased a car WiFi access point (e.g. ‘‘Guest WiFi’’) at a vulnerable
through a local car-sharing program that provides Toyota location like a charging station. When the vehicle owner
Prius and electric Chevrolet Bolt EVs with plans to spend the enters their real credentials on the spoof page, which
weekends in rural areas about 3 hours north of Oakland [30]. resembles the real website, the credentials are immediately

VOLUME 13, 2025 24211

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

FIGURE 5. Phishing attack in an IoAV-enabled network where user credentials are

stolen using a fake guest WiFi terminal.

stolen. Then, the attacker sniffs and uses these credentials underscored intelligent vehicle manufacturers’ substantial
to access the genuine service. Multifactor authentication challenge in securing their products. As we progress
(MFA) security can also be circumvented by displaying a fake towards V2X (Vehicle-to-Everything) communication sys-
prompt that mimics the legitimate one, deceiving the victim tems, where vehicles exchange data with each other and with
into providing their one-time passcode (OTP). Time-Locked surrounding infrastructure, similar concerns for cybersecurity
One-Time Passwords [35] with lightweight browser extension exist. Analogous to the Jeep Cherokee case, where external
uses temporary passwords with fast expiry, thus making it access to a vehicle’s systems enabled malicious control, V2X-
harder to intercept credentials. connected cars face the challenge of safeguarding against
unauthorized access and potential remote attacks. Low-
C. HIJACKING THROUGH VEHICLE CHARGING Latency Anonymous Authentication Protocol using elliptic
In a connected IoAV network, malicious vehicles behave curve cryptography (ECC) [39] can ensure high security
like legitimate vehicles by selectively dropping their packets. so that only trusted networks can communicate with the
A session refers to the data exchange between the vehicles vehicle’s system.
for a selective period. Unauthorized access to the IoAV
ecosystem during a session is called a session hijacking E. HEADLIGHT HACKING THROUGH BLACKTOOTH
attack. The case study in [36] shows a use case where SPEAKER
the attacker can hijack the session wirelessly and prevent In a recent incident [40], a smart vehicle (Toyota RAV4) fell
people from charging their vehicles. This use case further victim to hacking and theft. A thorough analysis revealed
presents us with the possibility that these vehicles connected that the attackers employed an innovative approach called
through an IoAV system could be victims of masquerading, ‘‘headlight hacking’’ to penetrate the car’s interconnected
man-in-the-middle attacks, and denial-of-service through systems through its headlight. The attackers exploited an
session hijacking. Mutual Certificate-Based Authentication apparent tool akin to a JBL Bluetooth speaker, which has
[37] with a minimal handshake process between vehicles and been illicitly circulated on the dark web under the guise of
charging stations for authenticating each other with encrypted an emergency ignition device for compatible smart vehicles.
certificates can prevent hijacking at the charging sessions. When interfaced with the vehicle’s Controller Area Network
(CAN), this device effectively bypasses all security protocols,
D. CHARLIE MILLER AND CHRIS VALASEK INCIDENT granting direct entry to the vehicle’s functionalities, including
The well-known incident [38] involving professional hackers ignition, without necessitating the car’s key. Data gathered
Charlie Miller and Chris Valasek seizing control of a from the car’s telematics system indicate that the attackers
Jeep Cherokee in 2015 highlighted the genuine risk of accessed the CAN system by exploiting the Electronic
connected car hijacking, bringing this threat to the forefront Control Unit (ECU) within the headlight assembly. This
of public consciousness. When commencing their research, vulnerability within the 2021 Toyota RAV4, classified
the investigators initially attempted to breach the multimedia as CVE-2023-29389, expands beyond the headlight entry
system of the Jeep using a WiFi connection, which is provided method. It is one of several potential routes to access the car’s
as a subscription service by the vehicle manufacturer, CAN systems, enabling attackers to simulate the vehicle’s
Chrysler. They discovered that Chrysler’s vehicles were key fob, unlocking and driving away with the car. Although
generating their WiFi passwords before the actual setup of that incident shows an attack on the in-vehicle network
the time and date. These passwords were established using of the car (CAN) using a physical connection, the same
the default system time during the short span of seconds effect may occur using a remote connection if the attacker
when the head unit initialized. In practical terms, for instance, utilizes the V2X communication vulnerabilities, which will
January 01, 2013, 00:00 GMT translated to 00:00:32 GMT. be discussed in Section IV. Proximity-based authentication
The permutations for password combinations were minimal, with PUFs (Physical Unclonable Functions) [41] connects
making it relatively effortless for even amateur hackers to devices using unique physical traits of the device that
predict the appropriate password accurately. Their successful cannot be replicated, thus it secures access to the legitimate
attack resulted in the recall of 1.4 million vehicles and vehicle.

24212 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

IV. VEHICULAR COMMUNICATION PROTOCOLS AND the vehicle’s overall weight by at least 50kg. Despite the
THEIR SECURITY ISSUES vehicle’s systems and sensors being able to communicate
Vehicular protocols are essential for enabling communication at very high speeds (25Kbps–1Mbps), it lacked security
between vehicles and their various supporting infrastructures, measures to thwart attacks and abnormal communication
as mentioned in Section I. These protocols enable a conditions. As highlighted in [49], CAN communications
wide range of applications, from safety-critical systems to face the following security issues:
infotainment. However, extensive surveys, such as the one • Absence of Message Authentication for Nodes: Since
in [42], [43], [44], [45], and [46], point out a number of each Electronic Control Unit (ECU) transmits and
their challenges, including some related to security. Table 2 receives all data on the same bus, there is the problem of
summarizes some of the prominent protocols and their points of failure due to lack of message authentication.
security issues. • Lack of Encrypted Messages: The CAN bus was not
initially designed with cyber attack prevention in mind,
A. DEDICATED SHORT RANGE COMMUNICATION lacking encryption. This exposes information, enabling
Dedicated Short Range Communication (DSRC) is used for unauthorized access, potential data alteration, and even
wirelessly connecting vehicles in short ranges. It operates data replacement, leading to operational errors.
at a band frequency of 5.9GHz. This protocol is based • Vulnerability to DoS Attacks and Replay Attacks:
on the IEEE 802.11p standard that supports vehicular CAN bus uses a message-oriented protocol, where all
communication and is also supported by the European devices on the network receive all messages and it is not
Telecommunications Standards Institute (ETSI). This proto- localized, but each device decides whether a message
col defines both V2V and V2I communication, with a range is relevant based on its identifier. This design makes
typically up to 1000 meters. DSRC is primarily used to enable it susceptible to DoS attacks and replay attacks, where
applications such as collision detection and cruise control. an attacker could send high-priority frames to disrupt
Due to DSRC’s ability to connect with road infrastructure, critical system functions.
it can also reduce pollution and bottlenecks by relaying
information on congestion levels, road conditions, and traffic C. FLEXRAY
signal status. FlexRay, introduced in the year 2000 and jointly developed
Some of the key security issues of the DSRC are as follows: by Daimler Chrysler, BMW, Freescale, and Phillips, is a high-
• Lack of scalability: Scalability refers to the system’s bandwidth communication protocol designed for high-speed
ability to handle a growing workload or accommodate technology applications in automotive vehicles [49]. It was
expansion. When the number of vehicles in the IoAV introduced to address the limitations of the CAN protocol
connected through DSRC increases, the management of in terms of speed, determinism, and reliability. It supports
cryptographic keys in the dynamic environment poses a high data rates of up to 10Mbps and ensures determinism
major challenge. in communication through time-triggered events. It also
• Need for privacy: The privacy refers to the protection supports network node synchronization, which is vital for
of sensitive information about drivers, passengers, novel ADAS-enabled IoAV vehicles. Some of the security
and vehicles from unauthorized access, collection, limitations of FlexRay include:
or misuse during data exchange between vehicles. This • Lack of built-in security mechanism: FlexRay’s
involves safeguarding location data, driving patterns, absence of inherent security mechanisms makes it sus-
and personal identification to prevent tracking, profiling, ceptible to spoofing and denial-of-service (DoS) attacks.
or other privacy breaches. This short-range protocol The current features fail to ensure dependable commu-
is susceptible to attacks such as phishing and data nication in the presence of adversarial threats [50].
interception, which puts the privacy of the clients using • Lack of authentication mechanisms: FlexRay com-
IoAV at a potential risk. munication is very susceptible to attacks like message
• Communication vulnerabilities: The communication spoofing, where the attackers can inject counterfeit mes-
is vulnerable to jamming DoS attacks where the attacker sages, resulting in erroneous actions within the vehicle
might try to disturb key functionality by using, for exam- control systems. However, there are ongoing efforts to
ple, malicious interference on wireless communication enhance its authentication and key management capabil-
systems, e.g., coordination and information regarding ities. For instance, the authors in [51] have introduced an
traffic safety and efficiency. [47]. authentication mechanism in the optional dual-channel
mode to facilitate backward compatibility and have
B. CONTROLLER AREA NETWORK (CAN) BUS also proposed various techniques for cryptographic key
The CAN bus is the most commonly used bus in the management and authentication. Also, [52] presents
automotive industry [48]. It was first introduced in the protection schemes based on splitting authentication
BMW 850 in 1986. The motivation for this technology is to tags using independent channels to counter spoofing
reduce vehicle wiring by 2 km. Eventually, this also reduced attacks.

VOLUME 13, 2025 24213

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

TABLE 2. Vehicular protocols, their aspects and their security issues.

D. AUTOMOTIVE ETHERNET account several factors, such as computational overhead,

Automotive Ethernet (AE) emerged in 2013, drawing inspira- interoperability, and trade-offs involving security. They
tion from audio-video Bridging (AVB) within the automotive utilize techniques like streamlined cryptographic algorithms
sector. AVB itself is rooted in a series of IEEE standards that require fewer computational resources to execute without
governing Ethernet networks, dictating audio and video imposing a significant burden on the system’s CPU or
transmissions’ transport, signaling, and synchronization. memory. Moreover, these protocols should be able to work
In Ethernet networks, a primary security concern involves with smaller key sizes, contributing to lower memory usage
the potential compromise of administrators, as adherence to and bandwidth requirements. However, they have to be
security policies is crucial during new software installations, designed by keeping in mind the trade-offs in the security
configuration adjustments, and the deployment of switches. of the system resulting from a smaller key size. There has
These vulnerabilities open routes for attackers to assess also been research where the size of transmitted data packets
the defences of the Ethernet network, probing until they has been minimized to reduce the strain on computational
identify susceptible points. Beyond this, attackers may resources. In this section, we discuss some of the state-of-the-
exploit network access to gain insights into the network art lightweight authentication schemes that aid in securing
topology, understand traffic patterns for future attacks, IoAV systems.
and potentially seize control over switches and routers or
manipulate information [49].
Despite Automotive Ethernet bringing high-speed com-
A. PRE-SHARED KEY (PSK) AUTHENTICATION
munication, it also introduces key security challenges to the
availability and confidentiality of vehicle systems: Pre-shared key (PSK) authentication is a method in which
a single symmetric key is distributed to all parties involved
• Need for isolation: Automotive Ethernet connects in a network, allowing them to authenticate and encrypt
several critical systems, such as the ADAS and power communications. This method can be utilized in a vehicular
train controls. However, there are not enough seg- network to secure communication between access points
mentation and isolation guidelines for these systems. (APs) and a trusted authority.
Network segmentation techniques using virtual local In [55], is presented a lightweight symmetric cryptography-
area networks are used for the isolation of these critical based session key agreement scheme between each ECU
systems from other critical systems [53]. and the manufacturer data center that uses a random nonce,
• Lack of Interoperability: When different vehicles in concatenation operator, a simple hash function, and a keyed
real-time IoAV scenarios employ different communica- hash message authentication code. The security parameters
tion protocols and generate different data types, there for in-vehicle Ethernet-based communications are defined
is a need to interoperate with different entities [54]. between the different ECUs. Pre-shared keys are used in
Improper integration of automotive Ethernet can lead to this approach to secure the transmission between the internet
compatibility issues between different systems, resulting protocol messages.
in the loss of integrity. Mutual authentication techniques [56] have been used
to enhance the security levels in the IoAV system. In this
V. LIGHTWEIGHT AUTHENTICATION PROTOCOLS method, the session key is shared by the vehicles, and the
Authentication protocols that decrease the computational random nonce is decided by both vehicles transmitting the
strain on the system are needed to improve efficiency messages. Based on their results, this method was able to
and resource utilization without compromising security. mitigate unknown key share attacks, replay attacks, and
These lightweight protocols must be designed to take into key-compromised impersonation attacks.

24214 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

The work in [57] generates and distributes a new session tication scheme (NoMAS) proposed in [60] mitigates this
key between the vehicle virtual sensor and the selected leakage problem. The NoMAS scheme offers a solution for
physical sensor at every time interval. This ensures that key disclosure issues by providing Hard Key and Soft Key
access to sensor data is tightly controlled and restricted Updates. The steps of NoMAS include the initialization,
to authorized time periods. The session key is established vehicle registration, authentication, message verification,
using the ephemeral Elliptic Curve Diffie-Helman (ECDH) password update, and key update stages, where symmetric
protocol [58]. This is done at the end of the authentication keys are generated and shared across the connected vehicles.
of the selected physical sensor and the vehicle virtual The work in [61] proposes a symmetric key for broadcasting
sensor using the PSK. This perfect forward secrecy strategy basic safety messages in the same roadside unit group in
prevents the decryption of past or future transmissions with a secured fashion. This group key is updated whenever a
compromised keys. vehicle joins or leaves the network, making the system very
secure by not even allowing the member leaving to access
B. SYMMETRIC KEY AUTHENTICATION any current communication process. Through heuristic and
A symmetric key, also known as a private key, is a crypto- Burrows-Abadi-Needham (BAN) [62] logic analysis which
graphic key that is used for both encryption and decryption of is used to verify the veracity of the security protocols, the
data. In symmetric key cryptography, the same key is shared authors verify the efficacy of this method.
between communicating parties using methods such as PSK The work in [63], proposed a quantum-key-based authen-
for key distribution. Symmetric key authentication verifies tication and key agreement scheme that follows a role-based
the identity of both vehicles, sending them to communicating data access control Strategy. In this work, the vehicle
parties using symmetric key cryptography. In this method, all privatization manages quantum session keys and the vehicles
the vehicles involved share a secret key, known only to them. generate quantum session keys randomly. Secure agreement
When one vehicle wishes to authenticate itself to the other, of the keys amongst vehicles is done through multilevel
it generates a message or a token and encrypts it using the quantum keys such as the quantum prefilled key, the quantum
shared secret key. The receiving vehicle decrypts the message protection key, and the quantum session key. Symmetric
using the same key and verifies its authenticity based on the encryption algorithms and hash algorithms are used for
decrypted content. authentication in this scheme to obtain better performance.
In symmetric key authentication, the integrity of the data In [64], the authors present a quantum-resistant key man-
must be verified using a pairwise symmetric key between two agement scheme tailored for Cellular-Vehicle-to-Everything
vehicles. Methods like RSA-based session key distribution or (C-V2X) networks. Recognizing the vulnerabilities intro-
Diffie-Hellman key exchange can be used for symmetric key duced by quantum computing to traditional cryptographic
sharing. The method in [59] enables the pairwise symmetric methods, the authors integrate blockchain technology with
key to be calculated based on its identification of the lattice-based cryptography to establish a decentralized sys-
communication counterpart for a message m with a message tem. The scheme ensures secure registration, key-agreement,
authentication code, as shown in Figure 6. periodic updates, and revocation processes. The results show
a blockchain simulation validates its scalability under varying
vehicle densities and transaction loads. Also, it provides a
lightweight, efficient, and secure solution for key manage-
ment in resource-constrained vehicular networks, addressing
critical challenges such as single points of failure and the
ever-growing threat of quantum computing.

C. ASYMMETRIC KEY AUTHENTICATION

Asymmetric key authentication is a cryptographic process
that uses a pair of keys—a public key and a private key—to
verify the identity of users. Contrary to symmetric encryption,
asymmetric key authentication uses the public key to encrypt
data and the private key to decrypt it, ensuring that only the
FIGURE 6. Pairwise symmetric key tobe calculated based on its intended recipient with the correct private key can access the
identification of the communication counterpart. information.
In [65], the authors propose a robust and scalable
Some prominent key disclosure security attacks include authentication scheme for vehicular networks based on
side-channel and leakage attacks, as side-channel attacks Elliptic Curve Cryptography (ECC) to meet the diverse and
involve extracting sensitive information from the physical growing service demands of autonomous vehicles. In this
characteristics of a system, such as timing, power usage, approach, vehicles register with a trusted authority (TA) once,
or electromagnetic emissions. A novel MAC-based authen- enabling fast and efficient authentication with Cloud Service

VOLUME 13, 2025 24215

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

Providers (CSPs). Any new CSP registered with the TA can

participate in vehicular services seamlessly. A cloud broker,
managed by the TA, connects all cloud services, simplifying
user CSP selection. The security analysis shows that the
scheme provides conditional privacy protection and meets the
security needs of vehicular networks. Also, by utilizing ECC
without complex bilinear pairings, the scheme is efficient and
well-suited for vehicular network applications.
In [66], the authors propose a comprehensive security
framework for vehicular networks featuring two main
components: authentication and intrusion detection. The
authentication module uses ECC to mutually authenticate
the certificate authority, cluster heads, and vehicles, ensuring
secure communications. The Intrusion Detection System FIGURE 7. ECC is computationally less expensive and it is 35.88% faster
(IDS) employs tensor-based dimensional reduction to reduce than RSA while creating digital signatures [9].
vehicular traffic data before analyzing it with Fuzzy C-Means
(FCM) clustering combined with Multi-Objective Evolu-
tionary Algorithm based on Decomposition (MOEA/D). authentication. The vehicle receiving the challenge generates
However, the proposed framework uses a controller that can a response, and if it is acceptable, a secret key is generated
pose as a spoof. In addition, it does not protect against DoS by the challenge-issuing Vehicle (as illustrated in Figure 8).
or DDoS attacks, which are the biggest threat to availability. The response is measured using hash functions or symmetric
Implementing lightweight authentication procedures is encryption schemes. The connected vehicles verify their
challenging while exchanging vehicle data in a vehicular authenticity using the same cryptographic algorithm and the
cloud (VC). The study in [67] proposes a Dual Hashing-based shared secret key.
Secure Hashing Algorithm-384 (DHS384) and Modified The security scheme in [70] proposes a challenge-response
Elliptical Curve Cryptography (MECC). This method has authentication protocol that uses upper-layer authentication
an optimal resource allocation (RA) scheme to allocate to define the legitimacy of the corresponding terminal, and
resources optimally using the Improved Harris Hawks when verified, a location-dependent shared key is generated
Optimization (IHHO) algorithm. with minimal mismatched bits. Minimal mismatched bits
Researchers have found vulnerabilities in RSA public keys between two vehicular entities improve the key agreement
in recent years. For example, in 2020, in 75 million active thus, enhancing the integrity of the exchanged messages
RSA keys across the Internet, it was discovered that 1 in every critical for safe vehicular operations [71], [72], [73].
172 certificates using RSA keys is vulnerable to a practical Based on the generated key, a PHY challenge-response
attack known as ‘factoring’ [68]. In one of our recent works, algorithm for multi-carrier communication is generated for
we compare the algorithms ECC and RSA and compare the re-authentication. Simulation results of this proposed scheme
average time for digital signature formation. At the same show lower signal-to-noise ratios and are effective against
security levels, we were able to conclude that ECC is 35.88% active and passive attacks such as signatures forgery in the
faster than RSA [9] (Figure 7). In order to obtain the same random oracle model.
security level with a digital signature created by RSA with The Wiggle protocol is a challenge-response scheme
3072 bits, ECC requires a 256-bit key, which leads to a proposed in [74]. In this protocol, a challenge is issued to the
less computationally expensive operation. These variations candidate to be admitted into a platoon. This protocol is called
in the execution time have a dire impact on the underlying wiggle because of the random longitudinal movements that
IoAV system. Researchers have compared the performance the platoon is challenged to execute. This protocol prevents
of ECC and RSA in line with key generation, encryption, any attackers from joining the platoon and injecting fake
and decryption time in [69]. RSA requires exponential vehicle control messages. They also utilize an adaptive cruise
calculation for encryption and decryption, on the other hand, control (ACC) algorithm to execute the challenges and ensure
the ECC algorithm (encryption and decryption) is faster and the imperceptible changes to the vehicle’s velocity while a
consistently displays better encryption (34%) and decryption Proof of Following (PoF) is executed.
speeds (35%). The work in [75] proposes a lightweight and secure identity
authentication protocol based on elliptic curve cryptogra-
phy to furnish an effective and secure data transmission
D. CHALLENGE RESPONSE AUTHENTICATION mechanism across a public communication channel for the
Challenge-response [17] is a lightweight authentication Internet of Vehicles. The Physical Unclonable Function
method used for verifying the identity of vehicles or their (PUF) is a hardware-based security technology that relies
components by issuing a challenge to the connected vehicle. on the microscopic irregularities and physical characteristics
This challenge can be a random or specific request for of hardware devices to generate unique identifiers. PUF

24216 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

FIGURE 8. The malicious attacker prevents the platooning operation by intercepting the candidates’ requests and
replacing them with its own. The attacker masks itself as the verifier and issues a challenge.

generally functions using a challenge-response mechanism. pseudo-identity and private keys for V2V authentication, the
When presented with a challenge, the PUF generates a overall computational strain is reduced.
distinct response based on its physical traits. This response
serves as a unique identifier for the device. Due to the VI. CHALLENGES AND RESEARCH DIRECTIONS
inherent irregularities in the hardware manufacturing process, Designing a secure authentication protocol for the IoAV
even devices of the same model exhibit slight variations in involves multiple critical aspects to ensure the safety,
their physical characteristics. This uniqueness makes PUF integrity, and efficiency of vehicle communication. In the
responses unclonable, meaning identical hardware models following, we give an outline of these aspects and their
cannot generate the same response. As a result, the security implementation considerations.
of the system is enhanced.
A. SECURE AND STRONG HASH FUNCTIONS
One research direction for IoAV is investigating the suitability
E. LOW-RESOURCE AUTHENTICATION of existing hash functions for developing a robust, lightweight
With minimal computational resources in hand, the confi- authentication protocol. On the other hand, it is vital
dentiality and privacy of data transmitted between vehicles, to tailor unique hash functions for IoAV applications to
roadside units (RSU), and control rooms is a critical issue provide sufficient defense against differential and linear
that must be addressed effectively. Blockchain technology cryptanalysis. This line of research can explore modifications
presents a potential solution for securing the IOVs because to the substitution box (S-box), a critical element in many
of its decentralization, stability, and transaction tracking symmetric cryptographic primitives, and assess the security
capabilities. To overcome these issues, [76] introduces a implications of the enhanced S-box to provide sufficient
blockchain-based lightweight authentication protocol for defense against differential and linear cryptanalysis. Addi-
generating trustworthy IOV communication. tionally, it is crucial to explore cost-effective hardware
implementation possibilities for deploying lightweight hash
The work in [77] introduces an efficient and safe identity
functions on resource-constrained devices with low power
authentication scheme based on the Feige-Fiat-Shamir (FFS)
consumption [80].
zero-knowledge identification scheme [78] to resist guessing
A solid cryptographic hash function must satisfy three
attacks. Unlike cryptographic schemes that require complex
computations such as elliptic curves and large integer criteria [81]:
factorizations, the FFS scheme uses simpler number theory • Preimage resistance: Retrieving the original message
operations, thus effectively reducing the computational when the hash value is given should be challenging.
overhead. • Second Preimage Resistance (Weak Collision Resis-
A Certificate Authority (CA) is necessary for vehicles tance): It should be challenging to find a different
within communication range to establish authentication. message with the same hash value as a given message.
However, as the vehicles move and the distance between • Strong Collision Resistance: Finding two messages with
them constantly varies, it is a challenge to make this the same hash value should be challenging.
happen. The work in [79] proposes a lightweight security A hash function takes an input of a fixed-size n-bit string
protocol for authentication for V2X. This scheme employs and produces an m-bit string where m < n. The original
a biometric device (BD) and a tamper-proof device (TPD), hash function breaks the message into fixed-size blocks
which together verify the driver and securely keep the and produces a final message digest. These message digests
keys. By decentralizing the CA’s tasks by locally generating can be of several types, such as MD2, MD4, and MD5,

VOLUME 13, 2025 24217

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

which are used primarily for ensuring data integrity and D. SYSTEM DESIGN
verifying the authenticity of data [82], [83]. The Secure Hash In the realm of embedded IoT edge systems, including
Algorithm (SHA) was developed by the National Institute of IoAV, computational resources are constrained, and tight
Standards and Technology (NIST). Examples of the Secure timing constraints require a tradeoff between security and
Hash Algorithm include SHA-1, SHA-2, and SHA-3, with performance. Encompassing the entire system by modelling
SHA-3 producing a message digest of 512 bits [81]. and analyzing end-to-end systems (hardware and software) is
preferred over solely focusing on modeling security protocols
B. BLOCKCHAIN SOLUTIONS like in [80], [85], and [86]. This is because it provides a
Blockchain technology offers promising solutions for infor- holistic view of the system’s security. This approach captures
mation security with its non-tampering, traceability, and the complex interactions between components, identifies
collective maintenance features. It addresses challenges such vulnerabilities that may arise from integration issues, and
as broadcast collision avoidance, resource scheduling, and addresses systemic risks beyond protocol-specific flaws.
privacy preservation in the IoAV context. It ensures that security measures are effectively balanced
Blockchain is deployed in a decentralized fashion between with performance, usability, and compliance requirements,
entities such as vehicles, roadside units (RSUs), and other enhancing overall resilience. By considering the full system,
entities that act as nodes in the network. In line with IoAV’s the possibility of developing accurate threat models, prior-
resource constraints and critical real-time requirements, itizing risks, and future-proofing systems against emerging
a permissioned blockchain (e.g., Hyperledger Fabric [84]) is threats increases, resulting in a more robust and comprehen-
utilized. Permissioned blockchain restricts and can validate sive security posture.
transactions, thus reducing computational overhead. There exist a number of models of computation (MoC)
Cryptographic-based authentication methods such as and frameworks especially adapted for modeling, designing,
Public Key Infrastructure (PKI), group signature, and and analyzing embedded systems. They usually focus on
identity-based schemes are adopted for IoAV. The PKI-based exposing parallelism, while providing tools for analyzing
scheme aims to guarantee data integrity and privacy through timing, memory use and communication bandwidth.
PKI signatures. Nevertheless, due to the significant con- For instance, the dataflow MoC is regarded as a natural way
sumption of computing and communication resources, there of modeling stream processing in embedded applications,
is a need to enhance the efficiency of PKI-based authenti- targeting both hardware and software implementations.
cation schemes. One method involves integrating PKI into For the automotive domain, [89] exposes a number of
decentralized blockchain systems to improve the security advantages of the dataflow MoC by introducing an approach
and integrity of the underlying networks. [85]. However, transforming a Synchronous Dataflow (SDF) model into
it is crucial to carefully assess PKI-based and Blockchain an automotive component. Some of these advantages come
schemes for their suitability in high vehicle density scenarios, from the static analysis techniques, able to provide accurate
as they require substantial computing resources [86]. information on throughput, latency, buffer sizes [90] and real-
time behavior [91], [92]. Such insights help the designer to
understand the system better and make informed decisions
C. MINIMIZATION OF OVERHEAD COMMUNICATION about the system architecture (hardware and software) and
IoAV relies on low-latency and highly reliable communi- behavior. Toolchains based on dataflow MoCs, such as CAL/
cation. With heavy authentication and strong cryptographic StreamBlocks [93], are becoming more mature and target
techniques, there can be large overheads in IoAV systems. both software and hardware implementations, starting with
For instance, in one of our previous works, by integrating a the same specification.
lightweight hybrid cryptography scheme called TAKS [87] However, specific challenges persist: (i) targeted applica-
in a deterministic network protocol for intra-vehicle com- tions are distributed, spanning vehicles and road infrastruc-
munication [88], we were able to witness an increase in ture rather than bound to a single computing platform [94].
delay between 20% and 60% based on the size of the This complexity requires intricate analyses and may call for
keys of the cryptographic protocol. Also, the parameters unique communication channel models. Modern frameworks
governing the MAC layer of the protocol played a major that are intended for distributed embedded applications,
role in determining the latency of the network and the such as LinguaFranca [95], are still in their infancy and
incorporation of security into it. In one of our recent studies target mainly software implementations. In software-defined
comparing various authentication schemes for cloud-enabled vehicles, optimizing dataflow and resource management is
platooning [9], we arrived at the same conclusion: the mandatory to achieve short response times, high throughput
authentication schemes, despite providing ample security, and predictability, which indirectly helps in ensuring the
can also lead to large overheads and eventual breakdown of network’s resilience under fluctuating conditions. Newer
the system. With these studies in mind, it is vital to define methods such as software-defined frameworks [96] refine
the tuning of several factors, such as network protocols, the resource allocation and optimal resource distributions.
authentication schemes, and levels of security, for enabling Also, (ii) classic actor models (e.g., dataflow) inherently
good security in IoAV networks. lack security considerations. For instance, incorporating
24218 VOLUME 13, 2025
H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

security layers into communication requires rethinking the performance, we should model the larger parts of the system
framework and MoC components or adopting dedicated using dataflow-based frameworks that are able to model both
solutions specific to each application. Methods such as hardware and software. Overcoming challenges such as the
Concurrent Task Analysis (CoTA), System-Theoretic Process distributed nature of IoAV and the lack of inherent security in
Analysis (STPA), and Fault Tree Analysis (FTA) have traditional modeling frameworks is essential for an accurate
been used in the literature to mitigate operational safety and realistic analysis of security protocols in these networks.
hazards [97] but they do not incorporate security layers
dedicated to hardware/software level. Solving the above REFERENCES
challenges will allow more accurate analyses factoring in [1] P. A. Hancock, I. Nourbakhsh, and J. Stewart, ‘‘On the future of
the overhead caused by security protocols and message transportation in an era of automated and autonomous vehicles,’’
Proc. Nat. Acad. Sci. USA, vol. 116, no. 16, pp. 7684–7691, Apr. 2019.
encryption/decryption. In turn, these will yield more accurate [2] M. L. Cunningham and M. Regan, ‘‘Autonomous vehicles: Human factors
estimates within realistic scenarios, which is crucial for issues and future research,’’ in Proc. Australas. Road Saf. Conf., vol. 14,
illustrating how security measures impact timing and system Oct. 2015, pp. 1–11.
[3] A. Nanda, D. Puthal, J. J. P. C. Rodrigues, and S. A. Kozlov, ‘‘Internet
resources. of Autonomous Vehicles communications security: Overview, issues, and
directions,’’ IEEE Wireless Commun., vol. 26, no. 4, pp. 60–65, Aug. 2019.
VII. CONCLUSION [4] H. Kurunathan, J. Santos, D. Moreira, and P. M. Santos, ‘‘Towards
safe cooperative autonomous platoon systems using COTS equipment,’’
The Internet of Things (IoT) is a network of physical objects in Proc. IEEE 24th Int. Symp. World Wireless, Mobile Multimedia
that are equipped with sensors, software, and other technolo- Netw. (WoWMoM), Jun. 2023, pp. 464–469.
gies to connect and exchange data over the Internet. Over [5] J. Pereira, H. Kurunathan, E. Filho, and P. M. Santos, ‘‘Edge-aided V2X
collision avoidance with platoons: Towards a hybrid evaluation toolset,’’ in
the past decade, its rapid expansion has impacted various Proc. 53rd Annu. IEEE/IFIP Int. Conf. Dependable Syst. Netw. Workshops
areas, including intelligent vehicular systems. The Internet (DSN-W), Jun. 2023, pp. 68–71.
of Autonomous Vehicles (IoAV), a subset of IoT, aims [6] R. Reddy, L. Almeida, H. Kurunathan, M. G. Gaitán, P. M. Santos
and E. Tovar, ‘‘Worst-case response time of mixed vehicles at complex
to enable autonomous driving without human intervention.
intersections,’’ IEEE Open J. Intell. Transp. Syst., vol. 5, pp. 186–201,
However, the dynamic environments in which autonomous 2024, doi: 10.1109/OJITS.2024.3368797.
vehicles operate present significant challenges, such as [7] R. Reddy, L. Almeida, M. Gaitan, H. Kurunathan, P. Santos, and E. Tovar,
‘‘Work-in-progress: Worst-case response time of intersection management
real-time communication and security vulnerabilities. These
protocols,’’ in Proc. IEEE Real-Time Syst. Symp. (RTSS), Dec. 2021,
challenges cannot be effectively addressed by traditional pp. 556–559.
cybersecurity solutions designed for static IoT environments. [8] M. Araújo, J. Silva, P. M. Santos, H. Singh, D. Gunjal, J. Fonseca,
This survey article presents an overview of the current P. Duarte, B. Mendes, R. Barbosa, P. Steenkiste, S. Sabamoniri, L. Lam,
J. Pereira, and H. Kurunathan, ‘‘Demo: Object detection under 5G-
security issues related to the Internet of Autonomous Vehicles edge mobility,’’ in Proc. IEEE 24th Int. Symp. World Wireless, Mobile
(IoAV) and their respective solutions. The survey highlights Multimedia Netw. (WoWMoM), Jun. 2023, pp. 343–345.
several crucial challenges when developing a lightweight [9] R. Rafael and E. T. Harrison Kurunathan, ‘‘An RT-cloud solution towards
security in vehicular platooning systems,’’ in Proc. 35th Euromicro
authentication protocol within an IoAV ecosystem. It empha- Conf. Real-Time Syst. Real-Time Cloud Workshop, vol. 2, 2023, pp. 1–6.
sizes the necessity of exploring cost-effective hardware [10] S. Vasavi, K. Aswarth, T. Sai Durga Pavan, and A. Anu Gokhale,
implementations for deploying lightweight hash functions ‘‘Predictive analytics as a service for vehicle health monitoring using
edge computing and AK-NN algorithm,’’ Mater. Today, Proc., vol. 46,
and evaluating PKI-based blockchain schemes for their pp. 8645–8654, Aug. 2021.
suitability in high vehicle density scenarios. Additionally, [11] U. Shafi, A. Safi, A. R. Shahid, S. Ziauddin, and M. Q. Saleem,
it emphasizes the importance of defining network protocol ‘‘Vehicle remote health monitoring and prognostic maintenance system,’’
J. Adv. Transp., vol. 2018, no. 1, pp. 1–10, 2018.
parameters, system designs, and authentication schemes to [12] E. Alalwany and I. Mahgoub, ‘‘Security and trust management in the
enable secure IoAV systems without making compromises. Internet of Vehicles (IoV): Challenges and machine learning solutions,’’
Last but not least, enhancing the security and trustworthiness Sensors, vol. 24, no. 2, p. 368, Jan. 2024.
[13] N. U. A. Tahir, Z. Zhang, M. Asim, J. Chen, and M. Elaffendi, ‘‘Object
of an IoAV ecosystem through lightweight protocols deserves detection in autonomous vehicles under adverse weather: A review of
significant attention. traditional and deep learning approaches,’’ Algorithms, vol. 17, no. 3,
In the future, creating a secure and efficient authentication p. 103, Feb. 2024.
[14] P. Zhao, Y. Huang, J. Gao, L. Xing, H. Wu, and H. Ma, ‘‘Federated
protocol for the Internet of Autonomous Vehicles (IoAV) is learning-based collaborative authentication protocol for shared data in
critical. This includes finding a balance between security social IoV,’’ IEEE Sensors J., vol. 22, no. 7, pp. 7385–7398, Apr. 2022.
and performance. Researchers must develop specific hash [15] H. Hemavathi, S. R. Akhila, Y. Alotaibi, O. I. Khalaf, and S. Alghamdi,
‘‘Authentication and resource allocation strategies during handoff for 5G
functions to protect against cryptanalysis and integrate Public
IoVs using deep learning,’’ Energies, vol. 15, no. 6, p. 2006, Mar. 2022.
Key Infrastructure (PKI) within decentralized blockchain [16] D. Zhang, F. R. Yu, R. Yang, and L. Zhu, ‘‘Software-defined vehicular net-
systems to achieve this. Since PKI and blockchain can be works with trust management: A deep reinforcement learning approach,’’
computationally demanding in environments with a high IEEE Trans. Intell. Transp. Syst., vol. 23, no. 2, pp. 1400–1414, Feb. 2022.
[17] M. A. Shawky, S. T. Shah, M. Abdrabou, M. Usman, Q. H. Abbasi,
density of vehicles, it is crucial to assess their performance D. Flynn, M. A. Imran, S. Ansari, and A. Taha, ‘‘How secure are our roads?
carefully to prevent system delays. Additionally, it is essential An in-depth review of authentication in vehicular communications,’’
to optimize network protocols and authentication schemes Veh. Commun., vol. 47, Jun. 2024, Art. no. 100784.
[18] I. Pali, R. Amin, and M. Abdussami, ‘‘Autonomous vehicle security:
to maintain low latency in IoAV communication. To fully Current survey and future research challenges,’’ Secur. PRIVACY, vol. 7,
understand how security measures affect overall system no. 3, p. 367, May 2024.

VOLUME 13, 2025 24219

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

[19] T. K. Mohd, S. Majumdar, A. Mathur, and A. Y. Javaid, ‘‘Simulation [41] J. Patel, M. L. Das, and S. Nandi, ‘‘On the security of remote key
and analysis of DDoS attack on connected autonomous vehicular network less entry for vehicles,’’ in Proc. IEEE Int. Conf. Adv. Netw. Telecom-
using OMNET++,’’ in Proc. 9th IEEE Annu. Ubiquitous Comput., mun. Syst. (ANTS), Dec. 2018, pp. 1–6.
Electron. Mobile Commun. Conf. (UEMCON), Nov. 2018, pp. 502–508. [42] N. H. Hussein, C. T. Yaw, S. P. Koh, S. K. Tiong, and K. H. Chong,
[20] A. Palanca, E. Evenchick, F. Maggi, and S. Zanero, ‘‘A stealth, ‘‘A comprehensive survey on vehicular networking: Communications,
selective, link-layer denial-of-service attack against automotive networks,’’ applications, challenges, and upcoming research directions,’’ IEEE Access,
in Proc. 14th Int. Conf. Detection Intrusions Malware, Vulnerability vol. 10, pp. 86127–86180, 2022.
Assessment (DIMVA), Bonn, Germany. Cham, Switzerland: Springer, [43] P. Bagga, A. Kumar Das, M. Wazid, J. J. P. C. Rodrigues, and Y. Park,
Jan. 2017, pp. 185–206. ‘‘Authentication protocols in Internet of Vehicles: Taxonomy, analysis, and
[21] F. R. Schreiber, Sybil: The True Story of a Woman Possessed by Sixteen challenges,’’ IEEE Access, vol. 8, pp. 54314–54344, 2020.
Separate Personalities. Chicago, IL, USA: Regnery, 1973. [44] T. Gebremichael, L. P. I. Ledwaba, M. H. Eldefrawy, G. P. Hancke,
[22] K. Rabieh, M. M. E. A. Mahmoud, T. N. Guo, and M. Younis, ‘‘Cross- N. Pereira, M. Gidlund, and J. Akerberg, ‘‘Security and privacy in the
layer scheme for detecting large-scale colluding Sybil attack in VANETs,’’ industrial Internet of Things: Current standards and future challenges,’’
in Proc. IEEE Int. Conf. Commun. (ICC), Jun. 2015, pp. 7298–7303. IEEE Access, vol. 8, pp. 152351–152366, 2020.
[23] D. Puthal, S. P. Mohanty, S. A. Bhavake, G. Morgan, and R. Ranjan, [45] M. Serror, S. Hack, M. Henze, M. Schuba, and K. Wehrle,
‘‘Fog computing security challenges and future directions [energy and ‘‘Challenges and opportunities in securing the industrial Internet of
security],’’ IEEE Consum. Electron. Mag., vol. 8, no. 3, pp. 92–96, Things,’’ IEEE Trans. Ind. Informat., vol. 17, no. 5, pp. 2985–2996,
May 2019. May 2021.
[24] M. Eldefrawy, I. Butun, N. Pereira, and M. Gidlund, ‘‘Formal security [46] S. B. ElMamy, H. Mrabet, H. Gharbi, A. Jemai, and D. Trentesaux,
analysis of LoRaWAN,’’ Comput. Netw., vol. 148, pp. 328–339, Jan. 2019. ‘‘A survey on the usage of blockchain technology for cyber-threats in
[25] L. Pankaczi and M. Eldefrawy, ‘‘Enhancing the security of ISO/IEC the context of Industry 4.0,’’ Sustainability, vol. 12, no. 21, p. 9179,
14443-3 and 4 RFID authentication protocols through formal analysis,’’ in Nov. 2020.
Proc. IEEE Int. Conf. Omni-layer Intell. Syst. (COINS), Jul. 2023, pp. 1–6. [47] N. Lyamin, A. Vinel, M. Jonsson, and J. Loo, ‘‘Real-time detection of
[26] X. Sun, F. R. Yu, and P. Zhang, ‘‘A survey on cyber-security of connected denial-of-service attacks in IEEE 802.11p vehicular networks,’’ IEEE
and autonomous vehicles (CAVs),’’ IEEE Trans. Intell. Transp. Syst., Commun. Lett., vol. 18, no. 1, pp. 110–113, Jan. 2014.
vol. 23, no. 7, pp. 6240–6259, Jul. 2022. [48] S. C. Hpl, ‘‘Introduction to the controller area network (CAN),’’
[27] S. Sharma and B. Kaushik, ‘‘A survey on Internet of Vehicles: Applications, Appl. Rep. SLOA101, Aug. 2002, pp. 1–17.
security issues & solutions,’’ Veh. Commun., vol. 20, Dec. 2019, [49] A. Martínez-Cruz, K. A. Ramírez-Gutiérrez, C. Feregrino-Uribe, and
Art. no. 100182. A. Morales-Reyes, ‘‘Security on in-vehicle communication protocols:
[28] S. Yu, J. Lee, K. Park, A. K. Das, and Y. Park, ‘‘IoV-SMAP: Secure Issues, challenges, and future research directions,’’ Comput. Commun.,
and efficient message authentication protocol for IoV in smart city vol. 180, pp. 1–20, Dec. 2021.
environment,’’ IEEE Access, vol. 8, pp. 167875–167886, 2020. [50] P.-S. Murvay and B. Groza, ‘‘Efficient physical layer key agreement
[29] A. Ghosal, S. Halder, and M. Conti, ‘‘STRIDE: Scalable and secure over- for FlexRay networks,’’ IEEE Trans. Veh. Technol., vol. 69, no. 9,
the-air software update scheme for autonomous vehicles,’’ in Proc. IEEE pp. 9767–9780, Sep. 2020.
Int. Conf. Commun. (ICC), Jun. 2020, pp. 1–6. [51] D. Püllen, N. A. Anagnostopoulos, T. Arul, and S. Katzenbeisser,
[30] T. Roccia, ‘‘Today’s connected cars vulnerable to hacking, ‘‘Security and safety co-engineering of the FlexRay bus in vehicular
malware,’’ Mcafee, Tech. Rep., Mar. 2018. [Online]. Available: networks,’’ in Proc. Int. Conf. Omni-Layer Intell. Syst., May 2019,
https://www.mcafee.com/blogs/other-blogs/mcafee-labs/todays- pp. 31–37.
connected-cars-vulnerable-hacking-malware/ [52] D. Püllen, N. A. Anagnostopoulos, T. Arul, and S. Katzenbeisser, ‘‘Secur-
[31] S. Rizvi, J. Willet, D. Perino, S. Marasco, and C. Condo, ‘‘A threat to vehic- ing FlexRay-based in-vehicle networks,’’ Microprocessors Microsyst.,
ular cyber security and the urgency for correction,’’ Proc. Comput. Sci., vol. 77, Sep. 2020, Art. no. 103144.
vol. 114, pp. 100–105, Aug. 2017. [53] P. Hank, S. Müller, O. Vermesan, and J. Van Den Keybus, ‘‘Automotive
[32] G. Avoine, M. A. Bingöl, S. Kardaş, C. Lauradoux, and B. Martin, Ethernet: In-vehicle networking and smart mobility,’’ in Proc. Design,
‘‘A framework for analyzing RFID distance bounding protocols,’’ J. Com- Autom. Test Eur. Conf. Exhib. (DATE), Mar. 2013, pp. 1735–1739.
put. Secur., vol. 19, no. 2, pp. 289–317, Mar. 2011. [54] A. Ioana, A. Korodi, and I. Silea, ‘‘Automotive IoT Ethernet-based
[33] D. Jacobs, K. R. Choo, M.-T. Kechadi, and N.-A. Le-Khac, ‘‘Volk- communication technologies applied in a V2X context via a multi-protocol
swagen car entertainment system forensics,’’ in Proc. IEEE Trust- gateway,’’ Sensors, vol. 22, no. 17, p. 6382, Aug. 2022.
com/BigDataSE/ICESS, Aug. 2017, pp. 699–705. [55] H. Khemissa and P. Urien, ‘‘Towards a centralized security architecture
[34] How a Credential Phishing Attack Could Lead to Tesla Car Theft for SOME/IP automotive services,’’ in Proc. IEEE 20th Consum. Com-
and How to Mitigate It. Accessed: May 3, 2024. [Online]. Available: mun. Netw. Conf. (CCNC), Sep. 2023, pp. 977–978.
https://vicone.com/blog/how-a-credential-phishing-attack-could-lead-to- [56] P. M. Rao, ‘‘2PSLAF: Privacy-preserving secure lightweight authenti-
tesla-car-theft-and-how-to-mitigate-it cation framework for vehicle-to-vehicle communications,’’ Res. Square,
[35] M. Hataba, A. Sherif, M. Mahmoud, M. Abdallah, and W. Alasmary, Jan. 2024, doi: 10.21203/rs.3.rs-1298797/v3.
‘‘Security and privacy issues in autonomous vehicles: A layer-based [57] S. Abbes and S. Rekhis, ‘‘Reinforcement learning for intelligent sensor
survey,’’ IEEE Open J. Commun. Soc., vol. 3, pp. 811–829, 2022. virtualization and provisioning in Internet of Vehicles (IoV),’’ IEEE
[36] G. Vailoces, A. Keith, A. Almehmadi, and K. El-Khatib, ‘‘Securing Access, vol. 12, pp. 54352–54370, 2024.
the electric vehicle charging infrastructure: An in-depth analysis of [58] J. Liu, S. Xu, R. Liu, Y. Tian, X. Zhu, and J. Xie, ‘‘LIA-V2X: A lightweight
vulnerabilities and countermeasures,’’ in Proc. Int. ACM Symp. Design identity mutual and batch authentication protocol for V2X communication
Anal. Intell. Veh. Netw. Appl., Oct. 2023, pp. 31–38. in Internet of Vehicles,’’ 2024.
[37] B. Vaidya and H. T. Mouftah, ‘‘Multimodal and multi-pass authentication [59] J. Yoo and J. H. Yi, ‘‘Code-based authentication scheme for
mechanisms for electric vehicle charging networks,’’ in Proc. Int. Wireless lightweight integrity checking of smart vehicles,’’ IEEE Access, vol. 6,
Commun. Mobile Comput. (IWCMC), Jun. 2020, pp. 371–376. pp. 46731–46741, 2018.
[38] A. Drozhzhin, ‘‘Black hat USA 2015: The full story of how that [60] H. Sikarwar and D. Das, ‘‘A novel MAC-based authentication scheme
jeep was hacked,’’ Kaspersky Lab Daily, vol. 6, Aug. 2015. [Online]. (NoMAS) for Internet of Vehicles (IoV),’’ IEEE Trans. Intell. Transp. Syst.,
Available: https://www.kaspersky.com/blog/blackhat-jeep-cherokee-hack- vol. 24, no. 5, pp. 4904–4916, May 2023.
explained/9493/ [61] Z. Zhou, X. Liu, C. Wang, and R. Lu, ‘‘Efficient and secure authentication
[39] Z. G. Al-Mekhlafi, H. D. K. Al-Janabi, A. Khalil, M. A. Al-Shareeda, scheme for Internet of Vehicles,’’ in Proc. Int. Conf. Intell. Inf. Process.
B. A. Mohammed, A. A. Alsadhan, A. M. Alayba, A. M. S. Saleh, Cham, Switzerland: Springer, Jan. 2024, pp. 469–485.
H. A. Al-Reshidi, and K. Almekhlafi, ‘‘Lattice-based cryptography [62] M. T. Burrows, M. Abadi, and R. M. Needham, ‘‘A logic of
and fog computing based efficient anonymous authentication scheme authentication,’’ ACM Trans. Comput. Syst., vol. 8, no. 1, pp. 18–36,
for 5G-assisted vehicular communications,’’ IEEE Access, vol. 12, Feb. 1990.
pp. 71232–71247, 2024. [63] Q. Shi, Z. Yang, T. Cheng, C. Wang, Z. Wu, X. Zhang, and P. Xu,
[40] Can Injection Attack—A New Tactic to Steal Smart Vehicles. Accessed: ‘‘QKBAKA: A quantum-key-based authentication and key agreement
Aug. 14, 2023. [Online]. Available: https://cyware.com/news/can- scheme for Internet of Vehicles,’’ IEEE Internet Things J., vol. 11, no. 7,
injection-attack-a-new-tactic-to-steal-smart-vehicles-1b1529f2 pp. 12292–12306, Jul. 2023.

24220 VOLUME 13, 2025

H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

[64] Y. Gong and B.-J. Hu, ‘‘A quantum-resistant key management scheme [85] H. Zhang and F. Zhao, ‘‘Cross-domain identity authentication scheme
using blockchain in C-V2X,’’ IEEE Trans. Intell. Transp. Syst., vol. 25, based on blockchain and PKI system,’’ High-Confidence Comput., vol. 3,
no. 11, pp. 16831–16844, Nov. 2024. no. 1, Mar. 2023, Art. no. 100096.
[65] J. Cui, X. Zhang, H. Zhong, J. Zhang, and L. Liu, ‘‘Extensible conditional [86] Z. Xu, X. Li, J. Xu, W. Liang, and K.-K.-R. Choo, ‘‘A secure and
privacy protection authentication scheme for secure vehicular networks in computationally efficient authentication and key agreement scheme
a multi-cloud environment,’’ IEEE Trans. Inf. Forensics Security, vol. 15, for Internet of Vehicles,’’ Comput. Electr. Eng., vol. 95, Oct. 2021,
pp. 1654–1667, 2020. Art. no. 107409.
[66] S. Garg, K. Kaur, G. Kaddoum, S. H. Ahmed, and D. N. K. Jayakody, [87] W. Tiberti, B. Vieira, H. Kurunathan, R. Severino, and E. Tovar,
‘‘SDN-based secure and privacy-preserving scheme for vehicular net- ‘‘Tightening up security in low power deterministic networks,’’ in
works: A 5G perspective,’’ IEEE Trans. Veh. Technol., vol. 68, no. 9, Proc. 16th IEEE Int. Conf. Factory Commun. Syst. (WFCS), Apr. 2020,
pp. 8421–8434, Sep. 2019. pp. 1–7.
[67] J. Varalakshmi and S. Dhanasekaran, ‘‘A dual hashing-based authentication [88] H. Kurunathan, R. Severino, Ê. Filho, and E. Tovar, ‘‘WiCAR—Simulating
and secure data transmission scheme for vehicular cloud environment towards the wireless car,’’ in Proc. Comput. Saf., Rel., Secur. SAFECOMP
using MECC with optimal resource allocation mechanism,’’ Soft Comput., Workshops, Lisbon, Portugal. Cham, Switzerland: Springer, Sep. 15, 2020,
vol. 28, nos. 17–18, pp. 10423–10437, Sep. 2024. pp. 136–147.
[68] J. Kilgallin and R. Vasko, ‘‘Factoring RSA keys in the IoT era,’’ in [89] M. O. Aybek, R. Jordão, J. Lundbäck, K.-L. Lundbäck, and
Proc. 1st IEEE Int. Conf. Trust, Privacy Secur. Intell. Syst. Appl. (TPS- M. Becker, ‘‘From the synchronous data flow model of computation
ISA), Dec. 2019, pp. 184–189. to an automotive component model,’’ in Proc. 26th IEEE
[69] M. R. Khan, K. Upreti, M. I. Alam, H. Khan, S. T. Siddiqui, M. Haque, Int. Conf. Emerg. Technol. Factory Autom. (ETFA), Sep. 2021, pp. 1–8,
and J. Parashar, ‘‘Analysis of elliptic curve cryptography & RSA,’’ J. ICT doi: 10.1109/ETFA45728.2021.9613621.
Standardization, vol. 11, no. 4, pp. 355–378, Nov. 2023. [90] M. Boulasikis, F. Gruian, G. Callanan, and J. W. Janneck, ‘‘Analysing
[70] M. A. Shawky, M. Bottarelli, G. Epiphaniou, and P. Karadimas, dataflow programs with causation traces,’’ in Proc. Int. Conf. Parallel
‘‘An efficient cross-layer authentication scheme for secure communication Architectures Compilation Techn., Oct. 2022, pp. 534–535.
in vehicular ad-hoc networks,’’ IEEE Trans. Veh. Technol., vol. 72, no. 7, [91] H. I. Ali, B. Akesson, and L. M. Pinho, ‘‘Generalized extraction of
pp. 8738–8754, Jul. 2023. real-time parameters for homogeneous synchronous dataflow graphs,’’ in
[71] H. Kurunathan, K. Li, W. Ni, N. Li, E. Tovar, and M. Guizani, ‘‘Exploring Proc. 23rd Euromicro Int. Conf. Parallel, Distrib., Netw.-Based Process.,
LSTM-assisted A2C for physical layer security in vehicular cyber-physical Mar. 2015, pp. 701–710.
systems,’’ in Proc. Int. Wireless Commun. Mobile Comput. (IWCMC), [92] H. I. Ali, B. Akesson, and L. M. Pinho, ‘‘Combining dataflow applications
May 2024, pp. 1177–1182. and real-time task sets on multi-core platforms,’’ in Proc. 20th Int. Work-
[72] H. Kurunathan, K. Li, E. Tovar, A. Mario Jorge, W. Ni, and shop Softw. Compil. Embedded Syst., Jun. 2017, pp. 60–63.
A. Jamalipour, ‘‘DRL-KeyAgree: An intelligent combinatorial deep [93] E. Bezati, M. Emami, J. Janneck, and J. Larus, ‘‘StreamBlocks: A
reinforcement learning-based vehicular platooning secret key generation,’’ compiler for heterogeneous dataflow computing (technical report),’’ 2021,
IEEE Trans. Intell. Transp. Syst., vol. 25, no. 11, pp. 16354–16369, arXiv:2107.09333.
Nov. 2024. [94] G. Callanan and F. Gruian, ‘‘Estimating stream application performance
[73] K. Li, H. Kurunathan, R. Severino, and E. Tovar, ‘‘Cooperative in early-stage system design,’’ in Proc. 56th Asilomar Conf. Signals, Syst.,
key generation for data dissemination in cyber-physical systems,’’ in Comput., Oct. 2022, pp. 816–823.
Proc. ACM/IEEE 9th Int. Conf. Cyber-Phys. Syst. (ICCPS), Apr. 2018, [95] M. Lohstroh, C. Menard, S. Bateni, and E. A. Lee, ‘‘Toward a lingua
pp. 331–332. franca for deterministic concurrent systems,’’ ACM Trans. Embedded
[74] C. Dickey, C. Smith, Q. Johnson, J. Li, Z. Xu, L. Lazos, and M. Li, Comput. Syst., vol. 20, no. 4, pp. 1–27, May 2021, doi: 10.1145/3448128.
‘‘Wiggle: Physical challenge-response verification of vehicle platooning,’’ [96] A. Nahar, K. K. Mondal, D. Das, and R. Buyya, ‘‘Clouds on the road:
in Proc. Int. Conf. Comput., Netw. Commun. (ICNC), Feb. 2023, A software-defined fog computing framework for intelligent resource
pp. 54–60. management in vehicular ad-hoc networks,’’ IEEE Trans. Mobile Comput.,
[75] S. Wang, Z. Fan, Y. Su, B. Zheng, Z. Liu, and Y. Dai, ‘‘A lightweight, vol. 23, no. 12, pp. 12778–12792, Dec. 2024.
efficient, and physically secure key agreement authentication protocol for [97] C. Correa-Jullian, M. Ramos, A. Mosleh, and J. Ma, ‘‘Operational
vehicular networks,’’ Electronics, vol. 13, no. 8, p. 1418, Apr. 2024. safety hazard identification methodology for automated driving sys-
[76] A. Singh, P. Rani, J. V. N. Ramesh, S. V. Athawale, A. H. Alkhayyat, tems fleets,’’ Proc. Inst. Mech. Eng., O, J. Risk Rel., 2024, doi:
A. N. Aledaily, T. A. Prola, and R. Sharma, ‘‘Blockchain-based lightweight 10.1177/1748006X241233863.
authentication protocol for next-generation trustworthy Internet of Vehi-
cles communication,’’ IEEE Trans. Consum. Electron., vol. 70, no. 2,
pp. 4898–4907, May 2024.
[77] M. Han, Z. Yin, P. Cheng, X. Zhang, and S. Ma, ‘‘Zero-knowledge identity
authentication for Internet of Vehicles: Improvement and application,’’
PLoS ONE, vol. 15, no. 9, Sep. 2020, Art. no. e0239043.
[78] J. M. Kizza, ‘‘Feige–Fiat–Shamir ZKP scheme revisited,’’ Int. J. Com-
put. ICT Res., vol. 4, no. 1, pp. 9–19, 2010.
[79] S. A. Abdel Hakeem, M. A. Abd El-Gawad, and H. Kim, ‘‘A decentralized
lightweight authentication and privacy protocol for vehicular networks,’’
IEEE Access, vol. 7, pp. 119689–119705, 2019.
[80] S. Windarta, S. Suryadi, K. Ramli, A. A. Lestari, W. Wildan, B. Pranggono,
and R. W. Wardhani, ‘‘Two new lightweight cryptographic hash functions
based on saturnin and beetle for the Internet of Things,’’ IEEE Access, HAZEM ISMAIL ALI received the Ph.D. degree
vol. 11, pp. 84074–84090, 2023. in electrical and computer engineering focused
[81] J. R. Vacca, Cyber Security and IT Infrastructure Protection. Rockland, on real-time embedded systems from the Faculty
MA, USA: Syngress, 2013.
of Engineering (FEUP), University of Porto,
[82] V. Fomichev, D. Bobrovskiy, A. Koreneva, T. Nabiev, and D. Zadorozhny,
Portugal, in 2017. He is currently an Assistant
‘‘Data integrity algorithm based on additive generators and hash function,’’
J. Comput. Virol. Hacking Techn., vol. 18, no. 1, pp. 31–41, Mar. 2022.
Professor with the Department of Computing
[83] C. Nist, ‘‘The digital signature standard,’’ Commun. ACM, vol. 35, no. 7, and Electronics for Real-Time and Embedded
pp. 36–40, Jul. 1992. Systems (CERES), School of Information Tech-
[84] J. Enare Abang, H. Takruri, R. Al-Zaidi, and M. Al-Khalidi, ‘‘Latency nology (ITE), Halmstad University (HH), Sweden.
performance modelling in hyperledger fabric blockchain: Challenges and His research interests include real-time embedded
directions with an IoT perspective,’’ Internet Things, vol. 26, Jul. 2024, systems, high-performance computing, the Internet of Things (IoT),
Art. no. 101217. hardware/software acceleration, and embedded system design.

VOLUME 13, 2025 24221

 H. I. Ali et al.: Navigating the Challenges and Opportunities of Securing IoAV

HARRISON KURUNATHAN received the bach- FLAVIUS GRUIAN received the Diploma of
elor’s degree in electronics and communication Engineering degree in computer science from
engineering from SRM University, India, in 2012, Politehnica University of Timisoara, Timisoara,
the master’s degree in very large-scale integration Romania, in 1996, and the Ph.D. degree in
from the SSN College of Engineering, Anna computer science from Lund University, Sweden,
University, India, in 2014, and the Ph.D. degree in 2002. He is currently an Associate Professor
in electrical and computer engineering from the in embedded systems with the Department of
University of Porto, Portugal, in 2021. He is Computer Science, Lund University, Sweden.
currently an Integrated Researcher with CISTER, From 2005 to 2006 he was a Postdoctoral Fellow
Porto, Portugal. Before his current role, he was an of the Department of Electrical and Computer
Assistant Professor with the Department of Electronics and Communication Engineering, University of Auckland, New Zealand. His research interests
Engineering, SRM University. His research interests include machine include scheduling for real-time systems, low-power and low-energy
learning, wireless sensor networks, visible light communication (VLC), systems, hardware/software co-design, the IoT, languages and tools for
network protocols, network security, the Internet of Things (IoT), and streaming applications, and compilation for ML hardware.
cooperative cyber-physical systems (CPS).

MAGNUS JONSSON (Senior Member, IEEE)

received the B.S. and M.S. degrees in computer
engineering from Halmstad University, Halmstad,
Sweden, in 1993 and 1994, respectively, and the
Licentiate of Technology and Ph.D. degrees in
computer engineering from the Chalmers Uni-
versity of Technology, Gothenburg, Sweden, in
1997 and 1999, respectively. From March 1998 to
March 2003, he was an Associate Professor in
data communication with Halmstad University
(acting between 1998 and 2000). Since 2003, he has been a Full Professor
MOHAMED HAMDY ELDEFRAWY received the in real-time computer systems with Halmstad University, where he is
Doctorate degree in electrical engineering from currently the Deputy Program Director for the university-wide profile area
Alexandria University, Egypt, in 2014. He was a smart cities and communities. Among previous assignments with Halmstad
Senior Researcher with the Center of Excellence University are, e.g., the Department Manager, the Vice Dean, and the
in Information Assurance, King Saud University, Director of Research. He has published over 135 scientific articles and
Riyadh, Saudi Arabia, and later as a Postdoctoral book chapters, most of them in the areas of vehicular communication,
Researcher with the Department of Information real-time communication, industrial communication, wireless networking,
Systems and Technology, Mid Sweden University, real-time and embedded computer systems, optical networking, and optical
Sundsvall, Sweden. Currently, he holds an asso- interconnection architectures. He has served on the program committees
ciate professor (senior lecturer) position with the of over 80 conferences and workshops and served as the International
School of Information Technology, Halmstad University, Sweden. With over Liaison Co-Chair for the 28th International Conference on Distributed
25 research articles published in esteemed journals and conferences, he is Computing Systems (ICDCS 2008), the General Co-Chair for the Fifth
also the inventor of two U.S./PCT patents in cybersecurity. His research International Workshop on Multiple Access Communications (MACOM-
interests include network security, digital authentication, and information 2012) and MACOM 2016, and the TCP Co-Chair for Nets4cars-Nets4trains
assurance. He was awarded the Bronze Medal at the 41st International 2013, MACOM 2013, Nets4cars-Nets4trains 2014, and MACOM 2015.
Exhibition of Inventions in Geneva, Switzerland, in April 2013, and serves He has got five best paper awards.
as a technical reviewer for numerous international journals and conferences.

24222 VOLUME 13, 2025