SMAD-LDS: enhanced secure message authentication and dissemination with lightweight digital signature in the Internet of Vehicles

Introduction

Nowadays, human life can’t be imagined without vehicles. However, the rise in traffic accidents is a significant issue with vehicles and the transportation system. According to the World Health Organization (WHO), about 1.3 million people annually die due to road traffic crashes. Moreover, traffic crashes in most countries cost around 3% of the gross domestic product (World Health Organization, 2023). There are many reasons for these accidents, such as driver mistakes, lack of awareness, insufficient traffic signs, and bad road design. These accidents highlight the urgent need to find solutions for saving human lives. The intelligent transportation system (ITS) may be the solution to these issues. The objectives of the ITS are to enhance road safety and alleviate congestion through effective traffic flow management.

VANET is a vehicular ad hoc network that allows wireless communication between vehicles and infrastructure along the road. VANET nodes possess specific components that facilitate the collection of road information. These components include sensors, cameras, GPS, and omnidirectional antennas (Prasan & Murugappan, 2016; Siddiqui, Khaliq & Pannek, 2017). VANETs play a crucial role in the ITS by enabling vehicles to act as mobile sensors and communication nodes, collecting, disseminating, and routing information about traffic conditions.

VANET has three main components, such as the on-board unit (OBU), roadside unit (RSU), and trusted authority (TA). OBU is a wave device mounted on each vehicle. It is responsible for exchanging information between vehicles and RSUs via dedicated short-range communication (DSRC) (Azam et al., 2021). RSUs are fixed units installed by the TA in dedicated locations along the road. All RSUs along the road can communicate with each other (Shetty & Manjaiah, 2022). The TA manages the entire VANET system and is responsible for the OBUs and RSUs.

Traditional VANETs primarily utilized vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication models (Afzal & Kumar, 2020). However, recent advancements necessitate additional models, including vehicle-to-sensor (V2S), vehicle-to-pedestrian (V2P), and vehicle-to-network (V2N) communication (Hakimi et al., 2021; Yoshizawa et al., 2023). To accommodate these evolving needs, a new VANET architecture, the Internet of Vehicles (IoV), has been proposed. The IoV enables vehicles to communicate with devices, people, and diverse networks, thereby enhancing the security and safety of intelligent transportation systems (ITS) (Verma & Sharma, 2021; Sharma & Sharma, 2020).

There are many distinguishing characteristics of the IoV, such as high and predictable mobility of vehicles, frequently disconnected networks, geographically constrained, highly dynamic topology, large network scale, sufficient energy (no power constraints), high computational resources, real-time communication, variable density, integration with other networks, and diverse applications (Hasrouny et al., 2017; Hozouri et al., 2023).

The IoV has the potential to enable a diverse range of applications, which can be broadly categorized into two main groups: safety-related and entertainment-related (Alalwany & Mahgoub, 2024; Tufail et al., 2021; Elsagheer Mohamed et al., 2022). Below are some examples:

• Collision avoidance: This system warns the driver of an impending collision with a vehicle in front by using sensors like radar and cameras. The IoV enhances this capability by enabling real-time data sharing about the leading vehicle’s braking status with following vehicles, providing earlier and more precise alerts.

• Blind spot detection (BSD): This system detects vehicles in the driver’s blind spots. The IoV enhances this feature by enabling data exchange about vehicles in the blind spots of nearby cars, extending the detection range, and significantly improving situational awareness.

• Road hazard warnings: This application allows vehicles to report road dangers like potholes, accidents, or slippery conditions to a central network. This information is then relayed to other vehicles in the vicinity, enabling drivers to take necessary precautions. Real-time updates on road conditions significantly improve driving safety.

• Driver drowsiness detection: This application can play a crucial role in mitigating the risks of driver exhaustion by using sensors to monitor driver behavior, such as eye tracking and steering patterns. Upon detecting signs of drowsiness, these applications can issue timely alerts or recommend rest stops, thereby enhancing road safety.

• Real-time traffic information for navigation: IoV plays a key role in traffic management by delivering real-time traffic information to navigation systems. This allows drivers to select the most efficient routes, avoiding congestion and reducing travel time.

• Connected car apps: IoV enables a wide range of connected car apps that provide information about nearby points of interest, restaurants, gas stations, and parking availability. These apps enhance travel convenience.

• Integration with smart home: These applications can connect with smart home devices, enabling drivers to manage home appliances remotely. This could include actions such as pre-heating the house or turning on lights before reaching home.

To ensure the safe and reliable operation of these applications, a robust and comprehensive security scheme is necessary. This scheme must fulfill all necessary security requirements, such as authenticity, confidentiality, anonymity and untraceability, integrity, non-repudiation, and resistance against well-known attacks (Yu et al., 2020). Moreover, this scheme should operate swiftly with low communication and computation overhead due to the stringent time constraints in most IoV applications. Without a robust scheme, attackers could physically or financially harm drivers and access sensitive information.

The security of the IoV depends critically on ensuring message authentication and data integrity. Verifying the authenticity of the message originator is essential, as the trustworthiness of information depends on the sender (Siddiqui et al., 2021; Hbaieb, Ayed & Chaari, 2022). The absence of effective and robust vehicle authentication in the IoV can lead to significant and varied problems, such as:

• The safety of critical vehicle applications, including collision avoidance, lane change warnings, and pedestrian detection, is threatened when malicious vehicles disrupt the necessary V2X (Vehicle-to-Everything) communication.

• Malicious vehicles could be able to send false routing information, creating chaos and increasing the risk of accidents

• Malicious vehicles can compromise the privacy of drivers and passengers. It can access sensitive driver information, such as personal data and financial information.

• The economic impact of malicious vehicles could be substantial, including damage to vehicles and infrastructure, increased insurance costs, and the potential for loss of life.

• A malicious vehicle will be able to impersonate another legitimate vehicle or even an RSU.

Nevertheless, implementing effective message authentication in IoV faces significant challenges, including computational and communicational limitations, real-time constraints, network scalability, diverse security threats, privacy concerns, and key management complexities. To overcome these challenges, a robust and efficient authentication scheme must be developed. This article proposes a secure message authentication and dissemination scheme with a lightweight digital signature (SMAD-LDS), which is designed for low computation and communication overhead. The main contributions of the SMAD-LDS scheme are as follows:

• The proposed scheme provides a novel lightweight message signature technique to meet the strict real-time message dissemination requirements in the IoV. This innovative technique reduces computation and communication overheads for fast authentication, ensuring timely message processing and delivery.

• The proposed scheme comprehensively addresses key security objectives. It provides message authentication with privacy preservation, ensuring verifiable origin without revealing sender details. Furthermore, it guarantees message integrity and non-repudiation, preventing unauthorized modifications, and providing proof of origin. Message confidentiality is maintained, protecting content from unauthorized access. Additionally, message freshness mitigates replay attacks by verifying message recency. Finally, access control restricts dissemination to authorized entities.

• This proposed scheme provides strong protection against a variety of common threats. It effectively safeguards against message tampering, ensuring the integrity of data. Additionally, it prevents impersonation, allowing only authorized entities to participate. Confidentiality is maintained, protecting against eavesdropping. The system also counters fake RSU attacks, stopping malicious entities from impersonating an RSU. Finally, it prevents replay attacks, ensuring that reused messages do not undermine security.

• The performance analysis demonstrates that our scheme significantly enhances both computation and communication overheads compared to existing state-of-the-art schemes. Additionally, experimental results show that the proposed scheme reduces computation overhead by approximately 94% and communication overhead by around 70% relative to other schemes.

This article is organized as follows: “Related Work” reviews recent related work. “The Proposed SMAD-LDS Scheme” presents a detailed explanation of the proposed SMAD-LDS scheme. “Performance Analysis” evaluates the scheme’s performance. “Analysis of Security Requirements and Attack Countermeasures for the SMAD-LDS Scheme” examines the scheme’s security, requirements, and countermeasures. Finally, “Conclusions and Future Works” concludes the article and outlines directions for future research.

Related work

Authentication is considered the first step of security in the IoV. There are two phases of authentication: the message authentication phase and the data authentication phase. In the IoV, when a vehicle becomes a part of a network, it must authenticate itself at the TA. Consequently, the TA can differentiate between legitimate and malicious vehicles. In addition, when a message is transmitted between two nodes, this message must be authenticated to ensure its integrity (Talpur & Gurusamy, 2021).

Numerous schemes exist for message authentication with privacy preservation in the IoV, as depicted in Fig. 1. All researchers in the literature are focused on enhancing cryptography algorithms, verification processes, or signature processes. In this section, we briefly present a few existing message authentication schemes.

Jahanian, Amin & Jahangir (2015) proposed a TESLA protocol that first broadcasts the encrypted message with its message authentication code (MAC). Then, after a specified time, broadcast the encryption key to allow receivers to verify the message’s authenticity. This scheme is not suitable for safety applications that have critical time constraints. TESLA++ is an enhanced version of TESLA, offering improved efficiency and security. It operates by initially broadcasting a MAC followed shortly after by the complete message and the corresponding authentication key (Manvi & Tangade, 2017).

Lim & Manivannan (2016) proposed an efficient authentication protocol. This protocol depends on the vehicle-to-infrastructure (V2I) communication model. Moreover, they used the Diffie–Hellman (DH) protocol to establish and share a symmetric key. They use both symmetric key and asymmetric key cryptography algorithms. Liu et al. (2016) enhanced the message dissemination technique of Afzal & Kumar (2020) by using an aggregate message authentication code instead of the onion signature.

Ahmed, Mohamed & Sadek (2017) proposed a low computation overhead protocol that uses CRT-RSA cryptography algorithm instead of using the traditional RSA. This algorithm can decrypt the message faster than the traditional RSA. Moreover, this protocol generates a shared symmetric key without using the well-known Diffie-Hellman protocol, which produces more computation overhead at each node. Mohamed, Ahmed & Sadek (2021) enhanced the authentication scheme of Islam (Ahmed, Mohamed & Sadek, 2017) by combining symmetric and asymmetric cryptography techniques efficiently.

Jiang, Ge & Shen (2020) proposed an anonymous authentication scheme based on a group signature (AAAS). This protocol combines the group signature and pseudonym mechanism to prevent a single authority from figuring out the real identity.

Goudarzi et al. (2022) proposed a secure and privacy-preserving authentication scheme. This scheme depends on a fog server to leverage the system throughput by moving part of the computational power to the edge of the network. Moreover, it utilizes the quotient filter (QF) to address node authentication and elliptic curve cryptography (ECC) for message authentication.

Mistareehi & Manivannan (2022) proposed a low-overhead message authentication and dissemination scheme. The RSU’s region in this scheme is divided into sub-regions (groups). The RSU appoints group leaders who oversee each sub-region. The group leader is responsible for gathering messages from vehicles, verifying their authenticity, compiling them, and sending them to the RSU. This scheme reduces the message authentication overhead at the RSU.

Li, Yang & Chen (2023) proposed a privacy authentication scheme with exculpability. This model applied a double-key approach to achieving trust communication between vehicles and the RSU. Moreover, it combined the group-based methods with identity-based methods. This combination enhanced the message authentication process.

Genc et al. (2023) proposed a novel identity-based privacy-preservation anonymous authentication scheme (NIBPA). The author focused on the vehicle-to-vehicle communication model. Moreover, this scheme utilized pairing-free elliptic curve cryptography (ECC) to minimize the computation cost. In addition, they applied batch message verification to improve the overall performance.

Naskar et al. (2024) introduced a distributed authentication and revocation scheme for Vehicle-to-Infrastructure (V2I) communication. Their protocol employs a hybrid approach, combining symmetric and asymmetric cryptography. Communication occurs in discrete time sessions, or epochs, each secured by a unique epoch key. Within each epoch, authenticated message dissemination among verified vehicles is permitted. Additionally, vehicles can report misbehavior to Roadside Units (RSUs).

Gu et al. (2022) proposed a multi-fog-based authentication architecture designed to reduce overall delay by enabling vehicles to be verified by fog nodes. However, their protocol’s centralized nature creates a communication bottleneck at the Certificate Authority (CA).

Cui et al. (2020) proposed a secure, privacy-preserving authentication scheme where each vehicle receives an internal pseudo-identity (IPID) from the Trusted Authority (TA), derived from its real identity. The vehicle stores this IPID, along with a self-chosen encryption key, in a tamper-proof device (TPD). Following authentication with the TA, the vehicle generates a public pseudo-identity (PPID) used for signing transmitted messages, which are then verified by the receiver. While the scheme enhances security through periodic updates to both the IPID and encryption key, it may introduce challenges related to scalability and the management of pseudo-identities in large-scale networks.

Manivannan, Moni & Zeadally (2020) proposed a secure privacy-preservation authentication scheme utilizing ID-based cryptography to reduce concerns with certificate management. Moreover, it employs identity-based batch signature verification for efficient RSU verification of multiple signatures. However, this scheme is susceptible to various attacks, including replay attacks.

Chen & Chen (2021), proposed a certificateless aggregate signature (CLAS) scheme that simplifies the authentication process by eliminating certificate management. It offers conditional privacy, enabling vehicles to use pseudonyms for message signing to protect real identities while preserving traceability by a trusted authority. The aggregate signature mechanism further enhances verification efficiency by combining multiple signatures into a single compact signature. However, unlinkability is not a feature of this scheme.

Badshah et al. (2022) present an anonymous authenticated key exchange scheme for blockchain-enabled Internet of Vehicles in smart transportation (AAKE-BIVT). It utilizes lightweight encryption and physical unclonable function (PUF) to reduce computational overheads. In this scheme, vehicles are organized into clusters, and each cluster has a cluster head that communicates directly with the RSU. Moreover, it utilizes a blockchain for decentralized trust management. These techniques suffer from high computational overhead and are unsuitable for real-time applications.

To conclude, an efficient message authentication scheme should satisfy the privacy preservation condition. Moreover, it should incur low computation and communication overhead. In addition, it must resist different types of known and unknown attacks. Accordingly, this article presents an enhanced secure message authentication and dissemination scheme with a novel lightweight digital signature (SMAD-LDS).

The proposed SMAD-LDS scheme

This section provides a detailed description of the proposed lightweight message authentication and dissemination (SMAD-LDS) scheme.

The scheme is designed to minimize both the computational burden and communication costs while providing robust security. The SMAD-LDS employs a regional architecture wherein the entire area is divided into distinct regions (groups), each under the control of an RSU. The Vehicle-to-Infrastructure (V2I) is the main communication model. Therefore, all communications must be directed through the RSU. Consequently, the RSU is responsible for verifying the authenticity and integrity of each message before broadcasting it to the group. This communication model is more effective regarding the cost of communication since the RSU verifies the message originator’s authenticity only once.

The SMAD-LDS scheme

The SMAD-LDS scheme has four main phases: the Initiation phase, the Group Registration Phase, the Send/Receive a Group Message, and the Leave and Join another Group Phase.

Table 1 lists all notations used in this article.

Table 1:

Notations and abbreviations.

Notation	Description
$R$	An RSU
$V$	A Vehicle
$TS$	A Timestamp
$X_{pos},Y_{pos}$	The Location coordinates (X, Y)
$M$	A Message
$Cer{t}_V$	The Vehicle’s Certificate
$I{D}_v$	The Vehicle’s real identity
$PI{D}_v$	The Vehicle’s pseudo-identity
$H()$	Hash Function
$E()$	Encryption Function
$SIG$(M)	The message digital Signature
$P{K}_i$	The public key of entity i
$P{R}_i$	The private key of an entity i
${\delta }_i$	The entity’s secret value
${\delta }_G$	The Group’s secret value
$K_{VR}$	The shared symmetric key between a vehicle and RUS
$K_G$	The shared group key
$Ms{g}_{ID}$	The ID of the current group message
$SS{V}_i$	The ith Secret Seed Value

DOI: 10.7717/peerj-cs.2982/table-1

Phase 1: Initiation

In this phase, we generate all the required group keys, along with their corresponding chains of N secret values. This ensures that all necessary secret values for group communication are pre-generated and readily available.

In the SMAD-LDS scheme, the communication time within a group will be divided into sessions. Each session has its group key and a chain of N secret values. Each secret value in the chain will be used once. When all secret values are used, the current session will be closed, and a new session will be started. For each session, RSU will choose random bits to be used as a symmetric key. This key will be used as the group key $K_G$. Moreover, it will choose a random secret seed value $SS{V}_i$ to create a chain of N secret values, a chain of N hashes, as shown in Fig. 2. These secret values in the chain, from back to front, will be used in signing the group’s messages. These secret values will be used, in a novel and efficient way, to sign the group’s messages. This is a completely novel idea inspired by blockchain.

Phase 2: Registration

The SMAD-LDS scheme requires all vehicles to register with the RSU. This registration process prevents malicious vehicles from participating in group communication, including broadcasting and sending direct messages.

In the SMAD-LDS scheme, each RSU regularly broadcasts a beacon message every 100 ms. This message contains the $I{D}_R,TS,$ the RSU’s location coordinates $X_{pos},Y_{pos}$, and The ID of the last message sent to the vehicles into its group $Ms{g}_{ID}$.

(1) $$\begin{array}{rl}& m=TS,Xpos,Ypos,I{D}_R,Ms{g}_{ID}\\ & Sig=E\left(H\left(m\right),P{R}_R\right)\end{array}$$

When a vehicle enters a new region covered by an RSU, it will select a secret value ${\delta }_V$. Then, it encrypts both secret the value ${\delta }_V$ and its real identity $I{D}_v$using the public key of the RSU $P{K}_R$. After that, it generates a message containing the current timestamp, the vehicle’s location coordination $X_{pos},Y_{pos}$, the vehicle’s certificate $Cer{t}_V$, and the previously encrypted data. Finally, it signs the message using its private key $P{R}_V$ and sends it to the RSU.

(2) $$\begin{array}{rl}& m=TS,Xpos,Ypos,Cer{t}_V,I{D}_R,E\left(\left(I{D}_v\parallel {\delta }_V\right),P{K}_R\right)\\ & Sig=E\left(H\left(m\right),P{R}_V\right)\end{array}$$

When the RSU receives the registration message, it will check the message’s integrity and the vehicle’s authenticity. Upon authenticating the vehicle, it will generate a new shared symmetric key as $K_{VR}$ besides a new vehicle’s pseudo-identity $PI{D}_v$ as in Eqs. (3) and (4), respectively:

(3) $$K_{VR}=H\left(TS\parallel I{D}_v\parallel {\delta }_V\right).$$

(4) $$PI{D}_v=H\left(I{D}_v\parallel {\delta }_V\right).$$

The current group key $K_G$, the current group message ID ( $Ms{g}_{ID}$), and the current group’s secret value ${\delta }_G$ are encrypted utilizing the previously generated shared symmetric key $K_{VR}$. After that, it will generate a response message that contains $I{D}_R,TS,$ the RSU’s location coordinates $X_{pos},Y_{pos}$, and the previously encrypted data. Finally, it signs the message utilizing its private key $P{R}_R$. and sends it to the vehicle.

(5) $$\begin{array}{rl}& m=I{D}_R,TS,Xpos,Ypos,E\left(\left(K_G\parallel Ms{g}_{ID}\parallel {\delta }_G\right),K_{VR}\right)\\ & Sig=E\left(H\left(m\right),P{R}_R\right)\end{array}$$

When the vehicle receives the registration’s response message, it will check its integrity and authenticity. Then, it will calculate the shared symmetric key and its pseudo-identity as in Eqs. (3) and (4), respectively. After that, it will decrypt the encrypted data to get and store the current ID of the group message, the group’s secret value, and the group’s key. The message sequence is depicted in Fig. 3.

Phase 3: Send/receive messages

During this phase, the previously registered vehicles will be able to send and receive different types of messages. The SMAD-LDS scheme utilizes two distinct message types: group and one-to-one. As mentioned before, all messages in this scheme must be sent through the RSU.

A. Send and receive a group message

When a vehicle wants to send data to a group, it will encrypt it using the shared secret key $K_{VR}$. Then, it will create a new message that contains the encrypted data, the current timestamp, the vehicle’s location coordination $X_{pos},Y_{pos}$, and the vehicle’s pseudo-identity. Subsequently, a digital signature is appended to the message. Traditional message signing involves hashing the message content and encrypting the resulting hash with the originator’s private key. This approach incurs substantial computation overhead for both signing and verification. SMAD-LDS addresses this limitation by employing a lightweight signing mechanism. In SMAD-LDS, the message is signed by hashing the concatenation of the message and the vehicle’s secret value ( ${\delta }_v$). This signature technique provides a reduction in the computation overhead of signature generation and verification compared to traditional methods. Note that the vehicle’s secret value ( ${\delta }_v$) is shared only between the vehicle and the RSU. Finally, it sends the message with its signature to the RSU. The message signature, in this work, is created by hashing the message with a secret value. Hence, there is no need to encrypt it.

(6) $$\begin{array}{rl}& m=TS,Xpos,Ypos,PI{D}_V,I{D}_R,E\left(Data,K_{VR}\right)\\ & Sig=H\left(m\parallel {\delta }_v\right)\end{array}$$

When the RSU receives the message, it will check the message’s integrity and the vehicle’s authenticity. This check is achieved by hashing the message with the vehicle’s secret value ( ${\delta }_v$) and verifying it against the message signature. If it is not valid, it will just drop it. Otherwise, if it is valid, it will decrypt the encrypted data using the shared symmetric key $\left(K_{VR}\right)$. After that, it will encrypt it again by using the shared group key ( $K_G$). Next, it will create a message that contains the encrypted data, the RSU’s identity ( $I{D}_R$), the current timestamp $TS$, and the next group’s secret value ( ${\delta }_{G-1}$). After that, it will sign the message by hashing the concatenation of the message and the current group’s secret value ( ${\delta }_G$). Finally, it broadcasts the message with its signature to the group.

(7) $$\begin{array}{rl}& m=I{D}_R,TS,Ms{g}_{ID},E\left(Data,K_G\right),{\delta }_{G-1}\\ & Sig=H\left(m\parallel {\delta }_G\right)\end{array}$$

Upon receiving the broadcast message, each vehicle within the group performs the following: First, it verifies the message’s integrity and the RSU’s authenticity. This check is achieved by hashing the message with the stored group’s secret value ( ${\delta }_G$) and verifying it against the message signature. If it is valid, it will hash the received ( ${\delta }_{G-1}$) and then compare the result against the stored group’s secret value ( ${\delta }_G$). If it is not valid, it will drop the message. Otherwise, it will update the stored group’s secret value ( ${\delta }_G$) with the new received value ( ${\delta }_{G-1}$). Finally, it will decrypt the encrypted data using the shared group key ( $K_G$). The message sequence is depicted in Fig. 4.

B. Send and receive a one-to-one message

When a vehicle ( $V_1$) wants to send data to a vehicle ( $V_2$), it will encrypt it using the shared secret key $\left(K_{V_{1}R}\right)$. Then, it will create a new message that contains the encrypted data, the current timestamp, the vehicle’s location coordination $X_{pos},Y_{pos}$, the vehicle’s pseudo-identity ( $PI{D}_{V_1}$), and the destination vehicle’s pseudo-identity ( $PI{D}_{V_2}$). After that, it will sign the message by hashing the concatenation of the message and the vehicle’s secret value ( ${\delta }_{V_1}$). Finally, it sends the message with its signature to the RSU.

(8) $$\begin{array}{rl}& m=TS,Xpos,Ypos,PI{D}_{V_1},PI{D}_{V_2},I{D}_R,E\left(Data,K_{V_{1}R}\right)\\ & Sig=H\left(m\parallel {\delta }_{V_1}\right)\end{array}$$

When the RSU receives the message, it will check the message’s integrity and the vehicle’s authenticity. If it is valid, it will decrypt the encrypted data using the shared symmetric key $\left(K_{V_{1}R}\right)$. After that, it will encrypt it again by using the shared key between the RSU and the destination vehicle ( $K_{V_{2}R}$). Next, it will create a message that contains the encrypted data, the RSU’s identity ( $I{D}_R$), the current timestamp $TS$, and the pseudo-identity of both vehicles. After that, it will sign the message by hashing the concatenation of the message and the secret value of the destination vehicle ( ${\delta }_{V_2}$). Finally, it sends it to the destination vehicle.

(9) $$\begin{array}{rl}& m=TS,Xpos,Ypos,PI{D}_{V_2},PI{D}_{V_1},I{D}_R,E\left(Data,K_{V_{2}R}\right)\\ & Sig=H\left(m\parallel {\delta }_{V_2}\right)\end{array}$$

When the destination vehicle receives the message, it will verify the message’s integrity and the RSU’s authenticity. If the message is valid, it will be accepted; otherwise, it will be dropped. The message sequence is depicted in Fig. 5.

Phase 4: Group key update

In the SMAD-LDS scheme, the group key is shared among all vehicles within the same group. Thus, if the key is exposed, an attacker could access the group’s communications until the key is updated. To maintain robust security and ensure secure message dissemination within the group, the RSU regularly updates the group key. This process involves identifying the specific situations that trigger a group key update and then replacing the current key with a pre-generated one. In the SMAD-LDS scheme, the group key will be updated based on the following situations:

• Exceeding the number of encrypted/decrypted messages (N messages). In this case, the group key will be updated after all values in the current chain of secret values are used (end of session). As mentioned before, the RSU updates the active group key and the chain of secret values at the beginning of each session.

• Exceeding a specified time interval. In this case, the RSU will update the group key after a specified time interval, even if it is not used frequently. Moreover, the time interval is proportional to the key length.

• Detecting a malicious vehicle. If the RSU detects and revokes a malicious vehicle, it will update the group key to ensure secure message dissemination within the group.

There are two different methods to update the group key as follows:

• Key encapsulation method (KEM). In this method, the RSU will share the new group key by sending it individually to each vehicle within the group. Consequently, the RSU will encrypt the new key using the public key of the vehicle. This method is suitable for key updates when exceeding a specified time interval or detecting a malicious vehicle.

• Key derivation method (KDM). In this method, the new key is derived from the old one. The derivation is done by hashing the old group key with some input parameters, such as a counter or a label, to generate the new group key. This method is suitable for key updates when the number of encrypted/decrypted messages (N messages) is exceeded.

Performance analysis

As previously mentioned, the growing number of message transmissions in the IoV, driven by its diverse communication types, necessitates an efficient and timely message dissemination scheme within a potentially congested network. Therefore, minimizing computation and communication overheads is critical. This reduction alleviates traffic congestion, consequently enhancing network latency and success rates. Additionally, the scheme utilizes a lightweight technique, which ensures minimal memory usage and does not impose any significant memory overhead. Consequently, evaluating the effectiveness of any message dissemination scheme requires an examination of the delivery time, which is directly impacted by both types of overhead. This section will analyze both the computation and communication overheads incurred by the proposed SMAD-LDS scheme during each of its phases. Furthermore, the performance of the SMAD-LDS scheme will be compared against existing schemes in the literature. Both the proposed and existing schemes were implemented and tested on a Windows 10 machine with a 2.60 GHz Intel Core processor and 8 GB of RAM. The results demonstrate the efficiency of the SMAD-LDS scheme in terms of both computation and communication overhead.

Analysis of security requirements and attack countermeasures for the SMAD-LDS scheme

The performance section demonstrates that the SMAD-LDS scheme provides improved and lightweight message dissemination, reducing computation and communication overheads. This lightweight design enables the SMAD-LDS to handle the growing number of vehicles and devices without system overload, thus effectively mitigating scalability issues while maintaining both security and performance. This section analyzes both the security requirements and attack countermeasures in detail.

Conclusions and future works

The Internet of Vehicles (IoV) relies on various applications designed to enhance driver safety. Therefore, secure and rapid message authentication and dissemination are crucial. This article introduces an enhanced secure message authentication and dissemination with lightweight digital signature (SMAD-LDS) scheme for the IoV, designed for efficient message exchange between vehicles and RSUs by minimizing communication and computation overheads. A novel, lightweight digital signature technique is employed for increased efficiency. Results demonstrate a significant reduction in computation costs for SMAD-LDS compared to existing solutions: at least 46.8% in the registration phase and at least 94% in the message send/receive phase. Communication costs are also improved, with reductions of 28% and 70% in the registration and message send/receive phases, respectively. Additionally, SMAD-LDS satisfies key security requirements, including entity and data authenticity, privacy preservation, and data confidentiality. The scheme’s robustness against various attacks, such as modification, impersonation, eavesdropping, fake RSU, traceability, and replay attacks, is also validated.

Future works will concentrate on four key areas to enhance the SMAD-LDS scheme. The first area involves extending the scheme to support V2V direct communication, enabling more efficient information exchange among vehicles. The second area will address security concerns related to the single-point-of-failure issue and RSU compromise by enhancing the scheme’s robustness in such scenarios. The third area will explore the potential of integrating recent quantum cryptography algorithms into the IoV architecture and analyze their impact on security and performance. Finally, we will examine and analyze various existing techniques for detecting malicious vehicles and revoking their certificates.

Supplemental Information