Enhancing Internet of Things security in healthcare using a blockchain-driven lightweight hashing system

Abstract

Background

The rapid expansion of Internet of Things applications in healthcare has created new opportunities for improving patient care through real-time monitoring and data sharing. However, this growth also introduces significant challenges related to data security, privacy, and system efficiency, especially for devices with limited processing power and energy resources. To address these issues, this study introduces a blockchain-based lightweight hashing system specifically designed for healthcare environments with resource-constrained devices. The goal is to ensure secure, efficient, and scalable handling of sensitive medical data without overwhelming the capabilities of connected devices.

Results

The proposed system combines a collision-resistant, lightweight hash function with blockchain technology to enhance data integrity, authentication, and privacy. The hash function minimizes computational demands, making it ideal for wearable and embedded healthcare devices. Blockchain integration enables decentralized data management, preventing unauthorized access and tampering. The system generates unique, immutable patient identifiers and protects electronic health information from common security threats, including collision attacks, Sybil attacks, and cryptographic analysis. Simulation results show improved computational efficiency, lower latency, and effective handling of high transaction volumes with minimal resource usage.

Conclusions

This research presents a secure and efficient framework for managing medical data in healthcare Internet of Things applications. By leveraging lightweight cryptographic techniques and decentralized data structures, the system addresses key limitations in current solutions while supporting scalability and real-world deployment. Potential applications include secure patient monitoring, real-time sharing of health data, and decentralized management of medical records. The proposed approach provides a foundation for future advancements in digital healthcare systems, particularly in remote care, emergency response, and wearable health technologies.

Introduction

In recent years, IoT has experienced significant growth, particularly in wireless networks and communication technologies. It focuses on interconnecting mobile devices, sensors, and RFID tags to enable virtualized information access through unique addresses. IoT is increasingly adopted across education, transportation, agriculture, and healthcare [1]. In healthcare, continuous remote monitoring via sensors in smartphones and wearables enhances patient care while reducing hospital visits, transportation, and insurance costs. This approach enables efficient data collection, allowing healthcare providers to improve service quality while optimizing resources [2]. A primary goal of IoT in healthcare is establishing secure communication between providers and patients. However, security and privacy challenges arise, particularly in protecting medical data. Addressing authentication, availability, data integrity, confidentiality, and non-repudiation is crucial [3]. Authentication ensures mutual device verification, enabling secure data transmission [3]. The rise of wearable technology and electronic health records introduces risks like data breaches, necessitating advanced security solutions. While past research has tackled these issues, further efforts are needed to strengthen IoT-based healthcare systems [4, 5]. Blockchain technology offers a promising solution due to its immutable, time-stamped records secured by cryptography. Implementing blockchain in electronic health records (EHRs) improves data integrity and protects against cyber threats. Its decentralized approach ensures sensitive patient data remains secure from unauthorized access [6, 7]. Research explores integrating collision-resistant hash functions into IoT healthcare applications to reinforce security [8]. These functions prevent unauthorized modifications and fraudulent records, preserving data authenticity [9, 10]. Traditional hash functions often require substantial computational power, making them impractical for IoT applications. To address this, researchers have developed lightweight, collision-resistant hash functions optimized for resource-limited devices, balancing security and efficiency [11, 12–13]. These functions secure data transactions with minimal computational overhead, making them ideal for healthcare IoT systems. Their integration into blockchain enhances data integrity, improves patient safety, and fosters trust in digital health solutions [14, 15]. Blockchain also strengthens the security of the Internet of Medical Things (IoMT), where reliable communication is essential [16, 17]. By leveraging blockchain, healthcare systems create decentralized networks that protect patient data and enable seamless device interactions [18, 19]. Hash functions generate unique identifiers for each data transaction, ensuring any alterations are detected [20]. Additionally, blockchain enables secure storage of digital health certifications and document signatures, granting patients greater control over their health data while maintaining privacy [21]. Collision-resistant hashing prevents preimage and second preimage attacks, ensuring robust security. As quantum computing advances, developing new cryptographic methods becomes crucial for safeguarding blockchain applications in healthcare [22, 23].

This study advances secure healthcare IoT systems by introducing a novel, lightweight hash function tailored for healthcare applications. It optimally balances security and computational efficiency, minimizing processing demands for IoT devices. By integrating blockchain, this approach enhances data integrity and protects against unauthorized access. Compared to existing models, it offers superior security, efficiency, and scalability, demonstrating its practical applicability in healthcare IoT settings [24].

The motivations for this work can be summarized as follows:

• Addressing Vulnerabilities: Many existing hash functions exhibit vulnerabilities to collision attacks, necessitating the development of more secure alternatives to protect data integrity.

• Enhancing Security: There is a critical need for stronger collision resistance in hash functions, particularly in applications where data security is paramount, such as in healthcare IoT systems.

• Improving Performance: Current hash functions often have high resource requirements and slower processing speeds, motivating the search for lightweight designs that can operate efficiently in resource-constrained environments.

• Expanding Output Options: The limitations of existing hash functions, especially regarding fixed output sizes, highlight the need for methods that offer variable output lengths to enhance security.

• Tailoring to Specific Needs: There is a growing demand for hash functions optimized for specific applications, such as healthcare IoT, which require both security and efficiency.

• Combining Best Practices: The motivation to integrate the strengths of various hash functions while incorporating lightweight principles aims to create a more robust and adaptable hashing solution.

These motivations collectively drive the current research to develop a hash function that not only addresses the limitations of existing methods but also meets the evolving demands of modern applications.

This research not only contributes to the field through innovative cryptographic techniques and blockchain integration but also provides a viable, efficient framework that meets the unique demands of secure, low-resource healthcare IoT applications.

The contributions of this paper can be summarized as follows:

• Development of a Novel Collision-Resistant Hash Function: This research develops and implements an innovative, immutable new lightweight hash function that significantly enhances collision resistance and offers flexible output options, addressing critical limitations of existing methods. This advancement ensures greater adaptability and security in various applications, particularly in resource-constrained environments like healthcare IoT systems.

• Enhancing Healthcare Security: Design a secure healthcare system by integrating a collision-resistant hash function in digital signatures, enabling secure signing of patient reports and electronic health records (EHRs), facilitating innovative management, storage, and security approaches in healthcare.

• Designing a secure blockchain framework for healthcare records management systems: Blockchain technology and electronic health record (EHR) systems combine to form a decentralized network for secure patient data management. A robust healthcare system is designed to significantly enhance data protection and integrity, fostering trust in healthcare transactions and addressing the growing need for secure and efficient digital health solutions.

A table of abbreviations has been included in Appendix I to enhance clarity and understanding of the terminology used in the study. This resource is a quick reference guide that helps readers navigate the many acronyms and terms used in the research. By including this table, the study aims to promote accessibility and comprehension for more effective engagement with the content.

The structure of this paper is organized as follows: Sect. 2 provides an overview of the background and related work, discussing existing lightweight hash functions and their limitations in constrained environments. Section 3 presents the proposed lightweight hash function, including its design, security features, and efficiency considerations. Section 4 discusses the simulation results, evaluating the performance of the proposed function in terms of speed, memory usage, and cryptographic robustness. Section 5 explores the implementation of the proposed hash function in healthcare, demonstrating its role in securing data and communications in healthcare applications. Finally, Sect. 6 concludes the paper, summarizing the key findings and suggesting directions for future work, including further optimizations and testing in diverse real-world applications.

Background and related work

This section reviews recent advancements in the integration of blockchain and IoT technologies, particularly focusing on enhancing data privacy, security, and integrity within the healthcare sector and other applications. Key studies include: The authors of [25] examined security challenges in smart healthcare systems that connect multiple devices and aggregate diverse patient data online. While traditional security surveys identified risks, they often overlooked modern technological solutions. This study highlighted recent threats such as man-in-the-middle (MITM) attacks, denial of service (DoS) attacks, data integrity issues, and phishing attempts. It proposed an AI- and blockchain-based secure architecture for analyzing malware and network attacks, achieving 93.14% accuracy with RF models and utilizing blockchain for secure data storage. Future research may focus on enhancing defenses against additional threats like buffer overflow and phishing attacks. In 2024, the authors of [26] introduced blockchain-based deep learning in secure smart home networks (BPDL-SSHN) to integrate IoT devices with secure smart home networks. This cloud-assisted model enhances security, reliability, and privacy by employing blockchain for decentralized data control, thus reducing vulnerabilities associated with centralized servers. It incorporates machine learning for threat detection, achieving 98.91% accuracy in detecting malicious activities, outperforming alternative methods. Future work may explore scalability, real-time recognition, and enhanced privacy measures. The study in [27] explored the integration of blockchain and IoT technologies to improve data privacy, security, and integrity in nuclear energy applications. It emphasized blockchain’s immutability and transparency for robust data validation, alongside IoT’s real-time data collection capabilities. A secure data management framework was developed, incorporating encryption and integrity checks, significantly enhancing security in the nuclear sector, although challenges in scalability and resource constraints remain. The research in [28] introduced a secure digital healthcare system that integrates blockchain-based IoT, deep learning, and biomimetic algorithms to protect sensitive patient data and enhance early disease detection. The system achieved an accuracy of 88.4%, outperforming existing standards, but noted challenges such as computational overhead in real-time IoT data processing and ethical considerations regarding patient consent. Ref. [29] proposed a secure IoT and healthcare blockchain model integrating a swarm-optimized Bayesian normalized neural network (SSO-BNNN) for real-time, scalable, and secure data processing. The model demonstrated high sensitivity (98%), specificity (98%), and accuracy (99%) in secure medical data handling, with future work exploring dictionary-based encoding techniques for enhanced efficiency. The study in [30] introduced the optimal deep learning-based secure blockchain (ODLSB) model for IoT and healthcare diagnostics, achieving high performance with 92.75% sensitivity, 91.42% specificity, and 93.68% accuracy, and optimizing blockchain hash compression. Future work may investigate dictionary-based encoding for greater efficiency. In [31], the HEART framework was developed to secure wearable healthcare data against cyber threats, achieving 93% training accuracy and 92.92% testing accuracy. Future improvements may include integrating ensemble models for enhanced robustness against ransomware. The study in [32] presented SBAKM-HS, a blockchain-based authentication and key management mechanism for IoMT healthcare applications, demonstrating superior performance in key management and authentication. In 2023, the authors of [33] introduced a blockchain-enabled secure smart health monitoring system (BSSHM) that integrates health sensors for real-time data collection, verified for resilience against cyberattacks. Future work may focus on enhancing features and integrating big data analytics. The research in [34] examined blockchain technology’s benefits in healthcare IoT applications, discussing trade-offs regarding scalability, privacy, and resource utilization. A choice matrix was developed to link healthcare IoT use cases to optimal blockchain structures. The study in [35] introduced a blockchain-empowered federated learning-based intrusion detection system (BFLIDS) to address security challenges in IoMT, achieving up to 98.21% accuracy in intrusion detection. The study in [36] proposed a blockchain-based zero-trust model within a fog-cloud computing framework for medical data security, confirming effectiveness in ensuring data fidelity and privacy. The research in [37] introduced HLOChain, a hierarchical blockchain framework for IoT, demonstrating improved performance in ledger storage cost and transaction confirmation latency. The study in [38] integrated collaborative threat intelligence with blockchain and machine learning, enhancing IoT security frameworks through real-time validation of threats. The authors in [39] examined how blockchain can enhance IoT security through decentralized data storage, proposing a secure computational model that integrates enhanced McEliece encryption. Ref. [40] introduced BC4Regu, an information management framework that enhances supply chain finance by integrating blockchain technology and IoT, improving transparency and collaboration. The authors of [41] proposed a group theory-based binary spring search (GT-BSS) algorithm integrated with a hybrid deep neural network to secure patient health records, demonstrating superior security and efficiency. The study in [42] focused on enhancing security and privacy in IIoT networks through a blockchain-based deep learning framework, achieving lower latency and reduced computational costs. According to [43], the study enhanced data privacy in IoT-based healthcare applications by integrating homomorphic encryption with blockchain, demonstrating superior accuracy in medical data prediction.

Table 1 gives a comprehensive summary of these studies, including references, topics, technologies, major contributions, limitations, and future directions.

Table 1. An overview and synopsis of recently related work

References	Focus area	Technologies used	Key contributions	Limitations	Future directions
[25]	Security in smart healthcare	AI, Blockchain	Proposed architecture with 93.14% accuracy in malware detection, secure data storage with blockchain	Limited to specific threats (e.g., DoS, MITM)	Defenses for buffer overflow, phishing
[26]	Secure smart home networks	Blockchain, Deep Learning, IoT	BPDL-SSHN model achieved 98.91% accuracy in threat detection	Scalability challenges	Focus on real-time recognition, privacy
[27]	Data security in nuclear energy	Blockchain, IoT	Framework enhances data privacy and security	Scalability, resource constraints	Standardization, post-quantum cryptography
[28]	Digital healthcare security	Blockchain, IoT, Deep Learning	88.4% accuracy in disease monitoring; optimized processing	Real-time IoT data processing overhead	Hyperparameter optimization for efficiency
[29]	Healthcare IoT data processing	Bayesian Network, Swarm Optimization	99% accuracy, high sensitivity and specificity	Potential inefficiency in some encryption schemes	Dictionary-based encoding for security
[30]	IoT and healthcare image sharing	Blockchain, Deep Learning, OPSO	High accuracy (93.68%) with optimized image transmission	File compression challenges	Explore dictionary-based encoding for optimization
[31]	Wearable healthcare data security	Blockchain, AI	92.92% accuracy in malicious detection, scalable with 6G	Limited ensemble model testing	LSTM-GRU ensemble for ransomware defense
[32]	Authentication in IoMT healthcare	Blockchain, Key Management	Outperformed existing schemes in key management	Limited real-world validation	Improve framework for scalability
[33]	Smart health monitoring	Blockchain, Real-time IoT	Verified security, testbed implementation for real-time use	Limited big data analytics integration	Incorporate big data in cloud environment
[34]	Healthcare IoT security frameworks	Blockchain, Cryptography	Developed choice matrix for blockchain in IoT	Trade-offs in scalability, privacy	Refine architectures for emerging technologies
[35]	IoMT Intrusion Detection	Blockchain, Federated Learning	98.21% accuracy in intrusion detection	Scalability constraints	Enhance adaptability to new threats
[36]	Medical data security in IoMT	Blockchain, Fog-Cloud	Reliable data auditing with zero-trust model	Computational efficiency	Integrate AI and 5G, advanced encryption
[37]	IoT device trust and security	Blockchain, Lightweight PoR	Improved transaction confirmation latency, ledger cost	Resource limitations in low-power devices	Smart contracts for privacy and efficiency
[38]	IoT threat intelligence	Blockchain, ML	Enhanced accuracy with ML and blockchain validation	Requires comprehensive evaluations	Quantify impact on IoT security frameworks
[39]	IoT data security	Blockchain, SHA-256	Significant improvement in PDR, delay, and throughput	High energy costs	Explore swarm-based optimization
[40]	SCF information management	Blockchain, IoT	Improved transparency, reduced SCF costs	Lacks legal exploration of BCT	Address legal challenges, expand BCT adoption
[41]	IoT Security in Healthcare	GT-BSS, Hybrid DNN, Blockchain, Homomorphic Encryption	Secure PHR storage, privacy-preserving search, improved efficiency and transparency	Limited scalability, dependency on Hyperledger Fabric	Enhancing scalability, integrating quantum security
[42]	Security and Privacy in IIoT Networks	Blockchain, Smart Contracts, VAE, BiLSTM, CNN, Homomorphic Encryption	Secure entity verification, intrusion detection, reduced latency (20 ms)	Computational overhead in hybrid deep learning	Optimizing performance, expanding cross-domain applications
[43]	Data Privacy in IoT Healthcare	Blockchain, Smart Contracts, Homomorphic Encryption, ODNN-GRU	Secure encrypted storage, access control, accurate medical data prediction	High computational complexity	Reducing complexity, real-time deployment in healthcare

Briefly, this section reviewed various recent studies related to the integration of blockchain and IoT technologies to enhance data privacy, security, and integrity, especially within healthcare and other critical sectors. While these studies offer promising solutions to modern challenges, several limitations and unresolved issues remain. Notably, many studies face scalability concerns, particularly when it comes to handling large-scale real-time data, as well as challenges in resource constraints and computational overhead. Furthermore, there is a need for standardized frameworks and practical validation to ensure the applicability of proposed solutions across diverse systems and environments. Ethical considerations, such as patient consent and equitable access, as well as the integration of emerging technologies like AI and 5G, also require more exploration to ensure the robustness and fairness of these systems. Future research directions could focus on improving system efficiency, enhancing defense mechanisms against evolving security threats, and addressing legal and privacy concerns that arise in blockchain and IoT integration.

Proposed lightweight hash function

In this section, we explain the details of the proposed lightweight hash function that will be used later in the designed healthcare system. The hash function is tailored to fit IoT-limited capability devices. The function design is shown in Fig. 1. As can be seen in the figure, the input to the hash function is the message (X). X is considered a variable length while the hash function works on only input with a fixed size of 512 bits. For compatibility issues, padding is utilized to form multiple 512 bits per message (X), X*. Then, the padded message (X*) is divided into multiple M messages denoted as, M₁, M₂,…,M_n that are sequentially processed by the compression H-function. Therefore, the hash function is designed based on four stages: padding, partitioning of the padded message, calculations using the compression hash function, and final hash generation. The padding is used to fit the required hash input, while the second stage is the message division into M messages. The padding is achieved by inserting a specific pattern of zeros between two 1’s. This guarantees the padding of borders. Otherwise, the 10 ∗ 1 padding technique can be employed. Those M messages are fed to the compression hash function for the final stage, which is the final hash generated.

Fig. 1 [Images not available. See PDF.]

Design of the proposed lightweight hash function

Algorithm 1 determines the number of padding bits required by calculating the remaining bits after dividing the message length (X) by the block size (M). Subsequently, the determined number of bits (P) is appended to the input message, forming X*. This represents the input message after padding and prepares it for further processing.

In the second stage of the proposed algorithm, the padded message (X*) of multiple block sizes is divided into 512-bit blocks of equal size (M₁, M₂,…,M_n). Four 128-bit words, i.e., representing a 512-bit block, and the message block M_i is represented by the words, $M_i^1$, $M_i^2$, $M_i^3$, and $M_i^4$. Each message block is subjected to the compression hash function calculation.

The compression hash function (H) accepts each message block (M_i) to be processed. The overall process can be summarized in five steps: block splitting, selecting, nonlinear layer (substitution box), linear layer (permutation box), and key mixing.

• Block Splitting: A 512-bit block of the message (M₁) has been split into four 128-bit sets which are $M_1^1$, $M_1^2$, $M_1^3$, and $M_1^4$.

• Sub-Block Selector: A multiplexing layer consists of four multiplexers, which are controlled by PRNG to forward a single distinguish sub-block $M_1^1$ or $M_1^2$ or $M_1^3$ or $M_1^4$ at a time. The control circuit is designed to ensure that the output of each multiplexer is unique and did not appear before in another multiplexer.

• Nonlinear Layer (Substitution Box): The nonlinear transformation layer consists of 32 ($4\times 4$) substitution boxes. Each box replaces a 4-bit input with another 4-bit output. In our proposed design, a feather S-box is utilized

• Linear Layer (Permutation Box): The linear transformation layer accepts 128 bits from the substitution transformation layer and then rearranges these bits to be permuted to the next step.

• Mix Key Layer: The mixing operation is processed between the subkeys and the output sub-blocks of the linear layer.

The proposed design uses the XOR function for the pair of consecutive outputs, resulting in a 128-bit stream. Two 128-bit streams are combined to generate the final hash, and the final hash is 256 bits is generated. Furthermore, the design is flexible to generate different hash lengths, including 256 and 512.

Figure 2 depicts the flowchart of the whole process of implementing the proposed lightweight hash algorithm. The hashing process begins with padding the input message X to ensure compatibility with the fixed block size of 512 bits. The padded message is then divided into smaller blocks, which are processed sequentially through the compression hash function. Each block undergoes substitution (nonlinear transformation), permutation (linear transformation), and key mixing to generate intermediate hash values. The final hash is computed by concatenating the outputs of these operations. Validation of the generated hash values was performed by comparing them against known test vectors to ensure correctness and collision resistance.

Fig. 2 [Images not available. See PDF.]

Flowchart of implementing the proposed lightweight hash algorithm

The following is the explanation of the security of the proposed hash function, including the linear, nonlinear layers, and analysis of the proposed lightweight hash function design. These subsection investigates the security of the proposed hash function.

Nonlinear layer (substitution box)

A lightweight substitution box (S-box) is designed for an efficient hash function. It ensures the nonlinearity of the output and is unpredictable. So, instead of a large S-box, our design uses a small S-box with many input–output pairs (e.g., $8\times 8$); however, a smaller robust S-box with a limited number of input–output pairs (e.g., $4\times 4)$ can be used. Certainly, the size of the hash function is reduced, and security is maintained. The values of the Feather S-box are chosen based on the algorithm and toolbox proposed in [44], as shown in Fig. 3.

Fig. 3 [Images not available. See PDF.]

Circuit diagram of feather S-box [41]

The Feather S-box area cost was calculated using Fig. 3’s S-box circuit diagram. The Feather S-box’s circuitry is composed of two-input XOR gates and 2-input AND gates. The Feather S-box has one of the smallest footprints, with a total area cost of 18A_XOR + 9A_AND. The total area of the S-box is determined by GE (gate equivalence), and the circuit has been implemented in VERILOG HDL using TSMC 0.18 $\mu m$ standard cell library and has an area of 15.68 GE ($\mu m^2$) [45].

In [45], an experimental analysis of the Feather S-box is carried out using simulations with the SET Match tool. The purpose of this simulator is to examine the cryptographic properties of S-boxes and Boolean functions. This tool is written in the ANSI C programming language. The outcomes have been proven and validated as correct and valid. Analytical metrics include nonlinearity, algebraic resistance, robustness to differential cryptanalysis, confusion coefficient variance for differential power analysis, and whether balanced and SNR (DPA). The complete study of the security analysis of the feather S-box in comparison to others is given in [45]. Figure 4 illustrates the cryptographic properties of the Feather S-box used in the proposed hash function. Key metrics such as nonlinearity, algebraic immunity, and resistance to differential cryptanalysis are highlighted. The results confirm that the Feather S-box achieves a balance between security and efficiency, making it suitable for lightweight cryptographic applications.

Fig. 4 [Images not available. See PDF.]

Simulator outcome of the feather S-box using a set tool

The effectiveness and robustness of the Feather S-box are evaluated through a security analysis utilizing diverse metrics. The Feather S-box has an input size (M) and output size (N) of 4, implying its functionality on a 4-bit input and generation of a 4-bit output. The cryptographic component in question is categorized as a balanced S-box due to its manifestation of a fair balance of 0s and 1s in its truth table.

The metric of nonlinearity holds significant importance as it measures the extent to which the S-box deviates from a linear function. The nonlinearity value of 4 achieved by the Feather S-box in this case indicates a notable deviation from linearity. The evaluation of resistance against linear and differential attacks, known as correlation immunity, has yielded a measurement of 0 for the S-box. This indicates a lack of correlation immunity in the S-box. The Feather S-box possesses an absolute indicator of 16, which serves as an upper bound for the output bits’ values. The value of the sum of the square indicator, which is a metric utilized to evaluate the cryptographic features of the S-box, has been determined to be 1024. The S-box can be described by an algebraic expression of maximum degree 3, as indicated by its algebraic degree.

The algebraic immunity of the Feather S-box is demonstrated to be 2. The assessment of the S-box’s capacity to function as a flawless random permutation is established as 3.533 through the transparency order. The evaluation of the propagation characteristic yields a value of 0, which denotes the average Hamming weight change between the input and output. Also, the S-box has been verified to conform to the strict avalanche criterion (SAC), which implies that a modification of a single bit in the input leads to a change of multiple output bits. The system exhibits a property of having a unique state of equilibrium whereby every input produces an identical output. In contrast, there are no fixed points that are opposite, indicating that there are no inputs that result in outputs that are exactly the inverse.

The value of composite algebraic immunity, which indicates the level of protection against composite field attacks, has been established as 2. The S-box demonstrates a resilience of 0.625 in the face of differential cryptanalysis, a commonly employed technique in cryptanalysis. The value of delta uniformity is 6, indicating the minimum number of differentials necessary to encompass all possible input differences. The signal-to-noise ratio (SNR) has been computed for differential power analysis (DPA) and has yielded a value of 2.399. This value serves as an indicator of the susceptibility of the S-box to power analysis attacks. Finally, the variance of the confusion coefficient, which quantifies the degree of variability in confusion coefficients across distinct input differences, is evaluated to be 0.4572.

Linear layer (permutation box)

In modern lightweight cryptographic algorithms, the linear layer is one of the most important blocks directly impacting the algorithm’s strength. Here, our emphasis on hardware efficiency necessitates a linear layer that can be implemented with the fewest processing elements, i.e., transistors. The permutation is the best implementation for this layer because it uses no space and only swaps wires. Table 2 gives the bit permutation used in our proposed hash algorithm. Bit i of the input state is moved to bit position P(i). The overall permutation process introduced by the permutation table is shown in Fig. 5.

Table 2. Permutation table of the proposed hash algorithm

S-box	S0								S1
i	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
P(i)	8	24	40	56	72	88	104	120	0	16	32	48	64	80	96	112
S-box	S2								S3
i	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31
P(i)	9	25	41	57	73	89	105	121	1	17	33	49	65	81	97	113
S-box	S4								S5
i	32	33	34	35	36	37	38	39	40	41	42	43	44	45	46	47
P(i)	10	26	42	58	74	90	106	122	2	18	34	50	66	82	98	114
S-box	S6								S7
i	48	49	50	51	52	53	54	55	56	57	58	59	60	61	62	63
P(i)	11	27	43	59	75	91	107	123	3	19	35	51	67	83	99	115
S-box	S8								S9
i	64	65	66	67	68	69	70	71	72	73	74	75	76	77	78	79
P(i)	12	28	44	60	76	92	108	124	4	20	36	52	68	84	100	116
S-box	S10								S11
i	80	81	82	83	84	85	86	87	88	89	90	91	92	93	94	95
P(i)	13	29	45	61	77	93	109	125	5	21	37	53	69	85	101	117
S-box	S12								S13
i	96	97	98	99	100	101	102	103	104	105	106	107	108	109	110	111
P(i)	14	30	46	62	78	94	110	126	6	22	38	54	70	86	102	118
S-box	S14								S15
i	112	113	114	115	116	117	118	119	120	121	122	123	124	125	126	127
P(i)	15	31	47	63	79	95	111	127	7	23	39	55	71	87	103	119

Fig. 5 [Images not available. See PDF.]

Overall permutation process

The permutation table in Table 2 is derived using a predefined bit-shuffling algorithm that ensures a uniform distribution of input bits across the output state. The mathematical explanation for the permutation process is provided in Table 3.

Table 3. Permutation mathematical expressions

S-box	Mathematical formula	S-box	Mathematical formula
$S_0$	$P\left(i\right)=(8+16i)mod128$	$S_1$	$P(i)=16(i-8)mod128$
$S_2$	$P\left(i\right)=(9+16i)mod128$	$S_3$	$P(i)=[16(i-24)+1]mod128$
$S_4$	$P\left(i\right)=(10+16i)mod128$	$S_5$	$P(i)=[16(i-40)+2]mod128$
$S_6$	$P\left(i\right)=(11+16i)mod128$	$S_7$	$P\left(i\right)=[16(i-56)+3]mod128$
$S_8$	$P\left(i\right)=(12+16i)mod128$	$S_9$	$P\left(i\right)=[16(i-72)+4]mod128$
$S_{10}$	$P\left(i\right)=(13+16i)mod128$	$S_{11}$	$P\left(i\right)=[16(i-88)+5]mod128$
$S_{12}$	$P\left(i\right)=(14+16i)mod128$	$S_{13}$	$P\left(i\right)=[16(i-104)+6]mod128$
$S_{14}$	$P\left(i\right)=(15+16i)mod128$	$S_{15}$	$P\left(i\right)=[16(i-120)+7]mod128$

Figures 6 and 7 depict the hardware architectures for generating 256-bit and 512-bit hash values, respectively. The process involves dividing the input into 128-bit blocks, applying substitution and permutation layers, and performing XOR operations with subkeys. Figure 6 illustrates the initial hash function of the proposed hash function family. The process entails dividing the input into four sets of 128 bits, which totals 512 bits, controlled using the RC4 method and four multiplexers. The nonlinear layer consists of thirty-two 4-bit S-boxes fed through the multiplexers. Then, the linear layer, represented by a permutation box combined with the keys, is processed through four XOR operations. The resultant outputs from these operations are then XORed together, with the first and third outputs being XORed and the second and fourth outputs being XORed. Finally, by concatenating the outcomes of these two XOR operations, a 256-bit hash value is obtained. Figure 7 depicts the second member of the new hash function family, the input is split into sixteen sets of 128-bit blocks, totaling 2048 bits. Likewise, sixteen multiplexers are regulated using the RC4 technique. The multiplexers feed thirty-two 4-bit S-boxes into the nonlinear layer. Next, proceed to a linear layer of the permutation box. The linear layer’s output is then mixed with Key using sixteen XOR operations. After that, XOR is applied to the first, second, third, and fourth sets. The fifth, sixth, seventh, and eighth sets are XORed in a similar manner. Likewise, the ninth, tenth, eleventh, and twelfth sets are also XORed. The twelfth, fourteenth, fifteenth, and sixteenth sets are XORed in a similar manner. Ultimately, a 512-bit hash value is produced by concatenating the four XOR operations.

Fig. 6 [Images not available. See PDF.]

Hardware architectures of swift-256 hash

Fig. 7 [Images not available. See PDF.]

Implementation of lightweight hardware architecture with a 512-bit hash

The validation of a proposed hash function is intrinsically linked to its collision resistance, a critical property that ensures the security and reliability of cryptographic applications. Collision resistance refers to the difficulty of finding two distinct inputs that produce the same hash output, which is vital for maintaining data integrity. When a hash function is validated, it must demonstrate that it can effectively minimize the chances of collisions, thereby safeguarding against potential vulnerabilities that could be exploited in various cybersecurity contexts. Therefore, it is crucial to precisely compute the chance of a hash collision. The computation of the hash collision probability is explained in Algorithm 2.

The proposed hash algorithms are distinguished primarily by their output lengths, specifically the 256-bit and 512-bit variants. These lengths are crucial as they directly influence the likelihood of collisions, which is a critical aspect of hash function performance. As previously discussed, a robust hash function must effectively minimize collision occurrences to ensure data integrity and security. The relationship between the number of output bits and collision probability is significant; for instance, a 256-bit hash output theoretically reaches a collision probability of 100% when the number of hashed values exceeds $12\times {10}^{38}$, while a 512-bit hash output reaches this threshold at $4\times {10}^{77}$ (see Fig. 8). This stark difference in collision resistance underscores the importance of selecting an appropriate hash length based on the expected volume of data to be hashed, thereby enhancing the overall security of the hashing process. To avoid collisions in the proposed hash algorithms, it is crucial to ensure that the number of hashed values remains below specific thresholds: for a 256-bit hash, the hashed values must not exceed $12\times {10}^{38}$, and for a 512-bit hash, they should remain below $4\times {10}^{77}$. Exceeding these limits would result in a 100% likelihood of collisions, thereby compromising the integrity of the hash function.

Fig. 8 [Images not available. See PDF.]

Collision probability versus the output hash (a. collision probability of 256-bit output and b. collision probability of 512-bit output)

Analysis of the proposed lightweight hash function design

The lightweight design of the proposed hash function ensures minimal computational overhead by utilizing a simplified substitution–permutation network that reduces the number of required operations. This design incorporates a 4-bit S-box and permutation layers optimized for hardware efficiency, requiring fewer gates and lower power consumption. Moreover, integrating IoT devices is facilitated through a modular architecture that offloads computationally intensive tasks, such as block validation, to more capable nodes within the blockchain network. This approach ensures that resource-constrained IoT devices, such as wearable sensors, can operate efficiently without compromising security or battery life. The complete analysis of the proposed design is provided as follows:

1. Nonlinear Layer (Substitution Box): Choosing an S-box in cryptographic algorithms directly impacts security, efficiency, and hardware/software performance. The Feather S-box is a lightweight 4-bit S-box designed for low-resource environments, whereas the AES S-box is an 8-bit S-box optimized for strong cryptographic security. Below is a detailed comparison of their advantages.

   1. Lower Hardware Complexity: The lookup table (LUT) size of a 4-bit S-box is only 16 entries (2⁴), whereas an 8-bit S-box requires 256 entries (2⁸). Requires fewer logic gates, leading to reduced transistor count. Less circuit area, making it ideal for resource-constrained devices such as smart cards and IoT sensors.

   2. Faster Computation: Since it processes 4-bit input instead of 8-bit, the substitution operation is inherently faster. Lower latency in hardware implementations due to reduced logic depth. Beneficial for applications requiring real-time encryption, such as wireless sensor networks (WSN).

   3. Lower Power Consumption: The reduced number of logic gates and simpler structure lead to less energy usage. Important for battery-powered IoT devices, where energy efficiency is a key requirement.

   4. Easier Implementation in Hardware: Smaller S-box means reduced complexity for hardware-based encryption. Less demand for FPGA/ASIC resources, making it viable for devices with constrained silicon area.

   5. Improved Side-Channel Resistance: Smaller and more randomized S-boxes are less vulnerable to side-channel attacks (SCA) such as differential power analysis (DPA). Since the S-box is simpler, its power consumption variation is lower, making it harder to exploit power-based side-channel attacks.

   6. The Feather S-box is optimized for lightweight cryptography, especially for low-power applications in IoT, RFID, and embedded systems.

   A summary of the benefits of 4-bit S-box (Feather) versus 8-bit S-box (AES) was provided in Table 4.

   Table 4. Feather S-box vs. AES S-box

   Feature	Feather S-box (4-bit)	AES S-box (8-bit)
   Lookup Table Size	16 entries (24)	256 entries (28)
   Gate Count (Hardware Complexity)	Low	High
   Computation Speed	Faster (low latency)	Slower (more operations)
   Power Consumption	Lower (efficient for IoT)	Higher (requires more power)
   Security Level	Moderate (lightweight apps)	High (resistant to major attacks)
   Side-Channel Attack Resistance	Better (lower power variations)	Moderate (higher power consumption)
   Suitability for IoT	Ideal	Too resource-intensive
   Suitability for High-Security Apps	Moderate (healthcare, education,..)	Ideal (military, banking …)

2. Linear Layer (Permutation Box): The permutation is a rearrangement process. A typical 8 × 8 P-box (used with 8-bit S-boxes) requires more complex bitwise operations and logical circuits. A wire swap permutation is simply a direct swapping of bits at predefined positions. By using fewer transistors, power consumption is reduced throughout the implementation process. Here, diffusion process can be implemented by simple wiring and costs no area and faster execution in hardware.

In the rapidly evolving landscape of IoT-based healthcare systems, ensuring data security and integrity is paramount. Hash functions play a critical role in safeguarding sensitive information by enabling secure data transactions and maintaining privacy. Table 5 compares various hash functions, including the SHA family, highlighting their security strengths and weaknesses. By identifying the limitations of existing techniques, we can better position our proposed method, which aims to enhance security and computational efficiency, making it particularly suited for resource-constrained IoT environments. The goal is to demonstrate how our approach improves upon traditional methods while addressing the unique challenges posed by healthcare IoT systems.

Table 5. Evaluation of different hash functions

Hash Function	Security Strengths	Weaknesses	Proposed Method Improvements
SHA-1	Widely used and well documented	Vulnerable to collision attacks	Enhanced collision resistance through improved design
	Fast processing speed	Limited to 160-bit output	Offers 256-bit and 512-bit output for better security
SHA-256-	Strong collision resistance	Slower than SHA-1	Lightweight design reduces computational overhead
	Good performance for most applications	Higher resource requirements	Optimized for IoT devices with limited resources
SHA-3	Flexible and secure	Less adoption and trust compared to SHA-2	Tailored for healthcare IoT security needs
	Does not suffer from the same weaknesses as SHA-1	Not yet widely implemented	Combining the best features from SHA-3 with lightweight principles
Proposed Method	Lightweight design ensures low overhead	Needs thorough evaluation against all attack vectors	Improved security features tailored for healthcare IoT
	Enhanced resistance to common attacks	Still under evaluation in diverse real-world scenarios	Provides specific optimizations for IoT healthcare applications

Accordingly, the proposed system introduces a novel lightweight hash function optimized for IoT healthcare applications. Unlike prior approaches that adapt existing cryptographic algorithms, our method is designed from the ground up to balance security and computational efficiency. Key innovations include using a compact substitution–permutation network, a flexible hash length mechanism (256/512 bits), and seamless integration with blockchain for decentralized data management. Compared to existing security systems, our approach achieves a reduction in latency, enhanced scalability, and resistance to common attacks such as collision, sybil, and differential cryptanalysis. The complete evaluation metrics of the proposed design is provided in the Simulation results section.

Simulation results

This section is dedicated to the simulation results of the proposed system with its three components: the proposed hash function, blockchain, and healthcare-designed system. The section shows two different measures which are performance measures and security measures, as stated before.

Experimental setup

The experiments were performed on a desktop computer equipped with an Intel i7-4790T CPU running at 2.70 GHz and 8GB of RAM. The proposed hash function was implemented using Python, and blockchain simulations were carried out with software agents. The process involved several key steps: generating test data for healthcare IoT devices, applying the proposed hash function to the data, simulating blockchain transactions using the hashed data, and then measuring performance metrics such as throughput, response time, and CPU utilization. The results revealed that the system achieved an impressive throughput of 1100 requests per second and a response time of 125 ms for 5500 operations, highlighting its potential suitability for real-time healthcare applications.

Evaluation metrics

Different criteria, including throughput, response time, scalability, and resource utilization, can be used to measure the overall system’s performance and security. Throughput measures the number of transactions processed per time unit, while response time is related to the time it takes to process a transaction. Scalability is the ability to add new users to the system without significant degradation in performance. On the other hand, resource utilization means monitoring the computational resources, such as memory and CPU.

Performance measures

The overall system is examined with the 1500, 2500, 3500, 4500, and 5500 operations, as shown in Fig. 9. The figure shows response time, throughput, scalability, and utilization (CPU). As can be seen in the figure, as expected, as the number of operations increases, the performance measure increases. However, the response time increases linearly with the number of operations, with values of 95, 105, 115, 120, and 125 ms, respectively. This indicates that the system response time increases within a controlled environment.

Fig. 9 [Images not available. See PDF.]

System performance matrix

The throughput seems associated with the number of operations. Again, the increase in the number of operations is almost straight, with 550 requests per second, 750 requests per second, 900 requests per second, 950 requests per second, and 1100 requests per second, respectively. Therefore, the system is able to handle the increase in the number of operations per second as the workload increases.

In terms of scalability, the system is going linear as other performance measures; it has a slightly fluctuating scalability percentage of 88, 82, 78, 82, and 88%, respectively. Once more, the scalability measure shows that the system is able to handle the increased operations as the workload increases, with slightly fluctuating scalability.

Finally, CPU utilization is increasing as the number of operations increases, with percentages of 65, 75, 80, 83, and 85%, respectively. This indicates that the system is approaching higher levels of utilization.

Figure 10 shows the block creation time, which proves that the proposed system can process transactions swiftly. For instance, the first transaction in the system takes 0.164 s, while the average creation time is 0.3 s. This indicates that the system swiftly creates transactions, and it is suitable for real-time healthcare systems, where time is vital for patient care and decision-making.

Fig. 10 [Images not available. See PDF.]

Block creation time

Security measures

Several scenarios are assigned to simulate the operation of the proposed healthcare system and showcase the impact of the proposed security functions on the overall system. Some of the performance measures for the system performance are block creation times, validity, difficulty level reflected by the number of leading zeros, and the proof of work values required for each block. Also, some attacks are simulated, simulating real-world attacks, including transaction tampering, double spending, 51% attacks, and sybil attacks. The block creation time represents the time taken to create each block in the blockchain. The validity means tracking the validity of each block in the blockchain, while the leading zeroes are the number of leading zeros in each blockchain block hash. The proof of work (PoW) algorithm is one of the important components of the blockchain, as it ensures the integrity of the blocks in the blockchain. The algorithm requires a mining process, which requires some computational work to find the required value called proof. The transaction tampering attack refers to modifying the data and recalculating the hash. It demonstrates the blockchain’s vulnerability to data modification. A double spending attack is creating a new block with the same data and trying to insert it at the same position in the blockchain. Another attack is simulated, which is a 51% attack that replaces the entire blockchain with a new controlled malicious one. That simulates the resistance of the blockchain single entity that controls most of the network’s mining power. Finally, Sybil’s attack is the generation of nodes with their own chains and tries to replace some of the main blockchains with some blocks from the Sybil nodes.

First, we examine the proof of work values, which represent the computational efforts invested in securing the blockchain. As the computational work is high, data integrity and immutability are ensured. Figure 11 indicates that computational work is in the range of thousands, which means the system resistance against potential attacks is high. The computational efforts refer to the trustworthiness and reliability of the healthcare records stored in the blockchain.

Fig. 11 [Images not available. See PDF.]

Values of the proof of work

The number of leading zeros in the hash value indicates the system’s security. It is well known that most catch blocks have leading zeros to set certain difficulty levels for the hash function and proof of work algorithm. However, there should be a balance between security and efficiency. The presence of one or two leading zeros could be enough for security and show resistance against potential threats. Figure 12 investigates the number of leading zeroes in the proposed hash test.

Fig. 12 [Images not available. See PDF.]

Leading zeros in the proposed hash test

The successful execution of integrity checks and maintenance of overall consistency of the blockchain is denoted by “ TRUE”; it represents the validity of all blocks. Figure 13 shows the results of the validity checks of the proposed hash. This validation process is a fundamental aspect of any secure healthcare system, as it guarantees the reliability and trustworthiness of the stored healthcare data. Ensuring data accuracy and consistency is paramount in healthcare, where patient safety and decision-making rely on the reliability of the information at hand.

Fig. 13 [Images not available. See PDF.]

Chain validity

In terms of attacks, Fig. 14 shows three types of attacks that were not successful. A Sybil attack was performed with 3 Sybil nodes, and a double spending attack was performed on transaction number 80 while transaction number 50 has been tampered with.

Fig. 14 [Images not available. See PDF.]

Attacks on the healthcare system

Based on the presented results, it is evident that the secure healthcare system utilizing the novel hash and blockchain technology exhibits commendable performance characteristics. The efficient block creation times, robust proof of work values, and appropriate difficulty level for the hash function collectively contribute to a secure and efficient healthcare data management system. These findings provide valuable insights into the system’s potential as a foundation for secure healthcare record-keeping, with implications for improved patient care and data security.

The proposed framework demonstrates a robust capability to manage real-world healthcare data loads, as evidenced by the scalability analysis included in Sect. 4.3. In the context of healthcare, where data is generated at an unprecedented rate from various sources such as electronic health records (EHRs), medical devices, and patient monitoring systems, the ability to efficiently process and analyze this data is crucial. The scalability analysis reveals that our system can effectively handle increasing data loads without compromising performance, which is essential for maintaining the quality of patient care and operational efficiency. One of the key aspects of our approach is its lightweight hash function, which is specifically designed to optimize performance in resource-constrained environments typical of healthcare settings. By minimizing computational overhead, the system can maintain high throughput and low response times even as the volume of incoming data escalates. For instance, the analysis shows that the system can sustain a throughput of 1100 requests per second, which is vital for real-time applications where timely access to patient data can significantly impact clinical decisions. Moreover, the design choices made in the hash function, such as the use of a small S-box and optimized permutation layers, contribute to its ability to scale efficiently. These features allow the system to process data in parallel, distributing the workload across multiple nodes, which is a proven strategy for managing large datasets in healthcare environments. This parallel processing capability aligns with best practices in big data analytics, where distributing and processing data across clusters enhance efficiency and reduce bottlenecks. Additionally, the system’s performance metrics indicate that it can handle heavy workloads while keeping CPU utilization within manageable levels, peaking at 85% during intensive operations. This level of resource management ensures that the system remains responsive and reliable, even under the strain of high data loads, which is critical for healthcare applications that demand continuous availability and performance. In summary, the scalability analysis confirms that our proposed framework is well equipped to handle the complexities and demands of real-world healthcare data loads. By leveraging a lightweight design and efficient processing strategies, the system not only meets the current needs of healthcare providers but also positions itself for future growth as data volumes continue to rise. This adaptability is essential for ensuring that healthcare organizations can harness the full potential of big data analytics to improve patient outcomes and operational efficiencies.

Implementation of the proposed lightweight hash function in healthcare

This section shows the implementation of the proposed lightweight hash function for an effective healthcare scenario. Also, it shows how the hash function is integrated into the blockchain-based security algorithm for patients’ medical information.

Problem description

Consider various measurement equipment attached to patient E, which are essential for continuous health monitoring and collecting real-time physiological data D. It defines a set of measurement instruments, denoted as E and D in the following way:

1

$$E=\{E_1,E_2,E_3,···,E_k,···,E_M\}$$

2

$$D=\{D_1,D_2,D_3,···,D_K,···,D_M\}$$where each device $E_k$ measures specific health parameters, contributing to a comprehensive understanding of the patient’s condition such that 1 ≤ k ≤ M. The notation indicates that M represents the total number of devices, allowing for a structured approach to monitoring patient E’s health status through these specialized instruments. Similarly, we can conceptualize the set of patients, denoted as P, in the following way:

3

$$P=\left\{P_1,,,P_2,,,P_3,,,·,·,·,,,P_j,,,·,·,·,,,P_N\right\}$$where P represents the entire group of patients under consideration, and each patient Pj is identified as a unique individual requiring medical attention such that 1 ≤ j ≤ N. Each patient is associated with specific measurement devices, referred to as k, that are attached to the patient’s body to facilitate the continuous monitoring of vital signs such as heart rate, blood pressure, and temperature.

4

$$P_j\stackrel{\mathit{Associate}}{\leftarrow }{E}_1{,E}_2,E_3,\cdots ,E_k$$

The health data for each patient is carefully gathered through these devices, ensuring comprehensive tracking of their medical status.

5

$$D_k\stackrel{\mathit{Produces}}{\leftarrow }{E}_k$$

This information is then stored within a centralized system, denoted as C, which serves as a secure repository for patient data.

6

$$C\stackrel{\mathit{Store}}{\leftarrow }D$$

An attacker A has the potential to penetrate the centralized system storing patient information, thereby gaining unauthorized access to sensitive medical records.

7

$$A\stackrel{\mathit{Attacks}}{\to }C$$

Once the attacker breaches the system, they can access specific patient medical records denoted as Dj, where j signifies a unique patient identifier.

8

$$A\leftarrow C_{\mathit{Comp}}(Dj)$$

Here, $C_{\mathit{Comp}}$ indicates the compromised version of the centralized system, illustrating that A is able to extract patient data Dj without authorization. And $\forall 1\le j\le N$ indicates that this vulnerability extends to all patients within the system, where N represents the total number of patients.

This series of equations highlights a critical security flaw; without adequate protective measures, the system is susceptible to infiltration, allowing attacker A to acquire patient data for nefarious purposes, potentially leading to identity theft, fraud, or other malicious activities.

To counteract these security challenges, advanced techniques such as encryption, hash functions, and blockchain technology are employed to develop a more resilient and secure method for storing patient information.

The proposed model

Protecting sensitive patient information is critically important for healthcare organizations, especially with the increasing reliance on centralized data systems, which are vulnerable to unauthorized access and data breaches. To address these concerns, a secure healthcare model has been proposed that uses advanced security techniques to safeguard medical records. The core aim of this model is to verify the identity of users before they can access or manage patient data. It relies on digital signatures and blockchain technology to ensure that the data remains authentic, secure, and tamperproof. This approach is essential for maintaining the privacy of sensitive health information and strengthening trust in the healthcare system. The architecture of the proposed system is illustrated in Fig. 15.

Fig. 15 [Images not available. See PDF.]

Architecture of the proposed system

The solution operates in two main phases. The first phase is called authorization verification and safe storage of patient data. It begins with data collection from medical devices connected to patients, such as heart rate monitors, temperature sensors, and blood pressure monitors. These devices continuously generate large volumes of medical data. Once collected, the data is digitally signed to confirm its origin and to prevent any alterations during transmission. The next step is data verification, where the system checks if the data has been correctly signed and remains unchanged. Any data that is altered or originates from an untrusted source is immediately rejected. Only verified data is accepted for storage. The verified data is then securely stored in a private blockchain network, which is accessible only to registered users. In this blockchain, data is stored in a chain of interconnected blocks, with each block containing a time stamp, a unique hash, the hash of the previous block, and a Merkle root—a structure that ensures fast and secure data verification. The second phase focuses on the secure access of patient data using blockchain technology. In this phase, authorized users such as doctors, pharmacists, family members, and remote healthcare providers are given access to the stored patient data. These users interact with the data through a blockchain-based interface. Access and modifications are managed through smart contracts, which are automated programs that enforce rules and allow secure, transparent communication without involving a third party. These smart contracts contain predefined logic and conditions, ensuring that only authorized individuals can read or update patient records.

This two-phase model ensures that patient data is collected, verified, stored, and shared in a highly secure and trustworthy manner, using the combined power of blockchain technology and digital signatures. The following section provides a detailed explanation of each phase.

Phase 1: authorization verification and safe storage of patient data

Digital signatures serve as a robust method for validating the authenticity of digital documents. They provide a way to confirm that the information has not been altered after it was signed, thereby maintaining the integrity of the data. This is particularly important in healthcare, where the accuracy of medical records is vital for patient safety and compliance with regulations.

By implementing digital signatures, healthcare organizations can strengthen their overall security framework. Digital signatures utilize cryptographic techniques to create a unique identifier for each document, which can only be verified by the intended recipient. This not only protects against unauthorized access but also ensures that any changes to the document can be detected.

To ensure the integrity and validity of data before secure storage, medical records Dj are first verified by a separate system S rather than being directly stored in the centralized system C. System S checks the report’s digital signature throughout this phase to confirm that an authorized person signs the medical data report. This procedure reduces the risk of unauthorized access and enhances the overall security architecture of the patient data management system.

9

$$S\stackrel{\mathit{Analyzes}}{\leftarrow }{(D_j)}_{\mathit{Signed}}$$

Once the medical records Dj have been thoroughly verified for authentication, integrity, and safety, they are securely recorded in a blockchain system B.

10

$$B\stackrel{\mathit{Stores}}{\leftarrow }{D}_j$$

Only authorized Auth persons can access this data, such as healthcare providers, specialists, and caregivers.

11

$$Auth\stackrel{\mathit{Access}}{\to }{D}_j$$

Furthermore, the entire system is integrated with a public blockchain network. This connection ensures that if any malicious or compromised data is detected, all participants in the patient’s care are promptly notified. In such cases, the compromised data will be systematically discarded to maintain the integrity of the patient’s medical records. This comprehensive approach not only safeguards patient data against cyber threats but also fosters trust among healthcare providers and patients by ensuring that their sensitive information is managed securely and transparently. Figure 16 depicts the suggested system model. Every patient has measuring equipment attached to their body that continuously tracks their health information. Fake information will be rejected; only valid information will be approved.

Fig. 16 [Images not available. See PDF.]

Patient data verification and storage

Phase 2: secure access of patient data using blockchain technology

The proposed structure ensures that only authorized personnel can view and utilize the patient’s medical records, thereby maintaining confidentiality and protecting sensitive information. The implementation of blockchain technology significantly enhances data security, making it exceedingly difficult for an attacker to gain unauthorized access to the medical records of patient Pj. Moreover, to detect any potential data manipulation or unauthorized alterations, the system continuously monitors incoming patient health data.

Blockchain technology offers a promising solution for enhancing the security and privacy of patient data in healthcare systems. This section uses Fig. 17 as a case study to apply the suggested blockchain algorithm to patient healthcare monitoring. The blockchain network connects doctors and the health authority to medical equipment and sensing devices.

Fig. 17 [Images not available. See PDF.]

A high-level architecture for the proposed blockchain-based healthcare system

The following is the framework of the suggested blockchain-based algorithm for patient health monitoring:

Doctors and the health authority are connected to medical equipment and sensing devices via the blockchain network. The data that passes across the blockchain network will be subject to some smart contracts, including registration, authentication, patient monitoring, consent management, and drug prescription verification. The patient’s data will be protected from any unauthorized usage through these smart contracts.

Medical equipment and sensing devices: patient measures are read by sensing devices in medical equipment. In order to gain access to the login for the decentralized storage systems linked to the health authority, the medical equipment first transmits device registration to the health authority over the blockchain network. The blockchain network sends an emergency notice if the login attempt is unsuccessful; if it is successful, the registration is successful. When doctors request the updated, signed measurement reports, which include the patient’s physiological parameters—the medical equipment then distributes them to them over the blockchain network. After phase 1 verification checking is complete, the medical equipment can then save the final measurements report immediately in the decentralized storage systems.

The doctors: Initially, the doctors will submit the registration form to have access to the data kept in the health authority’s decentralized storage systems via the blockchain network. The blockchain network will return an emergency warning if the registration is unsuccessful; if it is successful, the registration is successful. The doctors may then ask the blockchain network for authorization to acquire the measurement data for each patient, and they can update the reports depending on the study of patients. Lastly, in order to update the patients’ periodic reports, the doctors have direct access to the health authority’s decentralized storage systems.

The health authority: Doctors and medical equipment use the blockchain network to conduct all of their activities there. Doctors and medical equipment regularly register with the health authority, which then decides whether to approve or reject these registrations. The requested diagnosis report and medical procedure are sent when the registration is approved. It then gets a request from the doctors to get their consent to share the requested information. Lastly, the reputation score is computed and updated.

Decentralized Storage Systems: These systems are used to store patient information and various treatment methods. It supports a number of storage protocols, including SWARM, STORJ.IO, and IPFS.

The hash function was integrated into a blockchain-based healthcare system, where it was used to generate unique patient identifiers and secure medical records. Hence, healthcare data transactions on the blockchain are validated through a combination of digital signature verification and a proof of work mechanism adapted for lightweight operations. Each transaction is assigned a unique hash identifier, ensuring immutability and traceability. To maintain data privacy, the system employs encryption techniques combined with hash-based access control. Sensitive patient data is encrypted before being hashed, ensuring that only authorized parties with the correct decryption keys can access the raw data. The lightweight nature of the hash function further enhances scalability by reducing the computational burden of transaction processing, thereby maintaining the efficiency and privacy of the system in resource-constrained environments. The results confirm that the implementation meets the design objectives of low computational overhead and high security.

In the proposed healthcare system, doctors can only be added by the contract deployer, ensuring strict access control. The creation and execution process is outlined in Algorithm 3, incorporating cryptographic security measures. First, the system verifies if the user is the contract deployer; if not, an error message is displayed: “Only the contract deployer can add a doctor.” Next, it checks whether the doctor’s ID already exists. If the ID is found in the records, the process is halted, and a message is shown: “Doctor with this ID already exists.” If the ID is not found, the deployer must enter a password, which is compared with the stored hashed password. If the passwords do not match, access is denied with an error message: “Incorrect password. Access denied.” Upon successful verification, the system generates a private key for the doctor and derives the corresponding public key to be stored for authentication purposes. The doctor’s details and node address are then recorded securely in the system. Finally, once all necessary information is stored, a confirmation message is displayed: “New doctor successfully created.”

Algorithm 4 illustrates the process of adding a new patient to the blockchain network, ensuring that only an authorized doctor can perform this action while incorporating cryptographic security measures. First, the system verifies if the doctor is part of an authorized node; if not, an error message stating “Error: Patient cannot be created. Unauthorized doctor.” is displayed. If authorized, the doctor proceeds to enter the patient’s details. The system then generates a unique private key for the patient and derives the corresponding public key to ensure secure identification and access control. A unique patient ID is created using the Proposed Hash Function-256 hash function, enhancing data integrity. The patient’s information is stored in a struct, maintaining organization, and the unique ID is mapped to the struct address, securely linking the patient’s record. The patient’s public key is also stored for future authentication. Once all steps are completed, the system confirms successful patient creation by displaying the message “New patient successfully created.”

The patient’s previous records can now be easily accessed using their unique ID, and Algorithm 5 illustrates the process of securely adding new measurement records to the blockchain network in a decentralized and cryptographically secure manner. First, the system enters the patient’s ID and attempts to access the patient’s node; if the patient does not exist, an error message stating “Error: Patient does not exist.” is displayed. If the patient exists, a new record request is created, and the system verifies that the doctor’s public key matches an authorized entity. A unique digital signature is then generated using the doctor’s private key to authenticate the entry. The entered data, such as vital signs and test results, is stored in the patient’s node. The measurement data is then converted into JSON format for easy processing and signed with the patient’s private key, ensuring data integrity and authenticity. This JSON file is uploaded to IPFS (InterPlanetary File System), a decentralized storage network, which returns a unique hash (CID) as a reference for later retrieval. The IPFS hash and digital signature are stored on the blockchain, ensuring immutability and tamperproof records while preventing large files from being stored directly on-chain. Finally, the system displays the final records along with the stored data reference, confirming successful record creation with the message “New records successfully added.”

The following comparison table (Table 6) illustrates the differences between a basic authorization approach and a cryptographic security approach in a blockchain-based healthcare system. The table highlights key factors such as security, privacy, data integrity, authorization, and scalability. While the basic authorization method relies on simple access control, the cryptographic security model enhances protection through private–public key encryption, digital signatures, and decentralized identity verification. As shown in Table 6, the cryptographic approach offers a more secure, tamperproof, and scalable solution for managing healthcare data on the blockchain.

Table 6. Basic authorization vs. cryptographic security in blockchain-based healthcare systems

Aspect	Blockchain-based healthcare system (without keys)	Blockchain-based healthcare system (with private–public keys)
Security	Basic authorization checks (doctor verification)	Uses cryptographic authentication (private–public keys)
Data Integrity	No explicit verification for record integrity	Digital signatures ensure records are tamperproof
Privacy	Relies on general access control (authorized doctor)	Patients and doctors have unique keys for privacy
Blockchain Usage	Stores only hashes in the blockchain	Stores signed data and public keys for verification
Authorization Check	Simple IF conditions (doctor/patient existence)	Ensures valid doctors and patients via keys
Scalability	Works for basic use cases	More scalable for decentralized, multi-user environments

By implementing these robust security measures, healthcare organizations can better protect sensitive patient information from cyber threats, ensuring confidentiality and integrity while maintaining trust in the healthcare system.

Use case scenario for hospital management system (HMS) with blockchain integration

A hospital management system (HMS) is a digital platform that enhances healthcare operations by managing patient records, doctor workflows, and administrative tasks, traditionally using a centralized database for appointments, medical histories, prescriptions, and billing. Integrating blockchain technology transforms HMS into a secure and transparent system, solving issues like data breaches and inefficiencies in record-keeping (see Fig. 18). The system features three user roles: Patients, who can book appointments and access their medical histories; Doctors, who diagnose and update electronic health records; and Administrators, who are responsible for overseeing operations and compliance. Blockchain introduces smart contracts for automating processes such as payment verification and access control, while patients maintain ownership of their data through decentralized identities (DIDs). Key functionalities include secure interactions for recording medical histories, with data encrypted and validated by smart contracts, ensuring integrity and privacy. This integration leads to an auditable, patient-centric healthcare system that enhances security and interoperability by eliminating intermediaries and reducing paperwork, thereby representing the future of efficient healthcare management. Table 7 shows how blockchain is integrated into the HMS use case, as well as the function of system stakeholders.

Fig. 18 [Images not available. See PDF.]

Use case for blockchain-based HMS

Table 7. Hospital management system (HMS) use cases with blockchain integration

Actor	Use Case	Description	Blockchain Integration
Administrator	Register Patient	Adds new patient details to the system	Patient records stored as hashed transactions on blockchain for immutability
	Manage Staff	Handles doctor/nurse onboarding and roles	Staff credentials and permissions stored securely on-chain
	Generate Reports	Produces operational/financial reports	Audit logs stored on blockchain for transparency
Doctor	Schedule Appointment	Books patient consultation slots	Appointment time stamps recorded on blockchain
	Record Medical History	Updates patient diagnoses/treatments	Encrypted EHRs (electronic health records) stored on-chain
	Prescribe Medication	Issues e-prescriptions	Prescription details logged on blockchain to prevent tampering
Patient	View Medical History	Accesses personal health records	Decentralized identity (DID) allows secure access via blockchain
	Pay Bills	Settles hospital invoices	Payments processed via smart contracts for automated verification

In the hospital management system (HMS), key interactions involve two main processes: recording medical history and paying bills. When a doctor submits a diagnosis through the HMS frontend, the backend encrypts the data using SLIM algorithm and generates a hash using the proposed hash algorithm. A smart contract then verifies the doctor’s access rights via an NFT-based credential, and the hash along with the patient’s decentralized identity (DID) is stored on-chain, with a transaction ID (TxID) provided as proof of immutability. For bill payments, the patient initiates the payment, and the smart contract checks the patient’s wallet balance and the validity of the invoice. Once verified, funds are autonomously transferred to the hospital, and a TxID serves as the payment receipt. The system utilizes various blockchain components, including smart contracts for transaction validation, decentralized identities for secure data access, and NFT access tokens for granting doctors permissions. This integration ensures security through encrypted and hashed data, automates payment processes, and maintains compliance with auditable transactions. Figure 19 illustrates the sequence diagram of HMS integrated with blockchain technology.

Fig. 19 [Images not available. See PDF.]

Sequence diagram for blockchain-based HMS interactions

Conclusion and future work

The research highlights the pressing need for advanced hash functions that effectively address the inherent vulnerabilities of existing methods. By focusing on enhancing collision resistance, improving performance, and offering flexible output options, this study aims to redefine standards in cryptographic security. The proposed hash function is not only designed to meet the demanding requirements of critical applications, such as healthcare IoT, but also to ensure that security and efficiency go hand in hand. Furthermore, by integrating the best practices from various hashing techniques and adapting them to the specific needs of modern technology, this research paves the way for a robust and versatile solution. The implications of this work extend beyond theoretical advancements; it promises to significantly bolster data integrity and security in an increasingly interconnected world.

Future work will focus on enhancing the scalability and security of blockchain-driven healthcare IoT systems by optimizing performance for large-scale networks, incorporating AI for anomaly detection, and exploring side-channel attacks and quantum-resistant cryptographic methods. Additionally, evaluating the framework in real-world settings and addressing gas cost and storage optimization for platforms like Ethereum and Hyperledger—through techniques such as data compression, struct packing, and external storage (e.g., IPFS)—can improve efficiency, reduce costs, and ensure the system remains practical and reliable.

Acknowledgements

Not applicable.

Author contributions

The authors declare that the study was conducted in collaboration with each other with equal responsibility. The manuscript was read and approved by all authors.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Availability of data and material

No datasets were generated or analyzed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Abbreviations

BCT

Blockchain technology

BFLIDS

Blockchain-empowered federated learning-based intrusion detection system

BiLSTM

Bidirectional long short-term memory

BPDL-SSHN

Blockchain-based deep learning in the secure smart home network

BSSHM

Blockchain-enabled secure smart health monitoring system

DCGRU

Deep convolutional gated recurrent unit

DoS

Denial of service

DPA

Differential power analysis

EHRs

Electronic health records

GT-BSS

Group theory-based binary spring search

IoMT

Internet of medical things

IoT

Internet of Things

MITM

Man-in-the-middle

ML

Machine learning

NIS

Neighborhood indexing sequence

ODLSB

Optimal deep learning-based secure blockchain

ODNN

Optimized deep neural network

OPSO

Orthogonal particle swarm optimization

PDR

Packet delivery ratio

PHR

Patient health records

PoR

Proof of random

RFID

Radio-frequency identification

SCF

Supply chain finance

SHA

Secure hash algorithm

SNR

Signal-to-noise ratio

SSO-BNNN

Swarm-optimized bayesian normalized neural network

VAE

Variational autoencoder

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Zhang, C; Zhu, L; Xu, C; Lu, R. PPDP: an efficient and privacy-preserving disease prediction scheme in cloud-based e-healthcare system. Futur Gener Comput Syst; 2018; 79, pp. 16-25. [DOI: https://dx.doi.org/10.1016/j.future.2017.09.002]

2. Ali, HM; Liu, J; Bukhari, SAC; Rauf, HT. Planning a secure and reliable IoT-enabled FOG-assisted computing infrastructure for healthcare. Clust Comput; 2022; 25, pp. 2143-2161. [DOI: https://dx.doi.org/10.1007/s10586-021-03389-y]

3. Kumar, PM; Hong, CS; Afghah, F; Manogaran, G; Yu, K; Hua, Q; Gao, J. Clouds proportionate medical data stream analytics for internet of things-based healthcare systems. IEEE J Biomed Health Inform; 2021; 26, pp. 973-982. [DOI: https://dx.doi.org/10.1109/JBHI.2021.3106387]

4. Namasudra, S. Fast and secure data accessing by using DNA computing for the cloud environment. IEEE Trans Serv Comput; 2020; 15, pp. 2289-2300. [DOI: https://dx.doi.org/10.1109/TSC.2020.3046471]

5. Namasudra, S; Deka, GC; Johri, P; Hosseinpour, M; Gandomi, AH. The revolution of blockchain: state-of-the-art and research challenges. Archiv Comput Methods Eng; 2021; 28, pp. 1497-1515. [DOI: https://dx.doi.org/10.1007/s11831-020-09426-0]

6. Kish, LJ; Topol, EJ. Unpatients—why patients should own their medical data. Nat Biotechnol; 2015; 33, pp. 921-924.1:CAS:528:DC%2BC2MXhsVymt7%2FP[DOI: https://dx.doi.org/10.1038/nbt.3340] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26348958]

7. Tariq, N; Qamar, A; Asim, M; Khan, FA. Blockchain and smart healthcare security: a survey. Procedia Computer Science; 2020; 175, pp. 615-620. [DOI: https://dx.doi.org/10.1016/j.procs.2020.07.089]

8. Grassi, L; Khovratovich, D; Schofnegger, M. Poseidon2: A faster version of the poseidon hash function. International Conference on Cryptology in Africa; 2023; Cham, Springer: pp. 177-203.

9. Cheval V, Cremers C, Dax A, Hirschi L, Jacomme C, Kremer S (2023) Hash Gone Bad: Automated discovery of protocol attacks that exploit hash function weaknesses. In: 32nd USENIX Security Symposium (USENIX Security 23)

10. Wu, ST; Chang, JR. Secure one-way hash function using cellular automata for IoT. Sustainability; 2023; 15, 3552. [DOI: https://dx.doi.org/10.3390/su15043552]

11. Al-Zubaidie, M. Implication of lightweight and robust hash function to support key exchange in health sensor networks. Symmetry; 2023; 15, 152. [DOI: https://dx.doi.org/10.3390/sym15010152]

12. Rasool, M; Belhaouari, SB. From collatz conjecture to hash function. Chaos, Solit Frac; 2023; [DOI: https://dx.doi.org/10.1016/j.chaos.2023.114103]

13. Nanda, SK; Mohanty, S; Pattnaik, PK. An optimized 128-bit cellular automata-based hash function for authentication of data at rest and in transit. Int J Electrical Comput Eng; 2023; 13, 1858. [DOI: https://dx.doi.org/10.11591/ijece.v13i2.pp1858-1866]

14. Davda Y (2023) Design of hash algorithm for blockchain security. In: blockchain applications in cryptocurrency for technological evolution. IGI Global

15. Shi, S; He, D; Li, L; Kumar, N; Khan, MK; Choo, KKR. Applications of blockchain in ensuring the security and privacy of electronic health record systems: a survey. Comput Secur; 2020; 97, [DOI: https://dx.doi.org/10.1016/j.cose.2020.101966] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32834254][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7362828]101966.

16. Chenthara, S; Ahmed, K; Wang, H; Whittaker, F; Chen, Z. Healthchain: a novel framework on privacy preservation of electronic health records using blockchain technology. PLoS ONE; 2020; 15, 1:CAS:528:DC%2BB3cXis1emsb7I[DOI: https://dx.doi.org/10.1371/journal.pone.0243043] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33296379][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7725426]e0243043.

17. Kumar, S; Bharti, AK; Amin, R. Decentralized secure storage of medical records using blockchain and IPFS: a comparative analysis with future directions. Secur Priv; 2021; 4, [DOI: https://dx.doi.org/10.1002/spy2.162] e162.

18. Bhushan, B; Kumar, A; Agarwal, AK; Kumar, A; Bhattacharya, P; Kumar, A. Towards a secure and sustainable internet of medical things (IoMT): requirements, design challenges, security techniques, and future trends. Sustainability; 2023; 15, 6177. [DOI: https://dx.doi.org/10.3390/su15076177]

19. Srivastava, J; Routray, S; Ahmad, S; Waris, MM. Internet of medical things (IoMT)-based smart healthcare system: trends and progress. Comput Intell Neurosci; 2022; 2022, 7218113. [DOI: https://dx.doi.org/10.1155/2022/7218113] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35880061][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9308524]

20. Hirose, S. Collision-resistant and pseudorandom hash function using tweakable block cipher. International Conference on Information Security Applications; 2022; Cham, Springer: pp. 3-15.

21. Widhiara, B; Kurniawan, Y; Susanti, BH. RM70: a lightweight hash function. IAENG Int J Appl Math; 2023; 53, 1.

22. Freitag, C; Ghoshal, A; Komargodski, I. Optimal security for keyed hash functions: avoiding time-space tradeoffs for finding collisions. Annual international conference on the theory and applications of cryptographic techniques; 2023; Cham, Springer: pp. 440-469.

23. Hou, P; Shang, T; Zhang, Y; Tang, Y; Liu, J. Quantum hash function based on controlled alternate lively quantum walks. Sci Rep; 2023; 13, 5887.1:CAS:528:DC%2BB3sXnvVehtrc%3D[DOI: https://dx.doi.org/10.1038/s41598-023-33119-w] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37041296][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10090072]

24. Ananth, P; Hu, Z; Yuen, H. On the (im)plausibility of public-key quantum money from collision-resistant hash functions. International conference on the theory and application of cryptology and information security; 2023; Singapore, Springer: pp. 39-72.

25. Alabdulatif, A; Khalil, I; Rahman, MS. Security of blockchain and AI-empowered smart healthcare: application-based analysis. Appl Sci; 2022; 12, 11039.1:CAS:528:DC%2BB38XivVGqurrJ[DOI: https://dx.doi.org/10.3390/app122111039]

26. Alruwaili, FF. Blockchain-powered deep learning for internet of things with cloud-assisted secure smart home networks. IEEE Access; 2024; [DOI: https://dx.doi.org/10.1109/ACCESS.2024.3450796]

27. Rai, HM; Shukla, KK; Tightiz, L; Padmanaban, S. Enhancing data security and privacy in energy applications: integrating IoT and blockchain technologies. Heliyon; 2024; 10, [DOI: https://dx.doi.org/10.1016/j.heliyon.2024.e38917] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39430499][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11490785]e38917.

28. Ramachandraiah, KRD; Bommagani, NJ; Jayapal, PK. Enhancing healthcare data security in IoT environments using blockchain and DCGRU with Twofish encryption. Inf Dyn Appl; 2023; 2, 173185.

29. Chaturvedi, S. IoT-based secure healthcare framework using blockchain technology with a novel simplified swarm-optimized bayesian normalized neural networks. Int J Data Inf Intell Comput; 2023; 2, pp. 63-71.

30. Veeramakali, T; Siva, R; Sivakumar, B; Mahesh, PCS; Krishnaraj, N. An intelligent internet of things-based secure healthcare framework using blockchain technology with an optimal deep learning model. J Supercomput; 2021; 77, pp. 9576-9596. [DOI: https://dx.doi.org/10.1007/s11227-021-03637-3]

31. Jadav, D; Jadav, NK; Gupta, R; Tanwar, S; Alfarraj, O; Tolba, A; Marina, V. A trustworthy healthcare management framework using amalgamation of AI and blockchain network. Mathematics; 2023; 11, 637. [DOI: https://dx.doi.org/10.3390/math11030637]

32. Thapliyal, S; Wazid, M; Singh, DP; Das, AK; Shetty, S; Alqahtani, A. Design of robust blockchain-envisioned authenticated key management mechanism for smart healthcare applications. IEEE Access; 2023; 11, pp. 93032-93047. [DOI: https://dx.doi.org/10.1109/ACCESS.2023.3310264]

33. Thapliyal, S; Singh, S; Wazid, M; Singh, DP; Das, AK. Design of blockchain-enabled secure smart health monitoring system and its testbed implementation. Cyber Secur Appl; 2023; 1, [DOI: https://dx.doi.org/10.1016/j.csa.2023.100020] 100020.

34. Alhija, MA; Al-Baik, O; Hussein, A; Abdeljaber, H. Optimizing blockchain for healthcare IoT: a practical guide to navigating scalability, privacy, and efficiency trade-offs. Indo J Electrical Eng Comput Sci; 2024; 35, pp. 1773-1785.

35. Begum, K; Mozumder, MAI; Joo, MI; Kim, HC. BFLIDS: Blockchain-driven federated learning for intrusion detection in IoMT networks. Sensors; 2024; 24, 4591. [DOI: https://dx.doi.org/10.3390/s24144591] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39065989][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11280944]

36. Sutradhar, S; Majumder, S; Bose, R; Mondal, H; Bhattacharyya, D. A blockchain privacy-conserving framework for secure medical data transmission in the internet of medical things. Decis Anal J; 2024; 10, [DOI: https://dx.doi.org/10.1016/j.dajour.2024.100419] 100419.

37. Xie, Q; Dong, F; Feng, X. HLOChain: a hierarchical blockchain framework with lightweight consensus and optimized storage for IoT. Secur Commun Netw; 2023; 2023, 3412200. [DOI: https://dx.doi.org/10.1155/2023/3412200]

38. Nazir, A; He, J; Zhu, N; Wajahat, A; Ullah, F; Qureshi, S; Pathan, MS. Collaborative threat intelligence: enhancing IoT security through blockchain and machine learning integration. J King Saud Univ-Comput Inf Sci; 2024; 36, [DOI: https://dx.doi.org/10.1016/j.jksuci.2024.101939] 101939.

39. Kumar, S; Singh, A; Benslimane, A; Chithaluru, P; Albahar, MA; Rathore, RS; Álvarez, RM. An optimized intelligent computational security model for interconnected blockchain-IoT system and cities. Ad Hoc Netw; 2023; 151, [DOI: https://dx.doi.org/10.1016/j.adhoc.2023.103299] 103299.

40. Guo, L; Chen, J; Li, S; Li, Y; Lu, J. A blockchain and IoT-based lightweight framework for enabling information transparency in supply chain finance. Digit Commun Netw; 2022; 8, pp. 576-587. [DOI: https://dx.doi.org/10.1016/j.dcan.2022.03.020]

41. Ali, A; Almaiah, MA; Hajjej, F; Pasha, MF; Fang, OH; Khan, R; Zakarya, M. An industrial IoT-based blockchain-enabled secure searchable encryption approach for healthcare systems using neural network. Sensors; 2022; 22, 572. [DOI: https://dx.doi.org/10.3390/s22020572] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35062530][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8779424]

42. Almaiah, MA; Ali, A; Hajjej, F; Pasha, MF; Alohali, MA. A lightweight hybrid deep learning privacy preserving model for FC-based industrial internet of medical things. Sensors; 2022; 22, 2112. [DOI: https://dx.doi.org/10.3390/s22062112] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35336282][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8953567]

43. Ali, A; Al-Rimy, BAS; Alsubaei, FS; Almazroi, AA; Almazroi, AA. Healthlock: blockchain-based privacy preservation using homomorphic encryption in internet of things healthcare applications. Sensors; 2023; 23, 6762. [DOI: https://dx.doi.org/10.3390/s23156762] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37571545][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10422473]

44. Picek, S; Batina, L; Jakobović, D; Ege, B; Golub, Ma. Naccache, D; Sauveron, D. S-box, SET, match: a toolbox for S-box analysis. Information security theory and practice. securing the internet of things; 2014; Berlin, Heidelberg, Springer Berlin Heidelberg: pp. 140-149. [DOI: https://dx.doi.org/10.1007/978-3-662-43826-8_10]

45. Panchami, V; Mathews, MM. A substitution box for lightweight ciphers to secure internet of things. J King Saud Univ-Comput Inf Sci; 2023; 35, pp. 75-89. [DOI: https://dx.doi.org/10.1016/j.jksuci.2023.03.004]