[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Next Article in Journal
Enhanced Fall Detection Using YOLOv7-W6-Pose for Real-Time Elderly Monitoring
Next Article in Special Issue
IoT-Based LPG Level Sensor for Domestic Stationary Tanks with Data Sharing to a Filling Plant to Optimize Distribution Routes
Previous Article in Journal
Characterising Payload Entropy in Packet Flows—Baseline Entropy Analysis for Network Anomaly Detection
Previous Article in Special Issue
A Microservice-Based Smart Agriculture System to Detect Animal Intrusion at the Edge
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Future Internet
Volume 16
Issue 12
10.3390/fi16120471
futureinternet-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Enhancing Autonomous Vehicle Safety with Blockchain Technology: Securing Vehicle Communication and AI Systems

by
Stefan Iordache
*,
Catalina Camelia Patilea
and
Ciprian Paduraru
*
Department of Computer Science, University of Bucharest, 010014 Bucharest, Romania
*
Authors to whom correspondence should be addressed.
Future Internet 2024, 16(12), 471; https://doi.org/10.3390/fi16120471
Submission received: 29 October 2024 / Revised: 7 December 2024 / Accepted: 10 December 2024 / Published: 18 December 2024
(This article belongs to the Special Issue IoT Architecture for Smart Environments: Mechanisms, Approaches, and Applications)
Browse Figures
Figure 1
<p>Fog computing architecture with three layers: cloud layer (<b>top</b>) for centralized storage and large-scale processing; fog layer (<b>center</b>) for decentralized computation closer to data sources; and edge layer (<b>bottom</b>) for end devices like vehicles generating and collecting data.</p> "> Figure 2
<p>Types of denial-of-service (DoS) attacks in vehicular communication systems: (a) packet flooding, (b) V2V radio jamming, and (c) V2I radio jamming.</p> "> Figure 3
<p>Illustration of a Sybil attack in a vehicular network. A malicious vehicle (red car) creates multiple fake identities to inject false information into the network, misleading nearby vehicles and disrupting communication. Examples include spoofed hazard warnings, false location data, and network congestion, which compromise the safety and trustworthiness of vehicle-to-vehicle (V2V) communication.</p> ">
Versions Notes

Abstract

:
In recent years, the rapid development of autonomous vehicles (AVs) has brought new challenges in terms of data security, privacy, and communication integrity. Our research investigates the potential of blockchain technology to improve the security of AVs by securing vehicle communication systems. By integrating blockchain with AI-based predictive algorithms, this approach aims to secure vehicle peer-to-peer communication, reduce traffic congestion, and improve safety for drivers and pedestrians. Blockchain’s decentralized ledger ensures the integrity of data exchange between vehicles and smart city infrastructure and mitigates the risks of cyberattacks such as data manipulation and identity forgery. This paper also examines recent advances in vehicular ad hoc networks (VANETs) and vehicular social networks (VSNs), and it demonstrates how the immutability and cryptographic security of the blockchain can strengthen AV systems. The proposed architecture not only protects user privacy but also decentralizes access to critical data needed for AI-driven decisions, ultimately promoting a safer and more reliable environment for autonomous vehicles.

1. Introduction

As autonomous vehicles (AVs) become an integral part of modern transportation, ensuring the safety and reliability of vehicle communication systems is of paramount importance. Autonomous vehicles rely on interconnected networks for data exchange between vehicles and urban infrastructure, which poses significant risks, including cyberattacks, data tampering, and impersonation. These vulnerabilities can undermine the security of AVs and their ability to make accurate decisions in real-time traffic scenarios.
Blockchain technology offers a promising solution [1,2,3] to these challenges by providing a decentralized, immutable ledger that increases data integrity [4] and security. By integrating blockchain with AI-driven predictive algorithms, AVs can communicate securely in vehicular ad hoc networks (VANETs) [5,6,7], which are critical for reducing traffic congestion, improving decision-making, and increasing overall road safety. The cryptographic functions of the blockchain ensure that the data exchanged between vehicles are tamper-proof and that private user information remains secure.
Recent research shows the potential of combining AI and blockchain in AV systems to increase security, decentralize data access, and protect against common attacks such as data tampering and Sybil attacks [2,3,8,9]. This paper examines the role of blockchain technology in improving the security of autonomous vehicle networks. It analyzes the extent to which it is able to secure communication systems and improve the trustworthiness of AI-based decision-making processes. By addressing these key challenges, blockchain can serve as an important component in building a safe and efficient future for autonomous transportation.
In this paper, we propose an architecture that combines blockchain technology with AI-based predictive models, especially reinforcement learning, to improve the security of autonomous vehicle communication systems. Our implementation not only secures the data exchanged between vehicles but also improves decision-making processes by protecting against data manipulation. We have designed and implemented this architecture as a proof of concept to demonstrate its practicality in real vehicle networks. This implementation shows the potential of the blockchain to revolutionize the safety of autonomous vehicles and pave the way for more secure and trustworthy AV systems. By enabling secure data sharing and decentralized decision making, our solution lays the foundation for the integration of autonomous vehicles into broader urban infrastructures and advances the vision of smart and connected cities of the future.
In this work, we have implemented an architecture that addresses these challenges based on previous research. Our architecture allows vehicles to communicate either directly or through a master node. Road data and sensor information are collected from the nearest roadside units (RSUs) and forwarded to a blockchain ledger infrastructure deployed at the edge layer. These data are then processed and validated before being sent to a cloud layer where predictive AI models are stored and continuously trained with real-time inputs. These models are fed back to the vehicles and infrastructure to reduce congestion and improve road safety.
We have integrated the Byzantine Fault Tolerance protocol to ensure a robust consensus across the blockchain network, and smart contracts [10] manage participant authentication, data storage, and AI model sharing between entities. Implemented as a proof of concept, this architecture demonstrates the potential of the blockchain for securing AV systems and paves the way for its integration into advanced smart cities. It can be further customized to support more advanced autonomous driving simulators such as CARLA [11], and it can also be deployed in various cloud environments that support blockchain-as-a-service.
The contributions of our work can be summarized as follows:
  • A complete blockchain-based architecture proposal for autonomous vehicle communication, along with an extensible and reusable prototype.
  • The integration of smart contracts for cross-layer communication in the authentication of participants, data storage and processing, and management of deployed AI models across AVs.
  • A comprehensive study and comparison of the current state of the art in terms of the interaction between IoT devices and blockchain algorithms. This study is supported through implemented experiments.
The rest of the paper is structured as follows. The next section contains a general overview of the terminology and theoretical matters that recur throughout the paper. Section 3 discusses the state of the art in the field of components that we integrate in their entirety in this paper. The proposed architecture and implementation are described in Section 4. An evaluation and discussion of the results, threats to validity, and experimental design are discussed in Section 5 and Section 6. Finally, conclusions and a plan for future work are presented in Section 7 and Section 8.

2. Background

2.1. Vehicular Ad Hoc Networks (VANETs)

Vehicular ad hoc networks (VANETs) are a special form of mobile ad hoc networks (MANETs) that enable communication between vehicles and roadside infrastructure without relying on a central management system. The main objective of VANETs is to improve road safety, traffic management, and the overall driving experience by facilitating communication between vehicles (vehicle-to-vehicle or V2V) and between vehicles and infrastructure (vehicle-to-infrastructure or V2I).

2.2. IoT Devices

2.2.1. Roadside Units (RSUs)

Roadside units (RSUs) are critical components in vehicle communication networks such as vehicular ad hoc networks (VANETs). They are permanently installed infrastructure devices that are usually located along roads or at intersections and enable communication between vehicles (vehicle-to-infrastructure, or V2I), as well as between vehicles and external networks such as the Internet or traffic management systems.

2.2.2. Smart Traffic Lights

These IoT-enabled traffic lights can dynamically adapt based on real-time traffic data. They can communicate with vehicles to optimize traffic flow and reduce congestion, while also sending signals to vehicles about upcoming red lights or emergency situations.

2.3. Blockchain

2.3.1. Ledger

In the context of the blockchain, a ledger is a decentralized, distributed database that records transactions or data across multiple nodes in a network. Unlike traditional ledgers, which are usually managed by a central authority (e.g., banks or organizations), blockchain ledgers are managed collectively by all participants in the network. Each entry in the ledger is immutable, i.e., it cannot be changed once it has been recorded, which ensures transparency and security.

2.3.2. Smart Contracts

Smart contracts are self-executing contracts whose terms of the agreement are written directly into lines of code. These contracts automatically enforce and execute the obligations of both parties when predefined conditions are met, without the need for intermediaries such as lawyers or notaries. Smart contracts work with blockchain technology and ensure that all transactions and processes related to the contract are transparent, immutable, and secure.

2.3.3. Byzantine Fault Tolerance Protocol

Byzantine Fault Tolerance (BFT) is a consensus mechanism used in distributed networks, such as blockchain systems, to ensure that the network can function correctly even if some nodes (participants) act maliciously or fail. The term comes from the Byzantine Generals Problem, which describes the challenge of reaching an agreement in the presence of unreliable or treacherous actors.

2.3.4. Fog Computing

Fog computing (Figure 1) is a decentralized computing model that brings data processing closer to the source, reducing latency and improving real-time capability. It distributes computing and storage tasks to devices at the edge of the network, such as routers or IoT devices, instead of relying on distant cloud servers. This proximity enables faster processing and decision-making, which is crucial for applications such as autonomous vehicles and smart cities.
  • Cloud layer-handles centralized storage and large-scale data processing tasks, such as managing extensive databases or running computationally intensive applications.
  • Fog layer-acts as an intermediary between the cloud and edge devices, consisting of local servers, routers, and firewalls. This layer processes data closer to its source, reducing latency and ensuring that time-sensitive applications (e.g., vehicular networks) perform efficiently.
  • Edge layer-comprises end devices like vehicles, sensors, and other IoT systems that generate and collect data. These devices communicate directly with the fog layer or through roadside units (RSUs), enabling real-time interactions and local processing.
By adopting this layered approach, fog computing not only improves system responsiveness but also enhances security and privacy by keeping sensitive data local. At the same time, it reduces the load on cloud infrastructure by completing data-intensive tasks closer to their origin.

2.4. Attacks in V2X Scenarios

2.4.1. Denial of Service (DoS) Attack

A denial of service (DoS) attack in vehicle-to-everything (V2X) scenarios refers to an attack in which malicious entities intentionally flood the network with excessive or unauthorized requests to disrupt or completely stop communication between vehicles and infrastructure. V2X communication, which includes vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), and vehicle-to-pedestrian (V2P) interactions, relies on seamless data transmission to ensure safety, traffic efficiency, and real-time decision-making. In a DoS attack, attackers aim to overwhelm the network or specific components of the system, rendering them inoperable and compromising the reliability of critical vehicular systems.
DoS attacks in vehicular networks can take several forms:
  • Packet flooding-the attacker floods the network with a large number of unnecessary or malicious packets, overloading communication channels and preventing legitimate vehicles from sending or receiving critical data, such as collision warnings or traffic updates.
  • Radio jamming in V2V communication-the attacker disrupts direct communication between vehicles by emitting high-power signals that jam the radio frequencies used for vehicle-to-vehicle interactions. This prevents vehicles from exchanging safety-critical information, potentially leading to accidents or traffic congestion.
  • Radio jamming in V2I communication-the attacker targets the communication between vehicles and infrastructure, such as roadside units (RSUs), by jamming the radio frequencies used for vehicle-to-infrastructure interactions. This disruption can prevent vehicles from receiving vital data, such as traffic light signals, road conditions, or navigation assistance, affecting both safety and traffic flow.
These types of DoS attacks are illustrated in Figure 2, which shows examples of packet flooding, V2V radio jamming, and V2I radio jamming in a vehicular communication system. The figure demonstrates how these attacks can severely disrupt real-time data exchange, underscoring the importance of robust security measures to mitigate such threats.

2.4.2. Sybil Attack

A spoofed identity attack (often referred to as a Sybil attack, illustrated in Figure 3) occurs when a malicious actor on a network creates multiple false identities or nodes in order to manipulate, disrupt, or gain disproportionate control over a network. In the context of vehicular networks, such as V2X (vehicle-to-everything) communication systems, a fake identity attack can have serious consequences for traffic safety, communication integrity, and trust within the network.

3. Related Work

The intersection of blockchain technology, artificial intelligence, and autonomous vehicle (AV) security has attracted considerable attention in recent years, particularly in securing vehicle communications and mitigating cyber threats. Vehicular ad hoc networks (VANETs) and vehicle-to-everything (V2X) communications provide the basis for real-time peer-to-peer interaction between AVs, but they also expose these networks to vulnerabilities such as Sybil attacks and denial-of-service (DoS) attacks. Traditional centralized security methods have proven inadequate for the size and complexity of AV networks. Therefore, researchers have explored the decentralized ledger of the blockchain as a solution for securing data integrity and vehicle authentication [2,5].
The integration of the blockchain with AI has shown promise in strengthening the resilience of vehicle networks. Studies suggest that the consensus mechanisms Byzantine Fault Tolerance (BFT) and Practical Byzantine Fault Tolerance (PBFT) [12] increase security against identity spoofing attacks, a common problem in V2X networks [13]. In a recent study [14] applied these protocols in blockchain architectures and demonstrated their effectiveness in maintaining network reliability, even in scenarios with malicious nodes.
To mitigate latency issues, fog and edge computing models are often used alongside the blockchain. These approaches help reduce the computational load by distributing tasks to roadside units (RSUs) close to data sources, thus meeting the real-time processing requirements for AVs [15]. In one case, Patel et al. (2023) integrated [8] RSUs with lightweight blockchain clients to enable localized data validation and ensure faster decision making in dynamic environments such as V2I communication.
In addition, smart contracts play a crucial role in automating secure transactions in these networks, where they help maintain transparency in data sharing and manage privacy controls that are essential for regulatory compliance in vehicular communication systems [16]. Recent work by Tan and Li (2024) [17] has shown that smart contracts can effectively regulate the interaction of AV data by providing decentralized access only to authorized entities. The privacy mechanisms in these contracts, which are often reinforced by cryptographic algorithms such as Elliptic Curve Cryptography (ECC) [18], further protect user data.
As blockchain and AI technologies mature, studies are increasingly focusing on optimizing these systems for large-scale use, especially in the context of smart cities. Challenges such as scalability and energy efficiency [19] remain at the forefront, with researchers proposing adaptive strategies such as sharding and lightweight consensus models to efficiently process large amounts of data [20,21]. These advances not only contribute to scalability but also improve fault tolerance in AV networks, where maintaining low latency is critical to vehicle responsiveness and user safety.
While previous research has made significant strides in integrating blockchain and artificial intelligence (AI) to enhance vehicular network security, several critical challenges remain. Many studies focus on specific security threats, such as Sybil attacks or denial of service (DoS) attacks, without addressing the scalability and real-time processing requirements essential for practical deployment in dynamic vehicular environments.
Existing studies on vehicular ad hoc networks (VANETs) and V2X communication systems often employ blockchain for data integrity and identity authentication [22]. However, they rely on consensus mechanisms that are not optimized for real-time environments. For instance, most use traditional consensus protocols that introduce high latency, which is unsuitable for time-sensitive vehicular applications. The proposed system overcomes this limitation by utilizing a Practical Byzantine Fault Tolerance (PBFT) consensus mechanism. PBFT is specifically designed for environments with a moderate number of participants, ensuring both fast consensus and resilience to malicious nodes, a critical requirement for vehicular networks.
Another gap lies in the lack of comprehensive testing under real-world conditions. While existing studies often demonstrate the effectiveness of blockchain-based systems under idealized settings, they do not account for the impact of vehicle mobility, network jitter, or latency variability on system performance. This research fills this void by simulating dynamic network conditions and analyzing their effects on Sybil attack detection, DoS mitigation, and consensus latency. These evaluations provide valuable insights into the robustness and scalability of the proposed system in practical environments.
In summary, the proposed framework builds on existing research by combining decentralized consensus mechanisms, AI-driven decision-making, and modular architecture to address the challenges of scalability, real-time processing, and adaptability. By bridging these gaps, this work advances the field of vehicular communication systems, paving the way for secure and efficient deployment in smart city environments.

4. Proposed Architecture and Implementation

Our implementation proposes a secure communication system for autonomous and connected vehicles using blockchain technology. The system addresses the key challenges of data integrity, authentication, and privacy in vehicle-to-everything (V2X) communication. The architecture combines the blockchain with smart contracts and consensus mechanisms to ensure a secure, decentralized, and reliable data exchange between vehicles (V2V) and between vehicles and infrastructure (V2I).

4.1. System Architecture Overview

The architecture implemented in this research consists of the following core layers:
  • Vehicles layer: vehicles equipped with IoT devices and sensors on board generate real-time data.
  • Edge/fog layer: roadside units (RSUs) serve as edge nodes to facilitate local processing and reduce latency.
  • Blockchain layer: a decentralized ledger where all transactions (vehicle interactions and data exchanges) are securely recorded.
  • Cloud layer: for high-level processing and analysis, data that do not occur in real time can be outsourced to the cloud.
  • Smart contracts: to automate the enforcement of policies for secure communication, authentication, and data exchange.

4.2. Technical Implementation of the System

To ensure robust and secure communication within the vehicular network, the proposed framework integrates blockchain technology with cryptographic methods, consensus algorithms, and smart contracts. Below are the detailed components of the system implementation.
  • Initialization and node configuration. Each vehicle and RSU acts as a blockchain node and is equipped with the following:
    –
    Public–private key pairs-each vehicle and RSU is assigned a unique identity represented by a public–private key pair using elliptic curve cryptography (ECC). The public key acts as an identifier, while the private key is used to sign transactions securely.
    –
    Blockchain client-vehicles and RSUs are equipped with lightweight blockchain clients to minimize resource usage while ensuring compatibility with the decentralized network.
    –
    IoT sensors-vehicles are equipped with various sensors (e.g., GPS, LIDAR, and cameras) to collect data for secure communication and decision making. These sensor data are signed and securely transmitted to RSUs and other vehicles for validation.
  • Secure data generation and signing. When a vehicle generates data (e.g., location, speed, and hazard warnings), the system ensures data integrity through cryptographic signing:
    –
    The vehicle’s on-board unit (OBU) signs the data using its private key.
    –
    The signed data are packaged into a blockchain transaction and propagated across the network to nearby RSUs and vehicles.
  • Data validation via the blockchain.
    –
    Transaction propagation-the transaction (signed data) is propagated through the network and received via nearby RSUs (edge nodes) and vehicles.
    –
    Consensus mechanism-RSUs use the Practical Byzantine Fault Tolerance (PBFT) protocol to validate transactions:
    *
    Verify the signature using the sender’s public key.
    *
    Ensure the data’s integrity and authenticity.
    –
    Transaction recording-once the transaction is validated, it is bundled into a block. Each RSU maintains a local blockchain ledger, and validated blocks are propagated to other RSUs in the network to ensure consistency.
  • Smart contract deployment. Smart contracts play a critical role in automating network operations:
    –
    Authentication-when a vehicle or RSU joins the network, the smart contract verifies its identity using its public–private key pair.
    –
    Role-based access control-smart contracts enforce access restrictions to sensitive data. For example, hazard alerts are broadcast to all nearby vehicles, but personal vehicle information remains encrypted and accessible only to authorized nodes.
  • Tools and platforms.
    –
    Hyperledger fabric-used as the blockchain platform due to its support for permissioned networks and modular architecture.
    –
    Python-utilized for simulations and integration with Hyperledger Fabric.
    –
    Key libraries-cryptography is used for ECC key generation and signature verification, while SimPy is employed for network latency and mobility modeling.

4.3. Communication Proof of Concept

In our implementation, we developed and deployed two primary communication scenarios within the blockchain-based secure vehicle communication system: vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication.
  • Secure vehicle-to-vehicle communication (V2V)
    –
    Data broadcast-vehicles broadcast signed messages, such as emergency braking or hazard warnings, to nearby vehicles and RSUs.
    –
    Message verification (Algorithm 1)-the receiving vehicles verify the authenticity of the message by checking the sender’s public key and the blockchain entries. This ensures that the message does not originate from a malicious actor (protection against Sybil attacks).
    Algorithm 1 Message verification algorithm
      1:
    Input: Message m, Sender’s Public Key K p u b
      2:
    Output: Validation Status (True/False)
      3:
    Step 1: Extract the digital signature from m
      4:
    Step 2: Verify the signature using K p u b
      5:
    Step 3: Check the blockchain ledger for the sender’s identity
      6:
    if signature and sender identity are valid then
      7:
        return True
      8:
    else
      9:
        return False
    10:
    end if
    –
    Data sharing-after validation, the data are used to support the receiving vehicle’s decision making (e.g., adjusting speed to avoid a collision).
  • Secure vehicle-to-infrastructure communication (V2I)
    –
    Data aggregation with RSUs-RSUs collect and process data from nearby vehicles. They can take over local traffic management, store road condition data, or interact with traffic lights to optimize traffic flow.
    –
    Blockchain recording-important V2I interactions (such as toll payments or the prioritization of emergency vehicles) are recorded in the blockchain for transparency and verification. The message does not originate from a malicious actor (protection against Sybil attacks).
    –
    Fog computing-part of the data processing takes place at the edge (RSUs) so that decisions can be made in real time without having to send all data to the cloud, reducing latency.
In our implementation, we take into account the dynamic nature of vehicles moving across different geographical areas and ensure that each vehicle remains connected to the nearest available roadside unit (RSU). As vehicles are constantly changing position, it is crucial to maintain an efficient and secure communication link with the nearest RSU for reliable data exchange. To achieve this, we have developed an algorithm that calculates the distance between the moving vehicle and the nearby RSUs and determines which RSUs are within communication range. This algorithm ensures that the vehicle seamlessly connects to the RSU that offers the best signal strength and proximity, enabling continuous and secure vehicle-to-infrastructure (V2I) communication.
Algorithm 2 uses the current GPS position of the vehicle and calculates the distance to each RSU using Haversine’s formula [23], which is well suited to calculating distances on the earth’s surface. If the calculated distance is less than or equal to the RSU’s communication range, the RSU is considered "in range" and is added to the list of available RSUs for the vehicle. This allows the vehicle to dynamically adjust its connection as it moves through different regions.
Algorithm 2 Algorithm to determine RSUs in range
  1:
Input: vehicle position v c u r , list of RSUs { R S U } , RSU range r r a n g e
  2:
Output: List of RSUs within range { R S U i n _ r a n g e }
  3:
Initialize: R S U i n _ r a n g e ← ∅
  4:
function DETERMINE_RSUs( v c u r , R S U _ l i s t , r r a n g e )
  5:
    for  R S U in R S U _ l i s t  do
  6:
         d i s t ←  CALCULATE_DISTANCE( v c u r , R S U . p o s i t i o n )
  7:
        if  d i s t ≤ r r a n g e  then
  8:
            R S U i n _ r a n g e .add( R S U )
  9:
        end if
10:
    end for
11:
    return  R S U i n _ r a n g e
12:
end function
13:
function CALCULATE_DISTANCE( v c u r , R S U _ p o s )
14:
     l a t v , l o n v ← v c u r
15:
     l a t r s u , l o n r s u ← R S U _ p o s
16:
     d L a t ← r a d i a n s ( l a t r s u − l a t v )
17:
     d L o n ← r a d i a n s ( l o n r s u − l o n v )
18:
     a ← sin 2 ( d L a t 2 ) + cos ( r a d i a n s ( l a t v ) ) × cos ( r a d i a n s ( l a t r s u ) ) × sin 2 ( d L o n 2 )
19:
     c ← 2 × atan 2 ( a , 1 − a )
20:
     d i s t a n c e ← R e a r t h × c
21:
    return  d i s t a n c e
22:
end function

4.4. Blockchain Components

  • Blockchain platform-our implementation is based on Hyperledger Fabric [24], a permissioned blockchain platform that ensures that only authorized vehicles and RSUs participate in the network.
  • Consensus algorithm-the system uses a Byzantine Fault Tolerant (BFT) [25] consensus mechanism, in this case Practical Byzantine Fault Tolerance (PBFT) (Algorithm 3) [26], which is suitable for environments with a relatively small number of participants (RSUs and vehicles) and requires fast consensus.
  • Cryptographic techniques-the developed system relies on elliptic curve cryptography (ECC) [27] for secure key generation, as it provides strong security with relatively small key sizes, which is ideal for resource-constrained in-vehicle IoT devices.
Algorithm 3 Practical Byzantine Fault Tolerance (PBFT)
  1:
Input: Client request r e q
  2:
Output: Agreement on the state of the system
  3:
Initialization:
  4:
  All nodes n i are initialized with the same initial state
  5:
  The system can tolerate up to f faulty nodes out of 3 f + 1 total nodes
  6:
  Primary node P is selected as the leader
  7:
function PBFT( r e q )
  8:
    Phase 1: Pre-Prepare
  9:
     Primary P receives the request from the client
10:
     Primary P assigns a sequence number to the request and multicasts a pre-prepare message to all replicas
11:
    Phase 2: Prepare
12:
    for each replica r i  do
13:
        Replica r i verifies the pre-prepare message
14:
        if message is valid then
15:
            r i multicasts a prepare message to all other replicas
16:
        end if
17:
    end for
18:
    A replica waits until it receives 2 f  prepare messages from different replicas
19:
    Phase 3: Commit
20:
    for each replica r i  do
21:
        if  r i receives 2 f  prepare messages then
22:
            r i multicasts a commit message to all replicas
23:
        end if
24:
    end for
25:
    A replica waits until it receives 2 f  commit messages
26:
    The request is executed and the result is sent to the client
27:
    return Agreement reached and state updated
28:
end function

4.5. Security and Privacy Measures Implemented

In our implementation, we prioritized security and privacy by integrating several essential mechanisms into the blockchain-based vehicular communication system. This part of the research followed three key aspects:
  • End-to-end encryption-we implemented end-to-end encryption to ensure that all communication between vehicles (V2V) and vehicles and infrastructure (V2I) is encrypted. Public–private key pairs were used, with each vehicle and RSU having their own cryptographic keys.
    • When data (e.g., vehicle location, speed, and hazard warnings) are transmitted, they are first encrypted with the sender’s private key to ensure that the data cannot be read by unauthorized parties. Only the recipient who has the corresponding public key can decrypt the message. This guarantees that the data remain confidential and tamper-proof throughout the entire communication.
    • We used elliptic curve cryptography (ECC) for public–private key pair generation, as it offers strong security properties and efficient key generation, which are crucial for resource-constrained environments such as vehicular networks.
  • Tamper-proof ledger-the blockchain acts as a tamper-proof ledger that records all vehicle interactions and data exchanges. Each transaction between vehicles and RSUs is validated via a consensus mechanism and stored in blocks that are cryptographically linked to previous blocks.
    • Once data have been written to the blockchain, they are immutable. This means that no vehicle, RSU, or external company can change or delete previous data entries, ensuring an accurate and transparent record of all events.
    • We configured the blockchain to be permissioned so that authorized nodes (such as vehicles and RSUs) can access the ledger while maintaining transparency across the network. This ensures accountability, as all interactions are logged, and the integrity of the data can be verified by all participants.
  • Decentralization-to ensure decentralization, we established an approved blockchain with multiple participating nodes. Vehicles, RSUs, and infrastructure nodes (e.g., cloud servers) each act as decentralized nodes in the network. No single node controls the entire system, making the network resilient to attacks.
    • We used the Practical Byzantine Fault Tolerance (PBFT) consensus algorithm to reach an agreement across the network, even in the presence of faulty or malicious nodes. This protocol ensures that the network can still reach consensus and validate transactions as long as fewer than 1/3 of the nodes are compromised.
    • Decentralization makes the system resistant to denial of service (DoS) and Sybil attacks. For example, in a Sybil attack, where an attacker creates multiple fake identities to manipulate the network, the blockchain’s consensus protocol ensures that only authenticated nodes can participate in the decision-making process. In addition, the distributed nature of the system means that there is no single point of failure, which minimizes the likelihood of successful DoS attacks.

4.6. Considerations for Optimizing Performance

In our research project, we implemented several performance optimization measures to ensure that the blockchain-based communication system works efficiently, especially in high-traffic environments with numerous vehicles and roadside units (RSUs). These optimizations focused on edge processing and scalability to keep the system running smoothly without overloading the blockchain or causing significant delays.
  • Edge processing-to minimize the communication load on the blockchain, we deployed RSUs (roadside units) as edge computing nodes [28,29]. RSUs are strategically placed at the roadside and are responsible for local data processing tasks such as traffic monitoring, hazard detection, and vehicle coordination. By processing data locally, RSUs significantly reduce the need for constant interaction with the cloud or blockchain network.
    • RSUs are able to make real-time decisions at the border. For example, if a hazard is detected, the RSU analyzes the data locally and transmits the warning directly to nearby vehicles. Only important data, such as final decisions or important transaction details, are sent to the blockchain for logging, reducing the overall communication load of the system.
    • This implementation reduces latency by keeping processing close to the data source and minimizing the need to forward non-critical information to the cloud or blockchain. It also conserves bandwidth and improves the overall responsiveness of the system to real-time traffic conditions.
  • Scalability-to ensure that the system can handle a large number of vehicles, we designed the blockchain with a hierarchical architecture. The system includes lightweight nodes (vehicles) and more powerful nodes (RSUs) that manage the ledger and perform more resource-intensive tasks.
    • Lightweight nodes (vehicles): Vehicles act as lightweight nodes in the network. They primarily generate and verify smaller transactions (e.g., position updates or hazard warnings) and rely on the RSUs for more complex blockchain interactions.
    • Powerful nodes (RSUs): RSUs serve as more robust nodes capable of managing parts of the blockchain ledger and participating in the consensus mechanism. By distributing the more computationally intensive tasks to the RSUs, the system remains scalable even if the number of vehicles increases.

5. Evaluation

The main goal of our research was to improve the security and reliability of vehicular communication networks by implementing a blockchain-based system that is resistant to common cyberattacks. Our system was designed to mitigate Sybil attacks, prevent denial of service (DoS) attacks, and ensure the robustness of the Practical Byzantine Fault Tolerance (PBFT) consensus mechanism, even in the presence of faulty or malicious nodes.

5.1. Safety Simulations and Feedback

  • Sybil attack scenario-The first scenario in our evaluation involved the simulation of a Sybil attack where a malicious node attempted to create 50 false identities to disrupt the network. The attacker’s goal was to inject false data into the vehicle-to-vehicle (V2V) communication to manipulate and disrupt the coordination between the vehicles. By injecting these fake identities, the attacker attempted to confuse the network by generating false information about vehicle locations, traffic conditions or hazards, potentially leading to false actions by other vehicles. This type of attack can lead to significant safety risks, including traffic jams or collisions, as vehicles are misled about the actual state of the environment.
    Our system’s blockchain-based consensus mechanism was able to successfully detect and block all 50 fake identities, preventing any manipulation of V2V communication. This result demonstrates the robustness of the system’s authentication process, which relies on cryptographic identities validated via the blockchain to ensure that only legitimate vehicles can participate in the network. The system’s ability to mitigate this attack with a 100% success rate while minimally impacting latency underscores its effectiveness in securing vehicle communications against Sybil attacks.
  • Denial of service attack scenario-in the second scenario, we simulated a denial of service (DoS) attack in which a malicious node flooded the network with 1000 fake transaction requests per second and targeted the communication between vehicles and RSUs (vehicle-to-infrastructure, or V2I) to overload the blockchain nodes. The goal of the attack was to disrupt the network by overloading the roadside units (RSUs) and preventing legitimate transactions from being processed.
    The system’s consensus mechanism, combined with rate-limiting measures, successfully mitigated the attack by discarding 95% of malicious requests at the network level before they could reach the blockchain. The remaining 5% of requests were validated by the RSUs, identified as fake transactions, and discarded without being included in a block, ensuring that the malicious activity did not compromise the integrity of the blockchain ledger.
    • Mitigation rate-the system was able to neutralize 95% of attacks at the network layer, while the blockchain layer rejected 100% of forged transactions, maintaining data accuracy and security.
    • Impact on throughput-during the peak of the attack, the throughput of the system was reduced by 20% as resources were temporarily allocated to process and forged requests were rejected. However, no legitimate transaction was delayed by more than 50 ms, allowing communication between vehicles and RSUs to continue despite the attack.
It is important to note that the scenarios described are based on simulations and not on tests with actual vehicles or infrastructure. Due to the limitations of this project, including access to real vehicle networks, we used simulation environments to model and evaluate the performance of the system. These simulations allowed us to realistically assess the behavior of the system under certain attack scenarios, including Sybil attacks and denial of service (DoS) attacks, and to measure important metrics such as mitigation rates and impact on throughput.
By using simulation tools, we were able to recreate network conditions similar to those in real vehicle communication systems. While these simulations provided valuable insight into the effectiveness of the system, the results should be viewed with the understanding that real-world conditions may introduce additional complexities and variability that are not fully captured in a simulated environment.
For these simulations, we implemented all components using Python in a simplified form. Python provided a flexible and efficient framework for developing and testing the key aspects of the blockchain-based secured communication system. We simulated the interactions between the vehicles, the behavior of the RSUs, and the consensus mechanisms of the blockchain in a controlled environment and used libraries for cryptography, networks, and consensus modeling.
Through the use of Python, we were able to implement core processes such as the following:
  • The generation and validation of vehicle identities (to mitigate Sybil attacks);
  • The processing of large volumes of transaction requests during DoS attack simulation;
  • And the implementation of the Practical Byzantine Fault Tolerance (PBFT) algorithm for consensus.
This approach allowed us to focus on validating the underlying security protocols and communication structures without requiring the complexity of the actual physical infrastructure. Python’s simplicity and extensive libraries were ideal for simulating such blockchain-based communication systems while ensuring that key metrics and security objectives were effectively tested within the confines of our project. Table 1 showcases the scenarios listed above.

5.2. Consensus Mechanism Evaluation (PBFT)

Evaluating the Practical Byzantine Fault Tolerance (PBFT) was crucial, as it ensures that the blockchain-based vehicular communication system operates securely and reliably even in the presence of malicious or faulty nodes. PBFT allows the network to reach consensus and validate transactions as long as fewer than a third of participating nodes are compromised. This makes PBFT ideal for environments where trust and data integrity are critical, such as vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications.
By testing PBFT in simulated scenarios with malicious nodes, we were able to evaluate the system’s fault tolerance, verify its ability to maintain consensus stability, and ensure the system’s robustness to attacks or failures. This evaluation was crucial in proving that the system is capable of securing vehicle communication networks while ensuring reliable transaction validation and network stability.
  • Faulty node-tolerance scenario-in a simulated scenario, faulty or malicious nodes were introduced into the network, with the number of faulty nodes gradually increasing from 0 to 5 out of a total of 10 nodes to simulate Byzantine failures.
    As a result, the system successfully reached a consensus with up to three faulty nodes (in accordance with the PBFT tolerance of up to f = 1/3 Byzantine nodes). When the number of faulty nodes increased to four or more, consensus could not be reached, and the network could not reach an agreement concerning transaction validation.
    –
    Fault tolerance-the system was able to tolerate up to 3 faulty nodes in a 10-node network while maintaining consensus and data integrity.
    –
    Impact on latency-With one to two faulty nodes, the consensus time increased by 10%. With three faulty nodes, the consensus time increased by 25%. When the number of faulty nodes reached four, the system could no longer agree on the validity of the transaction.
    The PBFT consensus algorithm worked as expected and successfully maintained a consensus with up to f = 1/3 faulty nodes. Beyond this limit, the network became unstable, as predicted through PBFT theory.
  • Malicious node injection scenario-malicious nodes attempted to participate in the consensus process by sending corrupted or false blocks to disrupt the network.
    PBFT successfully identified and excluded the malicious nodes during the preparation and handover phase. The corrupted blocks were discarded, and the network reached a consensus with valid blocks from honest nodes.
    –
    Mitigation rate-100% of malicious nodes were detected and excluded from the consensus process.
    –
    Impact on latency-the presence of malicious nodes increased the consensus time by 15%, but the system remained functional and stable.
    The PBFT consensus mechanism proved to be effective in identifying and excluding malicious nodes, ensuring network operability without interruption, results that we summarized in Table 2.

5.3. Simulated Network Dynamics

To better understand the practical feasibility and performance of the proposed blockchain-based vehicular communication system, we conducted additional simulations incorporating real-world network dynamics. The focus was to evaluate the effects of vehicle mobility, communication jitter, and latency variability on system reliability and performance. These simulations allowed us to bridge the gap between idealized testing conditions and real-world scenarios and the results of this adjusted scenarios are showcased in Table 3.
  • Impact of vehicle mobility on communication latency-in real-world vehicular networks, the continuous movement of vehicles leads to dynamic changes in topology, which can introduce latency variations. Simulations were designed to account for the following:
    –
    Vehicles entering and exiting the communication range of roadside units (RSUs).
    –
    Changing distances between vehicles and RSUs, affecting the signal strength and communication delay.
    Observations:
    –
    As vehicles moved, latency increased slightly when transitioning between RSUs, averaging a 15% rise in the communication delay compared to static conditions.
    –
    The algorithm for selecting the nearest RSU (Algorithm 1 in the paper) effectively minimized disruptions, with vehicles reconnecting to optimal RSUs within 2–3 ms.
    –
    The dynamic handoff process demonstrated that the framework could maintain a consistent data exchange without losing critical information, even during high-mobility scenarios.
  • Effects of jitter and latency variability on performance-jitter refers to the variability in packet transmission times, which is common in mobile vehicular networks due to fluctuating signal strength and environmental interference. To simulate this, random delays ranging from 5 to 50 ms were introduced into the communication paths between vehicles, RSUs, and blockchain nodes.
    Observations:
    –
    For low to moderate jitter (up to 10 ms), the framework’s performance remained stable, with Sybil attack detection rates exceeding 97%. The system handled the variability without noticeable degradation in its mitigation capabilities.
    –
    When jitter increased to 20 ms or more, there were slight delays in the consensus process, which affected transaction throughput. During peak jitter conditions, the throughput decreased by approximately 5%, although no malicious transactions were validated.
    –
    Severe jitter (above 40 ms) occasionally caused temporary delays in the system’s attack detection capabilities, but blockchain consistency was preserved as malicious nodes were eventually identified and excluded during subsequent consensus rounds.
The results of these dynamic simulations were compared with those obtained under idealized conditions in which all nodes had consistent connectivity and minimal delays.
Resilience to mobility-the framework effectively managed vehicle handovers between RSUs without a loss of communication integrity. The rapid re-establishment of connections supports its robustness in high-mobility environments such as urban traffic scenarios.
Tolerance of jitter-the system demonstrated strong resistance to network variability, maintaining high detection and mitigation rates even under challenging conditions. While severe jitter impacted latency and throughput marginally, the overall security framework remained intact.
Performance trade-offs-The comparison highlights a slight degradation in performance under dynamic conditions, but this is an expected trade-off, given the complexity of real-world vehicular networks.

6. Discussion

Several important points emerge in the discussion of our research, reflecting both the strengths of our system and areas for future consideration.

6.1. Effectiveness of Safety Mechanisms

Our simulations have shown that the blockchain-based communication system is very effective in defending against Sybil and DoS attacks. In particular, the system achieved a 100% success rate in blocking Sybil attacks, and it neutralized 95% of malicious traffic in DoS attacks without significantly disrupting legitimate transactions. This confirms the robustness of using blockchain in vehicle networks to ensure secure, authenticated communication.
However, these results are based on simulated environments, and it is important to consider the limitations of simulations. Real use-cases, where network conditions and vehicle mobility are unpredictable, may present challenges that were not captured in the controlled scenarios of our study.

6.2. Weighing Up Latency vs. Security

An important observation in our experiments was the compromise between security and performance. While the system effectively blocked malicious transactions, the presence of faulty or malicious nodes increased latency, especially in the consensus process. With one to three faulty nodes, the consensus time increased by up to 25%, and with a higher number, a consensus could no longer be reached.
This raises the important question of how much performance degradation is acceptable in the pursuit of strong security measures. In critical applications such as autonomous driving or real-time traffic management, even small increases in latency can have serious consequences. Therefore, optimizing the balance between safety and real-time performance remains a challenge that future research must address.

6.3. Scalability and Network Dynamics

The system was designed with edge computing in the RSUs to offload processing and reduce the load on the blockchain. This approach proved successful in the simulations, but in the real world, the availability and distribution of RSUs may vary, and edge resources may be limited. As vehicle density increases, the ability of the system to scale seamlessly needs to be thoroughly tested.
In addition, the decentralized nature of the blockchain ensures that there is no single point of failure, which is crucial for defending against attacks such as DoS. However, decentralized systems are also susceptible to higher communication overhead and a slower completion of transactions on large networks. This scalability issue could be solved by exploring sharding [30] techniques or reducing the frequency of communication between vehicles and the blockchain.

6.4. Real-World Feasibility and Deployment

Although our results are promising, the real-world deployment of blockchain-based communication systems in smart cities and autonomous vehicle networks is still in its early stages. Factors such as infrastructure costs, energy consumption, and the legal framework for data protection need to be carefully assessed. In addition, ensuring interoperability with existing vehicle communication protocols (e.g., V2X standards) is essential for broad acceptance.
Another important aspect is the energy efficiency of blockchain operations. In a vehicle environment where both vehicles and RSUs have limited computing resources and battery power, the high energy consumption typically associated with blockchain must be taken into account. Techniques such as reducing consensus frequency or using energy-efficient consensus protocols could help to alleviate these concerns.

6.5. Consequences Under Data Protection Law

Although the system has implemented strong encryption for secure communication, data protection remains a major problem. As vehicles are constantly exchanging sensitive data such as location, speed, and identity, it must be ensured that this information is not misused or shared. Future work will need to address these privacy challenges more comprehensively, possibly through advanced privacy-preserving techniques such as zero-knowledge proofs (ZKPs) [31] or homomorphic encryption.
In addition, the use of anonymization to protect vehicle identity on the blockchain network could be further refined. Finding the right balance between maintaining transparency and protecting individual privacy will be crucial for both user acceptance and regulatory compliance.

6.6. Novelty and Differentiation

The proposed framework introduces several key innovations and enhancements over existing studies, addressing critical gaps in the literature and practical implementations of vehicular communication systems.
  • Enhanced security with smart contracts and PBFT consensus-unlike prior works that rely on centralized or static mechanisms for data validation and node authentication, our framework employs smart contracts and the Practical Byzantine Fault Tolerance (PBFT) consensus algorithm to ensure secure and decentralized operations.
  • Modularity and interoperability with existing V2X standards-a unique aspect of the proposed framework is its modular design, which supports interoperability with widely adopted V2X standards, such as ITS-G5 and Cellular-V2X (C-V2X).
  • Differentiation from existing RSU-based approaches-while existing works propose RSU-based architectures for forwarding data, they primarily focus on centralized data management or anomaly detection. In contrast, our framework achieves the following:
    –
    Distributes computation and decision-making across a decentralized blockchain network, eliminating single points of failure.
    –
    Utilizes RSUs not only as data aggregation points but also as lightweight blockchain clients. RSUs participate in the consensus process, validate transactions locally, and reduce latency in real-time data exchanges.
    –
    Integrates smart contract-based authentication directly at the RSU level, ensuring that only verified nodes can contribute to the network.
  • Real-world considerations-existing studies often evaluate their frameworks under idealized or static conditions, limiting their applicability in real-world scenarios. The current research achieves the following:
    –
    Includes simulations with dynamic vehicular mobility, network jitter, and latency variability, highlighting the framework’s robustness under real-world constraints.
    –
    Demonstrates how the modular design and decentralized consensus can adapt to diverse conditions, such as varying traffic densities and communication delays.

7. Future Work

For future work, we would like to expand our current research by addressing several key areas to further improve the performance, scalability, and security of the blockchain-based vehicle communication system.
  • Real-world testing: While our current simulations provided valuable insights into system behavior in attack scenarios, future efforts will focus on transitioning from simulations to real-world testing. Deploying the system in real vehicle environments will allow us to validate its performance and robustness under dynamic conditions, taking into account real-time factors such as network variability, physical obstacles, and hardware limitations.
  • Optimization for large-scale networks: As the number of connected vehicles and RSUs grows in future smart cities, scalability remains a critical challenge. In the future, we plan to explore more advanced optimization techniques, such as dynamic node allocation or sharding, to efficiently handle larger networks without overwhelming the capacity of the blockchain.
  • Energy efficiency improvements: Blockchain operations, especially in resource-constrained environments such as vehicles, can consume significant amounts of energy. We aim to investigate energy-efficient consensus algorithms or adaptive energy saving strategies, such as the integration of localized fog or edge computing solutions, to reduce the overall energy demand of the system.
  • Improved data protection measures: While we have implemented encryption and data anonymization techniques, in the future, we will explore even more advanced privacy-preserving methods, such as homomorphic encryption or secure multi-party computation [32]. These methods could further protect sensitive information without compromising the security or functionality of the system.
  • Advanced security mechanisms: We intend to enhance the security architecture by integrating machine learning-based intrusion detection systems (IDSs). These systems could dynamically detect and prevent novel or evolving attacks that could bypass traditional security mechanisms.
  • Evaluation of hybrid consensus models: PBFT performed well in our simulations, but exploring hybrid consensus models that combine PBFT with proof-of-stake (PoS) or other consensus algorithms could provide even better scalability and fault-tolerance, especially in environments with rapidly changing network topologies such as vehicular networks.

8. Conclusions

In sum, our research shows that a blockchain-based system can significantly improve the security and reliability of vehicular communication networks by effectively defending against common cyberattacks such as Sybil and denial of service (DoS) attacks. The implementation of the Practical Byzantine Fault Tolerance (PBFT) consensus method has demonstrated high fault tolerance by maintaining consensus and data integrity even in the presence of malicious nodes. The simulations provided a clear indication that our approach successfully ensures secure and robust vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication while demonstrating the importance of edge processing and scalability measures for managing network loads and ensuring timely communication.
However, our results also highlight the trade-offs between security and performance, especially in terms of latency and scalability in large networks. The increased latencies under certain conditions indicate that further optimizations are needed, especially for real-time applications in vehicles where fast data exchange is crucial. The fact that we rely on simulations also emphasizes the need for real-world testing to validate these results under dynamic conditions.
Overall, this research provides a fundamental framework for the use of the blockchain in vehicle networks and paves the way for further advances in secure autonomous vehicle systems. Future work will focus on real-world deployment, improved privacy measures, energy efficiency, and exploring hybrid consensus mechanisms to address the identified limitations. By further improving and expanding this system, we aim to contribute to the development of secure, scalable, and efficient communication networks for the evolving landscape of autonomous vehicles and smart cities. Overall, this research provides a fundamental framework for the use of the blockchain in vehicle networks, and the study paves the way for further advances in secure autonomous vehicle systems. Future work will focus on real-world deployment, improved privacy measures, energy efficiency, and exploring hybrid consensus mechanisms to address the identified limitations. By further improving and expanding this system, we aim to contribute to the development of secure, scalable, and efficient communication networks for the evolving landscape of autonomous vehicles and smart cities.

Author Contributions

Conceptualization, S.I., C.C.P. and C.P.; methodology, C.P.; software, S.I. and C.C.P.; validation, C.P.; resources, S.I.; data curation, C.C.P.; writing—original draft preparation, S.I. and C.C.P.; writing—review and editing, C.P.; visualization, S.I. and C.C.P.; supervision, C.P. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The study does not report any data.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Paduraru, C.; Jianu, A.; Stefanescu, A. Blockchain for Artificial Intelligence: An Industry and Literature Survey. In Proceedings of the 18th International Conference on Software Technologies (ICSOFT), Rome, Italy, 10–12 July 2023; pp. 135–150. [Google Scholar]
  2. Chen, L.; Zhao, Y.; Wang, M. A Secure and Efficient Blockchain-Based System for Vehicle-to-Everything (V2X) Communication. IEEE Trans. Intell. Transp. Syst. 2022, 23, 1–14. [Google Scholar]
  3. Rashid, M.; Alam, S.; Khan, M. Enhancing Security in VANETs with Blockchain and PBFT Consensus Mechanism. Comput. Netw. 2021, 184, 107644. [Google Scholar]
  4. Huang, J.; Wu, F.; Zhang, H. Blockchain-Based Decentralized Data Integrity Protocol for Autonomous Vehicles. IEEE Access 2023, 11, 7509–7522. [Google Scholar]
  5. Zhu, H.; Huang, X.; Zhang, W. Integration of Blockchain and AI in Autonomous Vehicle Safety. Int. J. Distrib. Sens. Netw. 2023, 19, 2023. [Google Scholar]
  6. Smith, A.; Jones, K.; Lee, D. Blockchain and Smart Contract Applications in Transportation Systems. Transp. Sci. 2023, 54, 130–144. [Google Scholar]
  7. Ramasamy, K.; Subramanian, D. Blockchain-Based Privacy Solutions in Vehicular Ad-Hoc Networks (VANETs). Comput. Secur. J. 2022, 108, 102456. [Google Scholar]
  8. Patel, S.; Kumar, A.; Verma, P. Implementation of Smart Contracts in AV Communication for Authentication and Privacy. J. Automot. Secur. 2023, 48, 102–118. [Google Scholar]
  9. Park, D.; Shin, J.; Lee, Y. Mitigating Sybil Attacks in AV Communication Networks via Blockchain. IEEE Veh. Technol. J. 2023, 13, 51–63. [Google Scholar]
  10. Chao, Y.; Lin, Z.; Wu, L. Smart Contracts for Autonomous Vehicle Interactions: A Case Study on Data Privacy. Int. J. Blockchain Res. 2023, 7, 289–304. [Google Scholar]
  11. Jeon, H.; Kim, Y.; Choi, M.; Park, D.; Son, S.; Lee, J.; Choi, G.; Lim, Y. CARLA Simulator-Based Evaluation Framework Development of Lane Detection Accuracy Performance Under Sensor Blockage Caused by Heavy Rain for Autonomous Vehicle. IEEE Robot. Autom. Lett. 2022, 7, 9977–9984. [Google Scholar] [CrossRef]
  12. Mohammed, R.; El-Sayed, A.; Lin, M. Performance Evaluation of PBFT and Hybrid Consensus in IoT Networks. IEEE Trans. Internet Things 2023, 7, 3023–3035. [Google Scholar]
  13. Rao, P.; Patel, S.; Gupta, M. Securing V2X Communication with Blockchain-Based Consensus Algorithms. J. Emerg. Transp. Technol. 2024, 19, 68–79. [Google Scholar]
  14. Kumar, A.; Sharma, P.; Gupta, R.; Singh, M. Enhancing Autonomous Vehicle Network Security Using Byzantine Fault Tolerance and Blockchain. Int. J. Intell. Syst. 2024, 9979371. [Google Scholar] [CrossRef]
  15. Wang, Y.; Zhao, L.; Chen, Z. Fog Computing in Autonomous Vehicles: Blockchain-Enabled Resource Management. Veh. Netw. J. 2022, 12, 200–215. [Google Scholar]
  16. Lee, J.; Park, S. Privacy Management in Vehicular Networks: A Smart Contract Approach. IEEE Veh. Commun. 2023, 17, 98–109. [Google Scholar]
  17. Tan, Y.; Li, J. Blockchain-Driven Solutions for Privacy and Security in Vehicular Networks. J. Comput. Telecommun. 2024, 60, 120–134. [Google Scholar]
  18. Liu, Z.; Zhao, M.; Xu, F. Leveraging Elliptic Curve Cryptography in AV Blockchain Systems for Efficient Authentication. Cybersecur. J. 2022, 13, 115–129. [Google Scholar]
  19. Davies, T.; Brown, R.; White, S. Energy-Efficient Consensus Mechanisms for Blockchain in Resource-Constrained Environments. Blockchain Sustain. Dev. 2023, 6, 54–72. [Google Scholar]
  20. Gao, Y.; Liu, Q.; Zhang, S. Scalability and Optimization in Blockchain for Autonomous Vehicular Networks. Smart Cities J. 2023, 5, 241–257. [Google Scholar]
  21. Zhao, Y.; Yu, H.; Liang, Y. Privacy-Preserving Distributed Collaboration Scheme for Autonomous Vehicles. IEEE Trans. Intell. Transp. Syst. 2024, 29, 201–214. [Google Scholar]
  22. Zhang, T.; Wang, J.; Han, L. Decentralized Authentication and Security Protocols for Autonomous Vehicle Networks. J. Netw. Secur. 2022, 12, 45–56. [Google Scholar]
  23. Van Brummelen, G. Heavenly Mathematics: The Forgotten Art of Spherical Trigonometry; Princeton University Press: Princeton, NJ, USA, 2013; ISBN 9780691148922. [Google Scholar]
  24. Wickremasinghe, G.; Frost, S.; Rafferty, K.; Sharma, V. Demonstrating a Hyperledger Fabric-based Blockchain with Knowledge Graphs for a Supply Chain Ecosystem. In Proceedings of the IEEE International Conference on Blockchain and Cryptocurrency (ICBC), Dublin, Ireland, 27–31 May 2024. [Google Scholar]
  25. Qi, X.; Chen, Z.; Zhang, Z.; Jin, C.; Zhou, A.; Zhuo, H.; Xu, Q. A Byzantine Fault Tolerant Storage for Permissioned Blockchain. In Proceedings of the ACM SIGMOD Conference, Xi’an, China, 20–25 June 2021; pp. 2770–2774. [Google Scholar]
  26. Wang, Z.-F.; Liu, S.-Q.; Wang, P.; Zhang, L.-Y. BW-PBFT: Practical byzantine fault tolerance consensus algorithm based on credit bidirectionally waning. Peer-to-Peer Netw. Appl. 2023, 16, 2915–2928. [Google Scholar] [CrossRef]
  27. Hankerson, D.; Vanstone, S.; Alfred Menezes, A. Elliptic Curve Cryptography Chapter in Encyclopedia of Cryptography, Security and Privacy; Springer: Berlin/Heidelberg, Germany, 2021; Available online: https://link.springer.com/referenceworkentry/10.1007/978-3-642-27739-9_245-2 (accessed on 28 October 2024).
  28. Liu, K.; Ma, R.; Yang, F. Edge Computing for Autonomous Vehicles: Enhancing Security and Reducing Latency. Int. J. Veh. Netw. Data Anal. 2023, 5, 95–109. [Google Scholar]
  29. Song, Q.; Wang, R.; Chen, L. Edge and Fog Computing in Autonomous Vehicles: Security and Privacy Aspects. J. Appl. Comput. 2023, 14, 102–118. [Google Scholar]
  30. Chang, Y.; Lee, C.; Yang, J. Optimizing Blockchain for Autonomous Vehicles Using Sharding Techniques. J. Distrib. Comput. 2024, 35, 445–462. [Google Scholar]
  31. Wong, P.; Li, X.; Huang, J. Enhancing Data Privacy in Autonomous Vehicle Networks through Zero-Knowledge Proofs. IEEE J. Priv. Data Secur. 2024, 16, 87–102. [Google Scholar]
  32. Kang, H.; Kim, S.; Woo, M. Secure Multi-Party Computation for Privacy in Vehicular Blockchain Networks. Int. J. Priv. Secur. 2023, 11, 78–91. [Google Scholar]
Futureinternet 16 00471 g001
Figure 1. Fog computing architecture with three layers: cloud layer (top) for centralized storage and large-scale processing; fog layer (center) for decentralized computation closer to data sources; and edge layer (bottom) for end devices like vehicles generating and collecting data.
Figure 1. Fog computing architecture with three layers: cloud layer (top) for centralized storage and large-scale processing; fog layer (center) for decentralized computation closer to data sources; and edge layer (bottom) for end devices like vehicles generating and collecting data.
Futureinternet 16 00471 g001
Futureinternet 16 00471 g002
Figure 2. Types of denial-of-service (DoS) attacks in vehicular communication systems: (a) packet flooding, (b) V2V radio jamming, and (c) V2I radio jamming.
Figure 2. Types of denial-of-service (DoS) attacks in vehicular communication systems: (a) packet flooding, (b) V2V radio jamming, and (c) V2I radio jamming.
Futureinternet 16 00471 g002
Futureinternet 16 00471 g003
Figure 3. Illustration of a Sybil attack in a vehicular network. A malicious vehicle (red car) creates multiple fake identities to inject false information into the network, misleading nearby vehicles and disrupting communication. Examples include spoofed hazard warnings, false location data, and network congestion, which compromise the safety and trustworthiness of vehicle-to-vehicle (V2V) communication.
Figure 3. Illustration of a Sybil attack in a vehicular network. A malicious vehicle (red car) creates multiple fake identities to inject false information into the network, misleading nearby vehicles and disrupting communication. Examples include spoofed hazard warnings, false location data, and network congestion, which compromise the safety and trustworthiness of vehicle-to-vehicle (V2V) communication.
Futureinternet 16 00471 g003
Table 1. Security simulations and feedback.
Table 1. Security simulations and feedback.
ScenarioDescriptionResult
Sybil attackA malicious node created 50 fake identities to manipulate vehicle-to-vehicle (V2V) communication100% success rate in detecting and blocking Sybil attacks, with a minimal impact on latency
DoS attackA malicious node flooded the network with 1000 fake transaction requests per second to overload the RSUs and blockchain nodes in vehicle-to-infrastructure (V2I) communication95% of malicious traffic was blocked, 100% of spoofed transactions were rejected, and throughput was reduced by 20% during the peak of the attack, with no legitimate transaction delayed by more than 50 ms
Table 2. Consensus mechanism evaluation feedback.
Table 2. Consensus mechanism evaluation feedback.
ScenarioDescriptionResult
Faulty node tolerance.Faulty or malicious nodes were introduced into the network, increasing the number from 0 to 5 out of 10 nodes in order to simulate Byzantine failures.The system successfully achieved consensus with up to three faulty nodes (in accordance with the PBFT tolerance of up to f = 1 / 3 Byzantine nodes). With four or more faulty nodes, consensus could not be achieved, and the network could not validate transactions.
Fault tolerance.The system was able to tolerate up to 3 faulty nodes in a 10-node network while maintaining consensus and data integrity.With one to two faulty nodes, the consensus time increased by 10%, and with three faulty nodes by 25%. With four faulty nodes, consensus was no longer possible.
Malicious node injection.Malicious nodes attempted to participate in consensus by sending corrupted or false blocks to disrupt the network.PBFT successfully identified and excluded 100% of the malicious nodes during the preparation and handover phase. The corrupted blocks were discarded, and the network reached a consensus with valid blocks.
Impact on latency.The presence of malicious nodes increased the consensus time.The Consensus time increased by 15% due to the presence of malicious nodes, but the system remained stable and functional.
Table 3. Comparative analysis: idealized vs. dynamic conditions.
Table 3. Comparative analysis: idealized vs. dynamic conditions.
MetricIdealized ConditionsDynamic Conditions (Mobility & Jitter)
Sybil attack detection rate100%97%
DoS attack mitigation rate95%90%
Consensus latency100 ms120 ms (average), peaks of 150 ms
Throughput reduction (DoS)20%25%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Iordache, S.; Patilea, C.C.; Paduraru, C. Enhancing Autonomous Vehicle Safety with Blockchain Technology: Securing Vehicle Communication and AI Systems. Future Internet 2024, 16, 471. https://doi.org/10.3390/fi16120471

AMA Style

Iordache S, Patilea CC, Paduraru C. Enhancing Autonomous Vehicle Safety with Blockchain Technology: Securing Vehicle Communication and AI Systems. Future Internet. 2024; 16(12):471. https://doi.org/10.3390/fi16120471

Chicago/Turabian Style

Iordache, Stefan, Catalina Camelia Patilea, and Ciprian Paduraru. 2024. "Enhancing Autonomous Vehicle Safety with Blockchain Technology: Securing Vehicle Communication and AI Systems" Future Internet 16, no. 12: 471. https://doi.org/10.3390/fi16120471

APA Style

Iordache, S., Patilea, C. C., & Paduraru, C. (2024). Enhancing Autonomous Vehicle Safety with Blockchain Technology: Securing Vehicle Communication and AI Systems. Future Internet, 16(12), 471. https://doi.org/10.3390/fi16120471

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop