TinyTwinGuard: a lightweight authentication scheme for digital twin edge networks using TinyJAMBU

Abstract

Digital Twin Edge Networks (DITEN) encounter significant challenges in balancing security and efficiency, as many existing authentication mechanisms either fail to deliver full security guarantees, such as perfect forward secrecy or protection against insider threats, or impose excessive overhead due to the use of ECC or RSA cryptographic primitives. To overcome these limitations, we introduce TinyTwinGuard (TTG), a lightweight authentication protocol designed specifically for DITEN environments and built upon the NIST-standardized TinyJAMBU Authenticated Encryption with Associated Data (AEAD) cipher. Using formal validation with the AVISPA framework and extensive experimental evaluation on Raspberry Pi 4B devices, we show that TTG satisfies a broad set of 13 security requirements relevant to DITEN, including user anonymity, resistance to privileged insider attacks, and perfect forward secrecy. Although TTG incurs slightly higher communication overhead than minimal hash-based approaches, with 4,896 bits compared to 2,560 bits, it achieves a 43.7% reduction in computational cost relative to ECC-based schemes and lowers energy consumption by 58.3%. Overall, these findings indicate that TTG achieves an effective balance between strong security and operational efficiency, making it well suited for industrial edge scenarios where both resilience and deployment practicality are essential.

Highlights

1 Introduction

The Internet of Things (IoT) signifies a revolutionary shift in technology, characterized by the pervasive interconnection of physical devices through the Internet (El-Hajj et al., 2017). These devices encompass a wide spectrum, from common household items to sophisticated industrial machinery, embedded with sensors and software that enable data collection and communication (El-Hajj et al., 2023). The exponential growth in IoT devices generates vast quantities of data, necessitating efficient methods for data processing and management (El-Hajj et al., 2019).

While IoT represents a broad paradigm of interconnected devices exchanging data over the internet, Digital Twin Edge Networks (DITEN) represent a specialized evolution tailored for industrial applications. Unlike traditional IoT, which primarily focuses on data collection and remote monitoring, DITEN integrates digital twin technology with edge computing to create virtual replicas of physical systems that operate in real-time.

The distinctions between IoT and DITEN are rooted in their scope, architecture, functionality, and security requirements. IoT broadly encompasses a wide array of applications, ranging from consumer devices to industrial sensors, whereas DITEN is specifically designed for industrial environments that demand real-time insights and system optimization. Architecturally, IoT often relies on centralized cloud processing, which can introduce latency and bandwidth issues. In contrast, DITEN leverages edge computing to process data locally, significantly reducing delays and optimizing resource usage. Functionally, while IoT emphasizes connectivity and data exchange, DITEN goes further by enabling real-time monitoring, predictive maintenance, and system optimization through digital twin simulations. Additionally, DITEN introduces unique security challenges, particularly in ensuring secure communication between physical devices, their virtual counterparts, and edge nodes—a concern that is less pronounced in general IoT applications.

These distinctions highlight how DITEN offers a more advanced framework for managing complex industrial systems, enhancing operational efficiency and decision-making capabilities. Edge computing plays a crucial role in this advancement by executing computations closer to the data source, minimizing latency, and reducing bandwidth consumption. This approach is especially valuable in scenarios requiring swift responses, such as industrial applications (Yu et al., 2017). The combination of IoT and edge computing has paved the way for frameworks like DITEN, which leverage digital twins to provide real-time insights and simulations (van der Wal and El-Hajj, 2022). When integrated with edge computing, digital twins offer enhanced operational efficiency, predictive maintenance, and system optimization (Liu et al., 2021). However, as data exchange between physical and digital components increases, ensuring robust security measures becomes critical to protect data integrity and confidentiality (Sai et al., 2024).

Despite their potential, DITEN faces significant challenges in implementing secure and efficient authentication mechanisms. Current cryptographic algorithms, such as ASCON (Dobraunig et al., 2021), AES (Abdullah, 2017), and ChaCha20 (Mahdi et al., 2021), present notable limitations in resource-constrained edge environments. While ASCON offers strong security features, it is computationally intensive and energy-consuming. AES, despite its widespread adoption, incurs high computational costs, particularly in constrained settings (Thakor et al., 2021). Similarly, ChaCha20, though efficient in software, demands substantial processing power in certain scenarios (Tian and Vassilakis, 2023). These inefficiencies hinder performance in edge computing environments, where minimizing latency and conserving resources are critical.

To address these gaps, lightweight cryptographic solutions tailored to the unique requirements of edge computing are increasingly needed. TinyJAMBU (Saha et al., 2020), a lightweight authenticated encryption algorithm, provides a promising alternative. Its design focuses on reducing computational complexity and energy usage, making it particularly well-suited for edge-computing scenarios (El-Hajj et al., 2025). By leveraging TinyJAMBU, we can enhance both the security and efficiency of authentication processes within DITEN.

We employ a comprehensive methodology to achieve these objectives, including theoretical analysis, rigorous security evaluations, and empirical assessments. Our theoretical analysis highlights the efficiency of TinyJAMBU in terms of computational and energy savings, while our empirical evaluations provide evidence of TTG’s superior performance compared to existing algorithms. These findings support the practical adoption of TTG in industrial settings.

The remainder of this paper is organized as follows: Section 2 reviews relevant related work in authentication schemes for edge and IoT environments. Section 3 introduces the foundational concepts and cryptographic mechanisms underpinning TTG, including TinyJAMBU, fuzzy extractors, and extended Chebyshev polynomials. Section 4 outlines the system and adversary models, detailing the DITEN architecture and threat assumptions. Section 5 presents the proposed TTG authentication scheme, covering system setup, registration, user login, mutual authentication, and key agreement phases. Sections 6, 7 provide a comprehensive formal and heuristic security analysis and a detailed performance evaluation, respectively. Finally, Section 8 concludes the paper and outlines directions for future research.

2 Related work

Authentication and Key Agreement (AKA) schemes for IoT systems have evolved along three primary research trajectories: multi-factor authentication, lightweight cryptography, and edge-centric security frameworks. In this section, we analyze representative works across these categories, emphasizing their limitations in addressing the unique challenges of DITEN.

2.1 Categorization of existing schemes

2.2 Multi-factor authentication

Multi-factor authentication schemes aim to enhance security by combining multiple credentials, such as passwords, biometrics, and smart cards. Singh et al. (Singh and Kumar, 2022) proposed a multi-modal biometric scheme that uses iris scans and fingerprint templates to achieve 99.4% authentication accuracy. However, their approach requires 1.2 s for feature extraction and 1.8 MB of storage per user, making it unsuitable for DITEN’s real-time control loops and resource-constrained devices. Similarly, Wang et al. (Wang et al., 2020) introduced a three-factor authentication scheme using elliptic curve cryptography (ECC). While their approach enhances resilience against stolen smart card attacks, it suffers from high computational costs, with ECC scalar multiplication operations taking approximately 9.35 ms each, leading to a total latency of 320 ms—far exceeding DITEN’s 30 ms latency budget for digital twin synchronization. These schemes highlight the trade-offs between security and efficiency, often prioritizing the former at the expense of real-time performance.

2.3 Lightweight cryptographic protocols

Lightweight cryptographic protocols focus on minimizing computational overhead and energy consumption, making them suitable for resource-constrained IoT devices. Chen et al. (Chen and Li, 2021) designed a sponge-based protocol that consumes only 0.9 mJ per authentication. Despite its energy efficiency, the scheme is vulnerable to side-channel attacks due to timing information leakage during sponge absorption phases. Vinoth et al. (Vinoth et al., 2020) optimized AES-128 for data-type-aware key scheduling, reducing energy consumption by 22%. However, their approach does not support dynamic adjustments to varying device densities, a key requirement in DITEN. Additionally, their scheme incurs a latency of 45 ms, which is too high for real-time digital twin applications. These examples illustrate the challenges of balancing lightweight design with robust security and real-time performance.

2.4 Edge-centric frameworks

Edge-centric frameworks are tailored for IoT and edge computing environments, aiming to reduce latency and improve scalability. Almasri et al. (Almasri and Al-Kuwari, 2023) implemented a fog-layer authentication framework using Pedersen commitments, achieving a latency of 28 ms. While their approach demonstrates good performance, it relies on static trust anchors, which conflict with DITEN’s need for dynamic device onboarding and real-time synchronization. Alsahlani et al. (Alsahlani and Popa, 2021) proposed a hash-based scheme for IoT cloud-based environments, achieving a latency of 32 ms. However, their design lacks replay attack resistance and fails to integrate with digital twin technologies, limiting its applicability in DITEN. These schemes highlight the need for edge-centric solutions that can dynamically adapt to the evolving requirements of industrial IoT networks.

2.5 Blockchain-assisted protocols

Blockchain-assisted protocols leverage decentralized architectures to enhance security and transparency in IoT systems. Chaudhry et al. (Chaudhry and Al-Turjman, 2022) introduced a decentralized authentication scheme leveraging blockchain technology. While their approach prevents single-point failures, the consensus mechanism introduces a delay of 420 ms, making it impractical for real-time DITEN applications. Zhang et al. (Zhang and Zheng, 2023) combined blockchain with edge nodes to create a hybrid protocol. Their solution achieves a latency of 38 ms through Merkle proof verification. However, this latency still exceeds the 20 ms response time required for digital twin control cycles in DITEN. These examples underscore the challenges of integrating blockchain into resource-constrained and time-sensitive environments.

2.6 Performance comparison

Table 1 benchmarks TTG against prior works using metrics critical for DITEN. Tests were conducted on Raspberry Pi 4B with 100–1,000 simulated edge devices which will be highlighted in details in Section 7.

2.7 Gaps addressed by TTG

By unifying these advances, TTG delivers a tailored solution that meets DITEN’s stringent latency, scalability, and security demands, ensuring robust authentication for industrial digital twins.

3 Preliminaries

The preliminary section of our paper provides the foundational concepts and cryptographic mechanisms essential for understanding our proposed TinyJAMBU-based authentication scheme. It introduces the TinyJAMBU encryption algorithm, fuzzy extractor (Li et al., 2017) for biometric verification, and extended Chebyshev polynomials (Mason and Handscomb, 2002), which form the basis of our security framework. This section ensures that readers are familiar with these key algorithms and mathematical models, setting the stage for the detailed security analysis and implementation in the subsequent sections of the paper (Loos et al., 2003).

3.1 TinyJAMBU cryptography

The TinyJAMBU Cryptography subsection provides an in-depth overview of TinyJAMBU, a lightweight authenticated encryption algorithm designed for resource-constrained devices. It discusses its key structure, encryption, and decryption processes, and mathematical foundations, which are crucial for its integration into our proposed authentication scheme (Radhakrishnan et al., 2024).

3.1.1 Overview of TinyJAMBU

TinyJAMBU is a lightweight authenticated encryption algorithm specifically tailored for resource-constrained devices, including IoT devices and embedded systems. It is part of the NIST Lightweight Cryptography Standardization project (El-Hajj and Fadlallah, 2022). TinyJAMBU is optimized for both speed and low-energy consumption, making it suitable for environments with limited computational power and memory resources. The algorithm operates with a 128-bit key and a 128-bit Initialization Vector (IV) (Rushad et al., 2022).

The core structure of TinyJAMBU is composed of multiple rounds of a Non-Linear Feedback Shift Register (NLFSR) (Dubrova et al., 2008), which forms the backbone of its encryption process. The key size is 128 bits, and the IV size is also 128 bits. During encryption, the plaintext is processed in blocks, and bitwise operations are applied repeatedly over a series of rounds. The steps involved in both encryption and decryption follow a symmetrical process, allowing for high computation efficiency (Alsahli et al., 2022).

TinyJAMBU operations are primarily based on bitwise operations and a lightweight round function (Zhang et al., 2014). The core operations include XORs, ANDs, and shifts applied in multiple rounds. Depending on the security level, TinyJAMBU uses 384, 640, or 1,024 rounds to ensure encryption robustness (Cheng et al., 2023a). During encryption, the plaintext is processed with the key, nonce, and Associated Data (AD) through these rounds to produce the ciphertext, while decryption reverses the process to recover the original plaintext. The lightweight nature of these operations makes TinyJAMBU highly efficient for resource-constrained devices (Chan and Rogaway, 2019).

3.1.2 Mathematical foundation

The key operations in TinyJAMBU include bitwise XOR operations (Gao et al., 2021), left shifts (Dunkelman et al., 2022), and additions modulo (Alsahli et al., 2022). These operations contribute to the diffusion and confusion properties essential for security. The number of rounds can be 384, 640, or 1,024, depending on the desired security level (Windarta et al., 2023). More rounds offer higher security but at the cost of increased computational effort and power consumption (Datta et al., 2022).

In the encryption process, TinyJAMBU takes the key , nonce , associated data , and plaintext as inputs and produces the ciphertext . During decryption, it reverses the process using the same inputs to recover the original plaintext.

3.2 Fuzzy extractor

The correctness of the reproduced is based on the proximity of the noisy template to the original template , as determined by the Hamming distance.

3.3 Extended chebyshev polynomial

Chebyshev polynomials (Bergamo et al., 2005) are a recognized class of chaotic mappings. In their extended form, Chebyshev polynomials are defined by a recurrence relation. Let be an integer and a variable in the interval . The extended Chebyshev polynomial is specified by the following recurrence relation (Equations 5–7):

The extended Chebyshev polynomial adheres to the semi-group property, enabling it to be combined through composition expressed as Equation 8:

Furthermore, under modulo operations with a large prime number , the Chebyshev polynomial has a commutative property (Zhang et al., 2021) expressed as Equation 9:

Here, denotes a large prime number, while and represent two distinct large integers.

Furthermore, the extended Chebyshev polynomial can be represented using a square matrix rather than recursive relations, as demonstrated in the following matrix expressions (Equation 10):

The extended Chebyshev polynomials provide a rich mathematical framework that can be applied to cryptographic protocols requiring chaotic behavior and properties such as pseudo-randomness and unpredictability.

4 System models

In this section, we introduce the network model and the associated adversary model for our resource-efficient authentication scheme utilizing TinyJAMBU, which we collectively refer to as the system model.

4.1 Network model

The proposed network model for DITEN in industrial applications is illustrated in Figure 1. The model consists of four key entities, each playing a crucial role in the authentication process.

FIGURE 1

4.1.1 Initialization phase

Table 2 summarizes the information stored in each entity during the initialization phase.

4.1.2 Registration authority (RA)

The Registration Authority (RA) is a completely trusted and dependable entity tasked with initializing and managing the system’s cryptographic parameters (Sal et al., 2021). The RA carries out the following essential functions:

4.1.2.1 System initialization

The RA generates the essential system parameters, including the master secret key , public parameters , and the cryptographic primitive initialization, specifically TinyJAMBU’s key and nonce . This step is represented as Step 1 in Figure 1.

4.1.2.2 Device and user registration

Smart industrial devices (SID) and mobile users (MU) must register online with the RA. During this process, the RA securely encrypts the registration credentials using TinyJAMBU and stores the results. This step is illustrated as Step 2 and Step 3 in Figure 1.

4.1.2.3 Smart gateway (SG) offline registration

The RA executes the offline registration for the smart gateway, preloading significant authentication credentials, including (ciphertext generated using TinyJAMBU) into the SG’s database. These credentials are subsequently utilized during the Mutual Authentication and Key Agreement (MAKA) phase, shown in Step 4 and Step 5 in Figure 1.

4.1.3 Smart industrial device (SID)

The Smart Industrial Device (SID) is an integral component of the network, responsible for performing industrial tasks and interacting with digital twins in the DITEN framework. The SID must:

4.1.3.1 Execute online registration

Before deployment, each SID must register with the RA. The RA encrypts the SID’s identity and other credentials using TinyJAMBU and stores these securely in the SID’s memory. This process is depicted in Step 2 of Figure 1.

4.1.4 Mobile user (MU)

The Mobile User (MU) represents any user entity, such as an employee or technician, who interacts with the DITEN using mobile terminals like smartphones, tablets, or laptops. The MU’s role includes:

4.1.4.1 Online registration

The MU registers with the RA, where their identity , password , and biometric template are encrypted using TinyJAMBU and stored in a smart card . This registration is facilitated by the RA and is shown as Step 3 in Figure 1.

4.1.5 Smart gateway (SG)

The Smart Gateway (SG) serves as the communication bridge between mobile users, smart industrial devices, and cloud-based industrial applications. The SG performs the following:

4.1.5.1 Data collection and secure communication

The SG securely collects private data or service information from various industrial devices and ensures secure communication between mobile users and devices through Peer-to-Peer (P2P) interactions. This role is essential for real-time data exchange in DITEN.

4.1.5.2 Offline registration and credential management

The SG is preloaded with encrypted credentials during the RA’s offline registration process. These credentials are later used to authenticate devices and users in the network.

Figure 1 illustrates the detailed network model for the proposed authentication scheme using TinyJAMBU.

4.2 Adversary model

We adopt and expand upon existing adversary models, particularly those presented in (Wang and Wang, 2018), (Wang et al., 2022). The adversary model captures the capabilities of an attacker, denoted as , in various scenarios.

4.2.1 C-1: offline enumeration attacks

An adversary can perform offline enumeration attacks by computing all possible combinations within the Cartesian product , where and represent the metric spaces of identities and passwords, respectively. The adversary’s computational power is polynomial in time, implying:

Where is the time complexity and is the number of elements in the Cartesian product.

4.2.2 C-2: Dolev-Yao (DY) attack model

In the Dolev-Yao (DY) attack model (Dolev and Yao, 1983), has full control over the communication channel. can intercept, eavesdrop, modify, delete, and replay messages exchanged between entities. This capability is formalized as:where represents the original message and the altered message.

4.2.3 C-3: compromised authentication factors

An adversary has the capability to compromise as many as factors in an -factor authentication scheme. The factors may consist of the victim’s password , biometric data , smart card , and mobile terminal . Mathematically, this is expressed as:

4.2.4 C-4: canetti-krawczyk (CK) attack model

4.2.5 C-5: device compromise and side-channel attacks

An adversary has the ability to compromise a specific number of smart industrial devices (SID) and retrieve sensitive data from their internal storage utilizing techniques such as power analysis or side-channel attacks (Zhao and Suh, 2018). This capability can be represented as:

Where represents the internal sensitive data of the SID.

4.2.6 C-6: digital forensic attacks

can use digital forensic techniques to acquire previously established session keys . The primary cause of this vulnerability is the improper erasure of data during the registration phase:

4.2.7 C-7: physical capture and tamper-resistance

End-point entities, including smart industrial devices (SID) and mobile terminals (MT), are not completely trusted since they can be physically seized or stolen. In contrast, smart gateway nodes (SG) are equipped with tamper-resistant features, rendering it impractical for to capture and extract information:where A (Capture (MT, SID)) represents the adversary’s ability to compromise MT or SID, and A (Capture (SG) represents the inability to compromise the SG due to tamper resistance.

This detailed system model lays the foundation for the security and performance analyses of the proposed TinyJAMBU-based authentication scheme. Each entity’s role is well-defined, and the adversary’s capabilities are rigorously modeled to ensure comprehensive security evaluations.

5 Proposed scheme using TinyJAMBU

5.1 System setup phase

The Registration Authority (RA) is responsible for initializing essential system parameters and cryptographic algorithms necessary for constructing the authentication scheme using TinyJAMBU. The specific notations are detailed in Table 3.

The RA publishes the public parameters:and securely retains the secrets , , , , and .

Figure 2 illustrates the key phases, including system initialization, registration of entities, mutual authentication, and session key agreement. Sequence of processes shows the interactions between the Registration Authority (RA), Smart Industrial Device (SID), Mobile User (MU), and Smart Gateway (SG).

FIGURE 2

5.2 Registration phase

This phase involves the registration of network entities, including smart industrial devices, gateway nodes, and mobile users. The RA conducts the registration phase through both online and offline methods.

5.2.1 Online smart industrial device registration

Figure 3 illustrates the step-by-step interactions between the Smart Industrial Device (SID), Registration Authority (RA), and Gateway Node (GWN) during the registration phase, highlighting the secure exchange of identities, cryptographic keys, and PUF-based authentication parameters.

FIGURE 3

5.2.2 User enrollment process

Figure 4 illustrates the step-by-step interactions between the Mobile User (MU) and the Registration Authority (RA) during the registration phase. It highlights the secure exchange of user credentials, cryptographic computations, and the issuance of a smart card containing encrypted parameters. The diagram captures key operations such as masked password generation, biometric template processing, and the storage of authentication factors, ensuring clarity in the registration workflow.

FIGURE 4

5.2.3 Smart gateway registration in offline mode

The RA employs a randomized approach to create a unique identifier for the smart gateway and proceeds to derive the associated pseudo-identity.and stores the parameters

In the SG’s secure database.

5.3 Authentication phase

The authentication phase includes both the gateway and user authentication processes, as described below.

5.3.1 Gateway authentication

5.3.2 User authentication

5.4 User login phase

When a registered user wants to access the services, they must log in to the DITEN system. The detailed steps are as follows expressed in Algorithms 3, 4:

5.4.1 Step ULG1

The user inserts their smart card into the card slot of their terminal device and enters their identity , password , and also imprints their biometric template . The system then calculates a stable cryptographic key from the biometric input using the Fuzzy Extractor mechanism, ensuring that the Hamming distance between the current biometric template and the original template registered at the time of user registration is less than or equal to (where is the predefined threshold).

Subsequently, the terminal device performs the following computations using standard cryptographic primitives: Computes

Computes

Computes

Computes the masked password.

Computes

Validates the fuzzy verifier condition: , where . If this condition does not hold, the login request is denied.

If the fuzzy verifier check passes, the user proceeds by selecting a random nonce and computing , where is the Chebyshev polynomial seed established during system setup.

The user then selects the desired smart industrial device along with its pseudo-identity , and calculates , where is a pre-computed authentication parameter stored on the smart card.

Finally, the user constructs and sends the login authentication message:to the smart gateway via a public channel.

Figure 5 illustrates the step-by-step user login and smart card operations in the proposed authentication scheme. It highlights the user’s interaction with their smart card, including biometric verification, cryptographic computations, and the generation of a login authentication message. The diagram captures key operations such as calculating the stable cryptographic key, validating the fuzzy verifier, and securely transmitting the authentication message to the smart gateway over a public channel.

FIGURE 5

5.5 Mutual authentication and key agreement phase

In this phase, mutual authentication and key agreement (MAKA) are established among the registered user , the smart gateway , and the targeted smart industrial device . Upon successful completion, and collaboratively determine a secret session key for secure communication. The steps are as follows:

5.5.1 Step MAKA1: gateway authentication and decryption

After receiving the login message from , the smart gateway retrieves from its secure database. It then checks: - If , this indicates the previous session failed to update the temporary identity. - If , this indicates the temporary identity was successfully updated in the last session. - If no record of exists, the mutual authentication process is immediately halted.

For the first two cases, extracts the corresponding long-term secret key , pseudo-identity , and registration password from its database. then calculates:and splits into two 128-bit parts. Using these, applies TinyJAMBU authenticated decryption with key and nonce to recover the original parameters. Finally, validates the fuzzy verifier condition . If the condition holds, the protocol proceeds; otherwise, it halts.

5.5.2 Step MAKA2: smart industrial device authentication

Upon receiving the message from , the smart industrial device performs the following: - Computes using its Physical Unclonable Function (PUF). - Calculates . - Computes . - Validates by performing TinyJAMBU decryption on .

If the validation succeeds, generates a new nonce and a new temporary identity . It then calculates the session key:and sends the mutual authentication response message to .

5.5.3 Step MAKA3: final user validation

Upon receiving the response message from , the user validates the message by checking:

If this condition holds, proceeds to calculate the shared session key:

Finally, stores for future secure communication with .

Figure 6 illustrates the step-by-step mutual authentication and key agreement (MAKA) process in the proposed scheme. It highlights the interactions between the user, smart gateway, and smart industrial device during the authentication phase. The diagram captures key operations such as gateway decryption, device validation, and final user verification, ensuring secure communication through the establishment of a shared session key. Each step includes validation checks to ensure the integrity and authenticity of the process.

FIGURE 6

5.6 TinyJAMBU’s user password and biometrics update phase (PBUP)

User can modify their password and biometric data locally for security reasons. This local update avoids involving the registry authority (RA), thus reducing communication and computational costs.

5.6.1 Step PBUP1: initial input and validation.

5.6.2 Step PBUP2: new password and biometric template update

5.6.3 Step PBUP3: parameter update on the smart card

The authenticated encryption and decryption processes during this phase use TinyJAMBU for secure and efficient cryptographic operations, ensuring the confidentiality and integrity of the updated password and biometric data.

5.7 User smart card revocation using TinyJAMBU

The ability to revoke a smart card is provided under this subsection. For instance, a user can cancel their prior smart card details if their smart card is misplaced, stolen, or lost.

5.7.1 Step SCRP1: revocation request

5.7.2 Step SCRP2: verification and revocation

5.8 Dynamic smart industrial device addition phase with TinyJAMBU

This phase focuses on the provision of the capability for the addition of new devices of smart industrial design. After the first installation, when some of the smart devices work poorly, and the normal function of the system is required, smart industrial devices are present for ensuring such.

5.8.1 Step DSDA1: identity submission.

5.8.2 Step DSDA2: key generation and challenge

5.8.3 Step DSDA3: PUF operation and key generation

5.8.4 Step DSDA4: database update

6 Security analysis of TinyJAMBU

This section presents a comprehensive security analysis of the TinyJAMBU authenticated encryption algorithm, which serves as the cryptographic core of the proposed TTG scheme. The analysis is structured into three main parts: (1) a formal security analysis using established AEAD security models, (2) a heuristic security analysis focusing on structural robustness, and (3) formal verification using the AVISPA tool suite. This multi-faceted approach ensures both theoretical rigor and practical validation.

6.1 Formal security analysis using AEAD security models

TinyJAMBU is an Authenticated Encryption with Associated Data (AEAD) scheme, standardized in the NIST Lightweight Cryptography portfolio (Saha et al., 2019). Its security must be evaluated under the standard notions for AEAD confidentiality and integrity (Bellare and Rogaway, 2000; Rogaway, 2002).

6.1.1 Security notions and model

For TinyJAMBU, security is based on the pseudorandomness of its underlying keyed permutation . The algorithm follows a duplex-sponge construction (Bertoni et al., 2011), where the permutation is applied to a state initialized with the key , nonce , and associated data .

6.1.2 Confidentiality of TinyJAMBU

6.1.3 Integrity of TinyJAMBU

This aligns with the security proof for sponge-based AEAD schemes (Bertoni et al., 2011). □

6.1.4 Critical security requirement: nonce uniqueness

The combination ensures probabilistic uniqueness with negligible collision probability.

6.1.5 Replay attack resistance protocol design

Clarification: Replay resistance is not an intrinsic property of TinyJAMBU but is achieved through protocol design. The statement in the original manuscript implying otherwise was incorrect.

Thus, while TinyJAMBU provides confidentiality and integrity for individual messages, replay protection is enforced at the protocol layer through explicit freshness mechanisms.

6.1.6 Alignment with TinyJAMBU’s design

This analysis corrects the earlier inaccurate representation that decomposed TinyJAMBU into . Instead, we adhere to the actual specification: TinyJAMBU uses a keyed permutation to update its internal state, and the tag is a direct extract of that state. This aligns with established security proofs for TinyJAMBU (Saha et al., 2019; Saha et al., 2020).

6.2 Heuristic security evaluation

This section presents an informal assessment of TinyJAMBU’s resistance to well-known cryptanalytic attacks. The analysis complements the formal security arguments provided in Section 6.1 by examining the algorithm’s structural properties and existing cryptanalytic evidence.

6.2.1 Resistance to differential and linear cryptanalysis

For differential attacks, the maximum probability of any differential characteristic over rounds is upper bounded bywhich renders full-round attacks computationally infeasible.

6.2.2 Resistance to side-channel attacks

These properties make TinyJAMBU well suited for constrained environments that require side-channel resistance without excessive performance overhead.

6.3 Formal verification using AVISPA

To formally validate the security properties of the TTTG protocol, we performed automated verification using the Automated Validation of Internet Security Protocols and Applications (AVISPA) framework (Armando et al., 2005). This verification provides formal assurance that the protocol design is free from logical flaws under a well-defined adversary model.

6.3.1 Verification environment setup

6.3.2 Adversary model

6.3.3 Formal security goals

6.3.4 Verification outcomes

FIGURE 7

6.3.5 Limitations and scope

Despite these limitations, the formal verification provides strong evidence that TTG’s design is logically sound and resistant to the class of attacks modeled. Combined with the theoretical security analysis of TinyJAMBU (Section 6.1) and heuristic evaluation (Section 6.2), this comprehensive approach ensures robust security for DITEN environments.

6.4 Multi-factor authentication security analysis

TTG adopts a three-factor authentication model consisting of possession (smart card), knowledge (password), and inherence (biometric data). This section evaluates the protocol’s robustness when up to authentication factors are compromised.

6.4.1 Security model

Let denote the authentication factors. An adversary is assumed capable of compromising at most two factors. The compromise function is defined as

The scheme is considered secure if, for any adversary satisfying , the probability of successful impersonation remains negligible.

6.4.2 Analysis of compromised factor combinations

6.4.2.1 Case 1: compromise of smart card and password

6.4.3 Case 2: compromise of smart card and biometric

6.4.4 Case 3: compromise of password and biometric

6.4.5 Overall security bound

In all cases, the probability of a successful impersonation by is bounded bywhere denotes the false acceptance rate of the fuzzy extractor for threshold .

This result demonstrates that TTG remains secure even if any two authentication factors are compromised, thereby satisfying the multi-factor security requirement listed in Table 5.

6.5 Security evaluation criteria

The security evaluation criteria for our proposed scheme using TinyJAMBU are summarized in Table 4. Each security indicator term is defined in the context of DITEN (Dynamic Industrial IoT Environments).

6.5.1 Security requirements

     The scheme must provide strong security even if factors are compromised in a -factor authentication system. This ensures resilience against partial breaches.

Table 5 summarizes the security requirements and their descriptions.

These security requirements form the foundation of the proposed scheme, ensuring comprehensive protection against a wide range of threats in DITEN environments. They are aligned with the evaluation criteria discussed in Table 4 and are rigorously validated through formal security proofs and heuristic analysis.

6.5.2 Proof of multi-factor security

The proposed scheme This system integrates three components: a smart card, a password, and a biometric template.To demonstrate its security, we assume an adversary compromises two of these factors and consider three possible scenarios:

6.5.3 Case 1: password and smart card compromise

Suppose knows the user’s password and has access to their smart card. In this scenario, is capable of retrieving all stored parameters from the smart card. Nevertheless, is unable to acquire the biometric secret key necessary for generating a legitimate login message. The adversary does not have access to the user’s biometric template , and is concealed within cryptographic parameters that are generated through hash functions and random values.

6.5.4 Case 2: password and biometric template compromise

Assuming knows the user’s password and biometric template, the adversary can calculate the user’s biometric secret key . In order to create a legitimate login message, must know , , , and other critical values. Even if extracts some smart card parameters, it cannot calculate the session key without the nonce . Therefore, the adversary is incapable of creating a valid login message.

6.5.5 Case 3: biometric template and smart card compromise

In conclusion, should adversary obtain the user’s biometric template along with the smart card, it can derive the biometric secret key and extract the parameters stored on the smart card. However, without the user’s password , it is computationally unfeasible for to create a valid login authentication message. In all instances, the proposed scheme provides multi-factor security. Even if two factors are breached, the third factor remains intact, rendering it impossible for the adversary to fabricate a legitimate login message.

7 Performance comparison analysis

This study presents a comparative assessment of the proposed scheme that incorporates TinyJAMBU, comparing it with seven relevant multi-factor user authentication schemes. The primary focus is on the security functionality features, in addition to the computational and communication overheads.

7.1 Comparison on security functionality features

In this analysis, we assess the proposed scheme in relation to other comparable schemes, utilizing the evaluation criteria detailed in Table 4. The findings of this comparison are illustrated in Table 6. The symbol “” is used to indicate that a scheme is reported to be secure or offers a certain security feature in its original publication, while the symbol “” denotes that a scheme either lacks explicit support for or has been shown to be vulnerable to attacks regarding that specific criterion.

According to our analysis based on the criteria in Table 4, the proposed scheme with TinyJAMBU appears to satisfy the 13 evaluated security requirements within the context of DITEN. This contrasts with the other surveyed schemes, which, based on available literature, exhibit various security limitations that may leave them vulnerable to certain attack vectors in industrial edge environments. It is important to note, however, that security evaluations are dependent on specific threat models and implementation assumptions, and direct comparison between schemes requires careful consideration of their respective design contexts.

Among the surveyed schemes, our analysis indicates that the proposed TTG scheme provides coverage for the widest range of security requirements defined for DITEN environments. This comprehensive coverage suggests its potential suitability for industrial applications where multiple security properties must be simultaneously addressed. The evaluation criteria (P1-P7, S1-S6) were specifically defined to capture the unique security challenges of DITEN, including real-time constraints, device heterogeneity, and integration with digital twin technologies.

For instance, Vinoth et al. (2020) fails to meet several key security properties, particularly (Password and Biometric Friendliness), (Provision of Session Key Agreement), (Mutual Authentication Verification), and (User Anonymity and Un-Traceability). This lack of support for crucial password, biometric, and session key provisions makes their scheme less secure in environments where user credentials and session communication integrity are paramount. Furthermore, the absence of mutual authentication means that their scheme does not effectively protect against impersonation or identity spoofing.

Similarly, Wang et al. (2020) covers most password- and session-related properties but fails to meet (Perfect Forward Secrecy), (Resistance to Known Attacks), and (Resistance to Node Capture Attacks). The lack of perfect forward secrecy indicates that, if a password or biometric template were compromised, previous sessions could be decrypted, presenting a significant risk in industrial environments. Additionally, the scheme’s inability to resist known attacks and node capture scenarios suggests that it may be susceptible to adversaries targeting weak points in the network infrastructure.

Alsahlani and Popa (2021) is also deficient in (Perfect Forward Secrecy), (Resistance to Node Capture Attacks), and (Password and Biometric Friendliness). While it covers most of the session key agreement and multi-factor security properties, the inability to protect previous session keys and resistance to device compromise leaves it vulnerable in high-security industrial environments where device security is crucial.

In contrast, the proposed scheme with TinyJAMBU fulfills all security requirements. This includes ensuring (Password and Biometric Friendliness), allowing for local password and biometric updates, and (Provision of Session Key Agreement), which enables the user and device to securely negotiate session keys. It also meets (User Anonymity and Un-Traceability), ensuring that an adversary cannot trace user activities or calculate their identities, and (Perfect Forward Secrecy), protecting past session keys even if current credentials are compromised.

Additionally, the proposed scheme demonstrates strong resilience against well-known attacks by satisfying (Resistance to Known Attacks), covering threats such as replay, man-in-the-middle, and impersonation. It also provides (Resistance to Smart Card Loss Attacks), ensuring that even if a smart card is compromised, the adversary cannot impersonate the user or misuse the card. Lastly, the scheme meets (Resistance to Node Capture Attacks), protecting the network even if industrial devices are captured by an adversary.

This comprehensive coverage makes the proposed scheme with TinyJAMBU far more secure and suitable for DITEN than the other schemes, ensuring robust user protection, communication security, and resilience against a wide range of attacks.

7.2 Performance comparison analysis

This section provides a comprehensive comparison of the proposed scheme TTG with existing multi-factor authentication schemes. The analysis focuses on computational overhead, communication overhead, energy consumption, and storage overhead.

7.2.1 Energy consumption

Energy consumption is a critical metric for evaluating the efficiency of cryptographic algorithms in resource-constrained environments. We measured the energy consumed during key operations (e.g., encryption, decryption, hashing) on both a high-performance server and a Raspberry Pi 4B. The results are summarized in Table 7.

The results demonstrate that TinyJAMBU consumes significantly less energy compared to AES and ASCON, particularly on resource-constrained devices like the Raspberry Pi.

7.2.2 Storage overhead

Storage overhead refers to the memory required to store cryptographic parameters on smart cards, devices, and gateways. Table 8 compares the storage requirements of the proposed scheme with existing schemes.

The proposed scheme requires minimal storage, with only 384 bytes on the smart card, 640 bytes on the device, and 896 bytes on the gateway. This low storage overhead makes it ideal for edge devices with limited memory resources. The results from both energy consumption and storage overhead analyses highlight the lightweight and efficient design of TTG. By leveraging TinyJAMBU, the scheme achieves significant reductions in energy usage and memory requirements, making it highly suitable for DITEN.

7.3 Comparison on computational overheads

Due to their remarkably short execution times, concatenation and bitwise XOR operations have been disregarded.

Based on (Wang et al., 2020), we estimate that ms, ms, and ms. The times for other operations are measured using the MIRACL (SDK, 2020) library and eBACS cryptographic benchmarking (Bernstein and Lange, 2009). The results are illustrated in Table 9 and Table 10 for server and Raspberry Pi settings, respectively. These values correspond to the ECC-based scheme of Wang et al. (Wang et al., 2020), against which our 43.7% computational overhead reduction is measured.

7.4 Comparison on communication overheads

The communication overhead for the proposed scheme with TinyJAMBU, and other related schemes, is compared in Table 11. In our scheme, there are three transmitted messages during the user login and MAKA phases. The total communication overhead is 4,896 bits.

7.5 Performance summary

TTG employs a hybrid cryptographic architecture that integrates the NIST-standardized TinyJAMBU AEAD primitive with established lightweight building blocks—including secure hash functions, Chebyshev polynomials, and fuzzy extractors—to achieve comprehensive security coverage.

As demonstrated in Tables 9–11, this design incurs moderately higher computational and communication overhead compared to certain ultra-lightweight authentication protocols. Specifically, the total communication cost of TTG is 4,896 bits, which exceeds that of schemes like Alsahlani et al. (Alsahlani and Popa, 2021) (2,560 bits) and Srinivas et al. (Srinivas et al., 2018) (2,912 bits).

However, this trade-off is deliberate and necessary to fulfill all 13 critical security requirements—including perfect forward secrecy, resistance to privileged-insider and node capture attacks, and full multi-factor security—none of which are simultaneously satisfied by the lighter alternatives. Furthermore, TTG reduces computational overhead by 43.7% compared to ECC-based approaches and achieves 58.3% lower energy consumption than conventional implementations, making it well-suited for resource-constrained edge environments where security completeness is non-negotiable.

8 Conclusion

This work addresses the critical security challenges in DITEN, where traditional authentication schemes often struggle to balance security with computational efficiency in resource-constrained environments. Our analysis indicates that TTG addresses a comprehensive set of 13 security requirements (S1-S6 and P1-P7) relevant to DITEN, while the surveyed competing approaches exhibit 3-7 limitations in various aspects of our evaluation framework. While TTG incurs moderately higher communication overhead than ultra-lightweight hash-based schemes (e.g., Alsahlani et al. (Alsahlani and Popa, 2021), Srinivas et al. (Srinivas et al., 2018)), it achieves a 43.7% reduction in computational overhead compared to ECC-based protocols such as Wang et al., with critical operations executing in ms and ms. Furthermore, energy consumption tests show a 58.3% reduction in power usage during authentication sequences, which may enable extended device lifespans in battery-constrained settings. These findings suggest that TTG offers a balanced approach to security and efficiency for lightweight authentication in industrial IoT. However, several limitations should be noted: (i) ARM Cortex-M4 optimization exhibits an 18.7% performance degradation on legacy hardware, which could impact real-time responsiveness in time-critical applications; (ii) integration challenges with existing PKI infrastructures may require transitional middleware for deployment in legacy systems; and (iii) the scheme currently lacks quantum resistance. Future work will explore quantum-resistant variants while maintaining computational efficiency; integrate lightweight anomaly detection mechanisms; investigate blockchain-anchored trust frameworks for reduced authentication latency; and expand architectural support with minimal performance impact. TTG represents an advancement in lightweight authentication for edge environments, demonstrating that broad security coverage and practical efficiency can be achieved without relying on traditional heavy cryptographic primitives.

References

• 1

  Abdullah A. M. (2017). Advanced encryption standard (Aes) algorithm to encrypt and decrypt data. Cryptogr. Netw. Secur.16 (1), 11.

  • Google Scholar

• 2

  Almasri M. Al-Kuwari S. (2023). Fog-layer authentication using pedersen commitments. IEEE Trans. Cloud Comput.11 (2), 1452–1463.

  • Google Scholar

• 3

  Alsahlani A. Y. F. Popa A. (2021). Lmaas-iot: lightweight multi-factor authentication and authorization scheme for real-time data access in iot cloud-based environment. J. Netw. Comput. Appl.192, 103177. 10.1016/j.jnca.2021.103177

  • CrossRef
  • Google Scholar

• 4

  Alsahli M. Borgognoni A. Cardoso dos Santos L. Cheng H. Franck C. Großschädl J. (2022). “Lightweight permutation-based cryptography for the ultra-low-power internet of things,” in International conference on information technology and communications security (Springer), 17–36.

  • Google Scholar

• 5

  Armando A. Basin D. Boichut Y. Chevalier Y. Compagna L. Cuéllar J. et al (2005). “The avispa tool for the automated validation of internet security protocols and applications,” in Computer aided verification: 17Th international conference, CAV 2005, Edinburgh, Scotland, UK, July 6-10, 2005. Proceedings 17 (Springer), 281–285.

  • Google Scholar

• 6

  Bellare M. Rogaway P. (2000). “Relations among notions of security for public-key encryption schemes,” in Advances in Cryptology—crypto’98, 26–45.

  • Google Scholar

• 7

  Bergamo P. D’Arco P. De Santis A. Kocarev L. (2005). Security of public-key cryptosystems based on chebyshev polynomials. IEEE Trans. Circuits Syst. I Regul. Pap.52 (7), 1382–1393. 10.1109/tcsi.2005.851701

  • CrossRef
  • Google Scholar

• 8

  Bernstein D. J. Lange T. , Ebacs: ecrypt benchmarking of cryptographic systems (2009).

  • Google Scholar

• 9

  Bertoni G. Daemen J. Peeters M. Van Assche G. (2011). “Duplexing the sponge: single-pass authenticated encryption and other applications,” in International workshop on selected areas in cryptography (Springer), 320–337.

  • Google Scholar

• 10

  Canetti R. Krawczyk H. (2001). “Analysis of key-exchange protocols and their use for building secure channels,” in Advances in cryptology — eurocrypt 2001. Editor PfitzmannB. (Berlin, Heidelberg: Springer Berlin Heidelberg), 453–474.

  • Google Scholar

• 11

  Chan J. Rogaway P. (2019). “Anonymous ae,” in Advances in Cryptology–ASIACRYPT 2019: 25Th international conference on the theory and application of cryptology and information security, kobe, Japan, December 8–12, 2019, proceedings, part II 25 (Springer), 183–208.

  • Google Scholar

• 12

  Chaudhry S. Al-Turjman F. (2022). Blockchain-based secure authentication for edge networks. IEEE Trans. Industrial Inf.18 (5), 3520–3529.

  • Google Scholar

• 13

  Chen W. Li H. (2021). Sponge-based lightweight authentication for edge devices. ACM Trans. Embed. Comput. Syst.20 (5), 1–22.

  • Google Scholar

• 14

  Cheng H. Großschädl J. Marshall B. Page D. Pham T. (2023a). Risc-v instruction set extensions for lightweight symmetric cryptography. IACR Trans. Cryptogr. Hardw. Embed. Syst., 193–237.

  • Google Scholar

• 15

  Cheng H. Grosschädl J. Marshall B. Page D. Pham T. (2023b). RISC-V instruction set extensions for lightweight symmetric cryptography. IACR Trans. Cryptogr. Hardw. Embed. Syst.2023 (2), 193–237.

  • Google Scholar

• 16

  Cui J. Cheng F. Zhong H. Zhang Q. Gu C. Liu L. (2023). Multi-factor based session secret key agreement for the industrial internet of things. Ad Hoc Netw.138, 102997. 10.1016/j.adhoc.2022.102997

  • CrossRef
  • Google Scholar

• 17

  Datta N. Dutta A. Ghosh S. (2022). “Int-rup security of saeb and tinyjambu,” in International conference on cryptology in India (Springer), 146–170.

  • Google Scholar

• 18

  Dobraunig C. Eichlseder M. Mendel F. Schläffer M. (2021). Ascon v1. 2: lightweight authenticated encryption and hashing. J. Cryptol.34, 1–42. 10.1007/s00145-021-09398-9

  • CrossRef
  • Google Scholar

• 19

  Dodis Y. Reyzin L. Smith A. (2004). “Fuzzy extractors: how to generate strong keys from biometrics and other noisy data,” in Advances in Cryptology-EUROCRYPT 2004: international conference on the theory and applications of cryptographic techniques, interlaken, Switzerland, may 2-6, 2004. Proceedings 23 (Springer), 523–540.

  • Google Scholar

• 20

  Dolev D. Yao A. (1983a). On the security of public key protocol. IEEE Trans. Inf. Theory29 (2), 198–208. 10.1109/TIT.1983.1056650

  • CrossRef
  • Google Scholar

• 21

  Dolev D. Yao A. C. (1983b). “On the security of public key protocols,” in 24th annual symposium on foundations of computer science (IEEE), 350–357.

  • Google Scholar

• 22

  Dubrova E. Teslenko M. Tenhunen H. (2008). “On analysis and synthesis of (n, k)-non-linear feedback shift registers,” in Proceedings of the conference on design, automation and test in Europe, 1286–1291.

  • Google Scholar

• 23

  Dunkelman O. Ghosh S. Lambooij E. (2022). “Full round zero-sum distinguishers on tinyjambu-128 and tinyjambu-192 keyed-permutation in the known-key setting,” in International conference on cryptology in India (Springer), 349–372.

  • Google Scholar

• 24

  El-Hajj M. Fadlallah A. (2022). “Analysis of lightweight cryptographic algorithms on iot hardware platforms,” in 2022 32nd international telecommunication networks and applications conference (ITNAC), 121–126. 10.1109/ITNAC55475.2022.9998413

  • CrossRef
  • Google Scholar

• 25

  El-Hajj M. Chamoun M. Fadlallah A. Serhrouchni A. (2017). “Analysis of authentication techniques in internet of things (Iot),” in 2017 1st cyber security in networking conference (CSNet) (IEEE), 1–3.

  • Google Scholar

• 26

  El-Hajj M. Fadlallah A. Chamoun M. Serhrouchni A. (2019). A survey of internet of things (iot) authentication schemes. Sensors19 (5), 1141. 10.3390/s19051141

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  El-Hajj M. Mousawi H. Fadlallah A. (2023). Analysis of lightweight cryptographic algorithms on iot hardware platform. Future Internet15 (2), 54. 10.3390/fi15020054

  • CrossRef
  • Google Scholar

• 28

  El-Hajj M. Itäpelto T. Gebremariam T. , (2025). Systematic literature review: Digital twins’ role in enhancing security for industry 4.0 applications, security and privacy e396.

  • Google Scholar

• 29

  Fuller B. Meng X. Reyzin L. (2020). Computational fuzzy extractors, information and computation, 275.104602

  • Google Scholar

• 30

  Galbally J. Ross A. Gomez-Barrero M. Fierrez J. Ortega-Garcia J. (2013). Iris image reconstruction from binary templates: an efficient probabilistic approach based on genetic algorithms. Comput. Vis. Image Underst.117 (10), 1512–1525. 10.1016/j.cviu.2013.06.003

  • CrossRef
  • Google Scholar

• 31

  Gao C. Xin X. Lu Y. Zhang Y. Yang J. Shu J. (2021). “Parabit: processing parallel bitwise operations in nand flash memory based ssds,” in MICRO-54: 54Th annual IEEE/ACM international symposium on microarchitecture, 59–70.

  • Google Scholar

• 32

  Jhawar K. Gandhi J. Shekhawat D. Upadhyay A. Harkishanka A. Chaturvedi N. et al (2024). Modeling, hardware architecture, and performance analyses of an aead-based lightweight cipher. J. Real-Time Image Process.21 (2), 32. 10.1007/s11554-024-01416-w

  • CrossRef
  • Google Scholar

• 33

  Khan S. H. Akbar M. A. Shahzad F. Farooq M. Khan Z. (2015). Secure biometric template generation for multi-factor authentication. Pattern Recognit.48 (2), 458–472. 10.1016/j.patcog.2014.08.024

  • CrossRef
  • Google Scholar

• 34

  Li N. Guo F. Mu Y. Susilo W. Nepal S. (2017). “Fuzzy extractors for biometric identification,” in 2017 IEEE 37th international conference on distributed computing systems (ICDCS) (IEEE), 667–677.

  • Google Scholar

• 35

  Liu T. Tang L. Wang W. Chen Q. Zeng X. (2021). Digital-twin-assisted task offloading based on edge collaboration in the digital twin edge network. IEEE Internet Things J.9 (2), 1427–1444. 10.1109/jiot.2021.3086961

  • CrossRef
  • Google Scholar

• 36

  Loos R. Giusti M. Pardo L. Niklasch G. Dentzer R. Landau S. et al (2003). “Topics of computer algebra,” in Computer algebra handbook: Foundations⋅ applications⋅ systems (Springer), 11–162.

  • Google Scholar

• 37

  Mahdi M. S. Hassan N. F. Abdul-Majeed G. H. (2021). An improved chacha algorithm for securing data on iot devices. SN Appl. Sci.3 (4), 429. 10.1007/s42452-021-04425-7

  • CrossRef
  • Google Scholar

• 38

  Mason J. C. Handscomb D. C. (2002). Chebyshev polynomials. Chapman and Hall/CRC.

  • Google Scholar

• 39

  Radhakrishnan I. Jadon S. Honnavalli P. B. (2024). Efficiency and security evaluation of lightweight cryptographic algorithms for resource-constrained iot devices. Sensors24 (12), 4008. 10.3390/s24124008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  Rogaway P. (2002). “Authenticated-encryption with associated-data,” in Proceedings of the 9th ACM conference on computer and communications security, 98–107.

  • Google Scholar

• 41

  Rushad M. Nambiar A. Chandavarkar B. (2022). Resource-aware cryptography: an analysis of lightweight cryptographic primitives. SN Comput. Sci.3, 1–12. 10.1007/s42979-021-00984-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  Saha D. Sasaki Y. Shi D. Sibleyras F. Sun S. Zhang Y. (2019). “Tinyjambu: a family of lightweight authenticated encryption algorithms,” in NIST lightweight cryptography project. Available online at: https://csrc.nist.gov/projects/lightweight-cryptography.

  • Google Scholar

• 43

  Saha D. Sasaki Y. Shi D. Sibleyras F. Sun S. Zhang Y. (2020). On the security margin of tinyjambu with refined differential and linear cryptanalysis, IACR transactions on symmetric cryptology, 152–174.

  • Google Scholar

• 44

  Sai A. M. V. V. Wang C. Cai Z. Li Y. (2024). Navigating the digital twin network landscape: a survey on architecture, applications, privacy and security. High-Confidence Comput.4, 100269. 10.1016/j.hcc.2024.100269

  • CrossRef
  • Google Scholar

• 45

  Saleem M. A. Shamshad S. Ahmed S. Ghaffar Z. Mahmood K. (2021). Security analysis on “a secure three-factor user authentication protocol with forward secrecy for wireless medical sensor network systems”. IEEE Syst. J.15 (4), 5557–5559. 10.1109/JSYST.2021.3073537

  • CrossRef
  • Google Scholar

• 46

  Sdk M. C. , Multiprecision integer and rational arithmetic cryptographic library (2020).

  • Google Scholar

• 47

  Singh R. Kumar P. (2022). Multi-modal biometric authentication for industrial iot. IEEE Trans. Industrial Inf.19 (1), 892–901.

  • Google Scholar

• 48

  Srinivas J. Das A. K. Wazid M. Kumar N. (2018). Anonymous lightweight chaotic map-based authenticated key agreement protocol for industrial internet of things. IEEE Trans. Dependable Secure Comput.17 (6), 1133–1146. 10.1109/tdsc.2018.2857811

  • CrossRef
  • Google Scholar

• 49

  Thakor V. A. Razzaque M. A. Khandaker M. R. (2021). Lightweight cryptography algorithms for resource-constrained iot devices: a review, comparison and research opportunities. IEEE Access9, 28177–28193. 10.1109/access.2021.3052867

  • CrossRef
  • Google Scholar

• 50

  Tian S. Vassilakis V. G. (2023). On the efficiency of a lightweight authentication and privacy preservation scheme for mqtt. Electronics12 (14), 3085. 10.3390/electronics12143085

  • CrossRef
  • Google Scholar

• 51

  van der Wal E. W. El-Hajj M. (2022). “Securing networks of iot devices with digital twins and automated adversary emulation,” in 2022 26th international computer science and engineering conference (ICSEC) (IEEE), 241–246.

  • Google Scholar

• 52

  Vinoth R. Deborah L. J. Vijayakumar P. Kumar N. (2020). Secure multifactor authenticated key agreement scheme for industrial iot. IEEE Internet Things J.8 (5), 3801–3811. 10.1109/jiot.2020.3024703

  • CrossRef
  • Google Scholar

• 53

  Wang D. Wang P. (2018). Two birds with one stone: two-factor authentication with security beyond conventional bound. IEEE Trans. Dependable Secure Comput.15 (4), 708–722. 10.1109/TDSC.2016.2605087

  • CrossRef
  • Google Scholar

• 54

  Wang D. Wang P. Wang C. (2020). Efficient multi-factor user authentication protocol with forward secrecy for real-time data access in wsns. ACM Trans. Cyber-Physical Syst.4 (3), 1–26. 10.1145/3325130

  • CrossRef
  • Google Scholar

• 55

  Wang C. Wang D. Tu Y. Xu G. Wang H. (2022). Understanding node capture attacks in user authentication schemes for wireless sensor networks. IEEE Trans. Dependable Secure Comput.19 (1), 507–523. 10.1109/TDSC.2020.2974220

  • CrossRef
  • Google Scholar

• 56

  Wazid M. Das A. K. Kumar N. Alazab M. (2020). Designing authenticated key management scheme in 6g-enabled network in a box deployed for industrial applications. IEEE Trans. Industrial Inf.17 (10), 7174–7184. 10.1109/tii.2020.3020303

  • CrossRef
  • Google Scholar

• 57

  Windarta S. Suryadi S. Ramli K. Lestari A. A. Wildan W. Pranggono B. et al (2023). Two new lightweight cryptographic hash functions based on saturnin and beetle for the internet of things. IEEE Access11, 84074–84090. 10.1109/access.2023.3301128

  • CrossRef
  • Google Scholar

• 58

  Yu W. Liang F. He X. Hatcher W. G. Lu C. Lin J. et al (2017). A survey on the edge computing for the internet of things. IEEE Access6, 6900–6919. 10.1109/access.2017.2778504

  • CrossRef
  • Google Scholar

• 59

  Zhang Y. Zheng Z. (2023). Lightchain: blockchain-empowered lightweight authentication. IEEE Internet Things J.10 (4), 3321–3333.

  • Google Scholar

• 60

  Zhang W. Bao Z. Lin D. Rijmen V. Yang B. Verbauwhede I. (2014). Rectangle: a bit-slice lightweight block cipher suitable for multiple platforms. Cryptology ePrint Archive. New Jersey, NJ: IEEE.

  • Google Scholar

• 61

  Zhang L. Zhu Y. Ren W. Wang Y. Choo K.-K. R. Xiong N. N. (2021). An energy-efficient authentication scheme based on chebyshev chaotic map for smart grid environments. IEEE Internet Things J.8 (23), 17120–17130. 10.1109/JIOT.2021.3078175

  • CrossRef
  • Google Scholar

• 62

  Zhao M. Suh G. E. (2018). “Fpga-based remote power side-channel attacks,” in 2018 IEEE symposium on security and privacy (SP) (IEEE), 229–244.

  • Google Scholar

• 63

  Ziauddin S. Martin B. (2013). “Formal analysis of iso/iec 9798-2 authentication standard using avispa,” in 2013 eighth Asia joint conference on information security (IEEE), 108–114.

  • Google Scholar