A lightweight blockchain framework for secure and scalable IoT communication

1 Introduction

The world is rapidly moving toward an era where billions of interconnected devices communicate seamlessly to provide intelligent services. This paradigm, or the Internet of Things (IoT), has transformed such fields as healthcare, smart cities, manufacturing, and transportation (Jain et al., 2025). Objects that are inanimate are becoming living, perceiving, processing and communicating information. However, this ubiquitous connectivity begs the important question of how to provide secure, efficient and reliable communication between resource-limited devices (Mohamed, 2023)?

The traditional IoT networks are very much based on the centralised model where data is collected, processed and stored in a central server. The design is not as hard to operate, but it gives rise to serious security threats that include single points of failure, manipulation of data and privacy. As the size of the IoT networks increases, these shortcomings become even more noticeable, and the traditional security protection mechanisms are no longer efficient in larger IoT networks (Lone et al., 2023).

One potential solution to these issues has been proposed as blockchain, which possesses the qualities of decentralization, immutability and transparency (Avasthi et al., 2024). Blockchain eliminates the necessity of a centralized system and offers impartiality of information. An additional capability that smart contracts offer is the ability to communicate securely and autonomously between devices (Kaushik and Jain, 2025). These features render blockchain one of the most appropriate options to secure IoT. However, blockchain cannot be comfortably implemented with the IoT. The existing blockchain technologies are typically resource-consuming and inefficient. Some consensus algorithms such as Proof of Work and complex smart contract executions cause substantial latency and energy use and are not practical in resource constrained IoT devices (Villegas-Ch et al., 2025). In turn, the simple addition of traditional blockchain solutions may negatively impact the performance.

Recent research seeks to address this issue through combining blockchain with edge computing, machine learning and distributed storage (Jain and Kumar, 2025). Despite the fact that these technologies increase security and intelligence, they tend to introduce additional complexity and overhead. Consequently, there has been an escalating need to develop strategies that focus on security, scalability and efficiency without the need to change the devices used in IoT (Cheikh et al., 2026; Agarwal et al., 2020). We propose a lightweight blockchain solution to avoid these problems and provide a secure communication in an IoT. The central idea of the proposed solution is to simplify the operations of blockchains and transfer computational processes of IoT devices to edge nodes. The proposed framework can accomplish this goal by exploiting hash-based data encoding, lightweight data validation and edge computing to process the data, which leads to a lower cost of computation, storage overhead, and communications delay. It also guarantees the vital security provisions of blockchain, such as data integrity and immutability.

The key contributions of this paper are three-fold:

• First, we develop a light-weight architecture that effectively combines IoT devices, edge computing and blockchain networks.
• Second, a novel lightweight data processing and verification technique is proposed to reduce energy.
• Third, the performance is thoroughly evaluated, revealing substantial gains in computational efficiency, latency, storage and energy under current solutions.

The remainder of this paper is organized as follows. Section 2 reviews related work. Section 3 describes our framework. Section 4 presents implementation and evaluation criteria. Section 5 provides results and comparisons. Finally, Section 6 concludes the paper.

2 Literature review

The IoT has rapidly expanded and turned the world into a smart, interconnected setting. In healthcare and industrial automation, IoT devices continuously create and share big amounts of data. This connectivity, however, has an enormous challenge, which is security and trust. The classical centralized systems have several problems that include tampering of data, single points of failure and low scalability making them inappropriate in the current IoT setting.

To overcome these problems, scholars have turned to the use of blockchain technology as a decentralized and secure solution. Among the first directions was the secure data storage. As an illustration, a blockchain-based IoT data storage system was suggested by de Brito Goncalves et al. to store IoT-related data with the help of smart contracts and IPFS, which allows a distributed storage with no possibility of tampering and low reliance on centralized servers (de Brito Gonçalves et al., 2022). Also, Hasan et al. proposed a framework of reliable IoT data streaming with blockchain and IPFS, which guarantees the integrity and reliability of data in real-time communication (Hasan et al., 2022). These articles shed light on the possibility of improving trust in IoT systems through the decentralization of data management with the help of blockchain.

Recent studies have also explored optimized blockchain architectures for efficient IoT data management. Al-madni et al. proposed a blockchain framework that improves secure IoT communication while reducing system overhead and improving scalability (Al-madni et al., 2024). As the IoT systems were further developed, scholars started to incorporate smart mechanisms in the blockchain systems. Cai et al. suggested a secure IoT blockchain architecture, GTxChain, with the integration of graph neural networks to enhance security analysis and improve decision-making (Cai et al., 2023). This article demonstrates how the combination of blockchain and new learning techniques can further improve IoT systems. However, the increased use of smart contracts in these systems also introduced vulnerabilities. He et al. studied the detection of vulnerabilities in blockchain smart contracts and they stressed the need for successful security verification (He et al., 2023).

To improve access control and privacy, several research has been dedicated to smart contract-based security. Zhonghua et al. developed a scheme of blockchain and edge computing-based attribute-based access control for IoT to enable fine-grained and secure access to IoT data (Zhonghua et al., 2024). Similarly, Guo proposed smart contracts based on blockchain to formally verify access control in IoT systems ensuring the correct and safe functioning of the system (Guo, 2025). These approaches highlight the need for secure and automated access control solutions in decentralized IoT.

Likewise, blockchain has been used in specific applications of IoT. For instance, Mofatteh et al. suggested a smart energy system framework, EnerChain, that creates blockchain and smart agents to share knowledge (Mofatteh et al., 2025). Meanwhile, Anand et al. combined blockchain and transfer learning algorithms to securely share data on the IoT, which has improved its adaptability and fault tolerance (Anand et al., 2025). These works demonstrate how blockchain can be applied to support different IoT applications.

While they have made these enhancements, there are still security issues, such as blockchain network and IoT attacks. Erfan et al. spoke about the problem of Sybil attacks in blockchain-based industrial IoT systems using decentralized federated learning and made it more resilient to malicious nodes (Erfan et al., 2026). Similarly, Qaisar et al. proposed a blockchain based trust management scheme of vehicle networks to ensure secure communication in V2X (Qaisar et al., 2026). The study identifies the need for trust and security in decentralised IoT systems.

The integration of new technologies in blockchain is also being investigated. For example, Crisostomo et al. reviewed the different machine learning based techniques to detect vulnerabilities in smart contracts, highlighting the role of AI in blockchain security (Crisostomo et al., 2025). In addition, Zeng et al. investigated integrating big language models with blockchain to enhance the smart contract intelligence and capabilities (Zeng et al., 2026). Blockchain has additionally been investigated for securing distributed sensing and object detection data in intelligent environments, further highlighting its applicability in real-time IoT systems (Shareef et al., 2021). These insights suggest the evolution towards intelligent, adaptable and self-driven blockchain-based applications.

Table 1 shows that current approaches are mainly focused on security and advanced features but do not have a lightweight design and scalability. This suggests a lightweight and efficient blockchain platform is required to provide secure communication with reduced computational and energy costs.

TABLE 1. Summary of related work.

Research Gap and Motivation: Despite large advances, the majority of solutions available are computationally intensive, complicated, and cannot be used with resource-constrained IoT devices. A variety of frameworks concentrate on security or intelligence and do not consider lightweight design, scaling, and real-time communication efficiency. This leaves a loophole that can be used to create a lightweight, simple, and scalable blockchain-based framework that can provide a secure communication channel of the IoT without exerting high computational load.

3 Proposed lightweight blockchain framework for secure IoT communication

3.1 System architecture

The proposed framework is driven by a basic fact that although blockchain provides powerful security assurances, its direct adoption with IoT is frequently not viable because of computational and storage demands. Thus, the architecture is highly engineered to push the complexity out of the devices of the IoT and maintain the security and trust (Liu et al., 2022). The system is designed into three collaborative layers as shown in Figure 1. The IoT device layer is comprised of heterogeneous sensors and smart devices, which constantly sense and produce data streams. Such devices are generally resource-limited in nature and due to this limit, they do little more than sense and transmit data. The generated data D i is forwarded to the edge layer, which acts as a local processing unit.

The edge layer is very important in ensuring that the system is lightweight. It carries out preprocessing activities such as filtering of redundant values, aggregation of data and data format validation. This is to minimize unwanted data transmission and enhance efficiency. On validation, the edge node derives a safe hash representation of the data and builds a transaction. Lastly, the blockchain layer is trusted and irreversible. Raw IoT data are not permanently stored on the blockchain. Instead, only compact cryptographic hash values are recorded on chain, while the original IoT data are maintained at trusted edge storage nodes or lightweight distributed storage repositories. This hybrid storage approach significantly minimizes blockchain storage overhead while still preserving data integrity, traceability, and auditability through hash verification. This layered structure enhances scalability as well as facilitating real-time communication by achieving lower latency and reducing computation load at the device level.

3.2 Lightweight data representation and security model

A key contribution of the proposed framework lies in its lightweight data handling mechanism, which ensures security without excessive resource consumption. Let D i denote the data generated by the i t h IoT device. A cryptographic hash function is applied (Chenchev and Aleksieva-Petrova, 2018) to produce a fixed-length digest as H i = H D i

This hash value uniquely represents the data while hiding its actual content. Instead of storing D i, only H i is transmitted to the blockchain. This design achieves two important goals: data integrity and storage efficiency. Any modification in D i will result in a different hash, making tampering easily detectable. Each block in the blockchain is defined as B j = H i , T j , P r e v H a s h j , M R j where T j represents the timestamp, P r e v H a s h j ensures linkage with the previous block, and M R j denotes the Merkle root for transaction grouping. The fact that Merkle root is included further improves the efficiency of verification since more than one transaction can be validated using a single hash structure (Nakamoto, 2008). This compact representation can be observed in the block structure depicted in Figure 1, in which only necessary metadata is stored. This methodology will greatly decrease the memory usage and make the system applicable in the large-scale IoT scenario with constant data streams.

3.3 Secure communication and transaction formation

The communication mechanism in the proposed system follows a structured yet efficient workflow. When an IoT device generates data D i, it is transmitted to the edge node, which verifies its authenticity and removes inconsistencies. After validation, the edge node computes the hash H i and constructs a transaction defined as follow in Equation 1 where I D i represents the unique identifier of the IoT device. This transaction encapsulates the essential information required for secure verification while avoiding redundant data transmission. The overall process can be summarized algorithmically as follows:

T x i = H i , T j , I D i (1)

• Step 1: IoT device senses and generates data D i
• Step 2: Data is transmitted to the edge node
• Step 3: Edge node performs validation and preprocessing
• Step 4: Compute hash H i = H D i
• Step 5: Form transaction T x i
• Step 6: Broadcast transaction to blockchain validators
• Step 7: Execute consensus mechanism
• Step 8: Append validated transaction to block B j

This algorithm guarantees that the computationally intensive tasks are processed by the edge nodes and blockchain validators, whereas the IoT devices can be kept lightweight. All the communication process, including data creation and block generation, is illustrated in Figure 1., and the smooth operation of all the elements is emphasized.

3.4 Lightweight validation consensus mechanism

Conventional consensus algorithms like Proof of Work cannot be used in the IoT because of their energy and computing-intensive nature (Dorri et al., 2017). Figure 2 illustrates the proposed, Lightweight Validation Consensus (LVC) mechanism by overcoming these limitation.

A group of authorized edge validators are involved in transaction verification in this mechanism. The validators assess the received transaction and provide a binary result as follow on Equation 2

V i = 1 , if transaction is valid 0 , otherwise (2)

Let N denote the total number of validator nodes. The cumulative validation score is computed by the following Equation 3

S = ∑ i = 1 N V i (3)

The transaction is accepted if: S ≥ N 2 . Otherwise, it is rejected. This majority-based validation ensures reliability while avoiding the complexity of mining. Once consensus is achieved, the transaction is added to a new block, and the updated block is broadcast across the network. The process illustrated in Figure 2 demonstrates how transactions are broadcast, validated by multiple edge nodes, and finalized through a simple yet effective decision rule. This mechanism significantly reduces energy consumption, improves validation speed, and enhances scalability. The security robustness of the proposed LVC mechanism under adversarial conditions is further discussed in Section 3.6.

3.5 Security analysis

The proposed framework ensures secure IoT communication through a combination of hashing, blockchain immutability, and lightweight consensus. Data integrity is maintained using cryptographic hash functions H D i, where any modification in data results in a different hash value, making tampering easily detectable. 1mmutability is achieved through block chaining using P r e v H a s h j, ensuring that recorded data cannot be altered.

To prevent replay attacks, each transaction includes a timestamp T j, allowing validators to reject duplicate or delayed messages. The LVC mechanism enhances resistance to Sybil attacks by allowing only trusted edge validators to participate, where majority voting ensures reliability. Authentication and access control are enforced at the edge layer, preventing unauthorized devices from injecting malicious data.

Confidentiality is preserved by storing only hashed data on the blockchain instead of raw IoT data, reducing the risk of data leakage. Additionally, the decentralized architecture ensures availability and fault tolerance even in case of node failures. The overall security features of the proposed framework and their corresponding attack resistance are summarized in Table 2.

TABLE 2. Security features and attack resistance.

3.6 Threat model and robustness analysis

The proposed Light-weight and Cost-efficient resource-constrained communication (LVC) mechanism is suitable for resource-constrained IoT environments, which require light operation and low communication overhead (Mohamed, 2026). The proposed threat model involves internal and external attacks on the IoT devices and edge validators. A malicious node can try to add erroneous transactions or replay validated transactions or can collaborate or use a Sybil attack to manipulate the consensus process. The proposed structure assumes that edge validators are semi-trusted entities and that more than 50% of the participating validators participate in the consensus process honestly. Let N denote the total number of validator nodes and V h o n e s t represent the number of honest validators. Consensus correctness is maintained when V h o n e s t > N 2.

In this condition, it is possible for malicious validators not to control the majority voting process in the LVC mechanism. A set of edge validators validate each transaction independently, and the validation is accepted if the majority agreement condition is met. This will help to minimize the chance of accepting counterfeit or altered transactions. Only registered edge validators are eligible to take part in the consensus process, as a measure against Sybil attacks. Identities of validators are registered when the network starts, so that any malicious node can’t join the validation layer. Moreover, timestamp verification of transactions by dropping duplicate (or delayed) packets will ensure that replay attacks are not possible.

The proposed framework does not involve multi-round communication or computationally intensive agreement process as in traditional Byzantine Fault Tolerant (BFT) protocols. This is intentional, since the main goal of the proposed work is to offer lightweight and low-latency consensus to be used with IoT systems. While the proposed mechanism may not be suited to large-scale validator compromises in highly adversarial settings, it offers a good trade-off between security, scalability and computational efficiency for practical IoT-edge deployments. The proposed LVC mechanism is more robust to common IoT attacks than conventional blockchain consensus mechanisms, while still having low computational and communication costs.

4 Implementation, performance metrics, and comparison

4.1 Implementation setup

We simulate the lightweight blockchain system in an IoT network of sensor nodes, edge nodes, and blockchain nodes. The design adopts a three-tiered architecture with IoT nodes generating data, edge nodes pre-processing and hashing data, and blockchain nodes performing secure storage and validation. In contrast to existing implementations, IoT devices do not perform complex cryptographic tasks, thus offloading their workload. A simulation with network sizes of 50, 100, 200 and 500 IoT nodes is used to study scalability. Intermediate validators (edge nodes) perform lightweight consensus instead of mining. Raw IoT data are maintained in lightweight off-chain edge storage repositories, whereas only cryptographic hash representations are stored on the blockchain.

The proposed work primarily focuses on lightweight blockchain validation and communication efficiency. Therefore, the latency and storage overhead associated with external distributed storage systems are not explicitly modeled in the current implementation and are considered part of future work.

4.2 Performance metrics

We assess the system based on five parameters: computational overhead, latency, storage overhead, energy consumption, and throughput. Computational overhead indicates the time required to process and validate transactions. Latency is the cumulative time of data generation to the final block confirmation. Storage overhead is a measure of the amount of memory needed to store transaction data in the blockchain. The energy consumption is a measure of the overall efficiency of the system especially on resource-constrained IoT devices. Throughput measures the scalability in terms of number of transactions that were completed within a unit time. The overall performance of the system can be mathematically expressed as follows in Equations 4–8 where C e d g e represents computation at edge nodes, C v a l i d a t i o n denotes blockchain validation cost, T p r o c e s s i n g, T c o n s e n s u s, and T b l o c k represent processing, consensus, and block confirmation delays, respectively. S denotes storage per transaction, E represents total energy consumption, and T indicates throughput.

C t o t a l = C e d g e + C v a l i d a t i o n (4)

L = T p r o c e s s i n g + T c o n s e n s u s + T b l o c k (5)

S = ∣ H D i ∣ (6)

E = E c o m p u t a t i o n + E c o m m u n i c a t i o n (7)

T = N t r a n s a c t i o n s t i m e (8)

4.3 Comparison framework

The performance is compared with three recent methods. (Hasan et al., 2022) uses blockchain with IPFS for secure storage, (Cai et al., 2023) proposes a GNN-based smart blockchain, and (Zhonghua et al., 2024) uses smart contracts for access control. This work is secure but has higher computation, delay and storage costs.

4.4 Complexity analysis

The computational efficiency of the proposed framework can be analysed in terms of time complexity. Let n denote the number of IoT devices in the network. In the proposed system, most computational tasks are offloaded to edge nodes, and each transaction undergoes lightweight validation. Therefore, the overall computational complexity can be approximated as: C p r o p o s e d = O n since each node contributes one transaction and validation is performed using a simple majority-based decision mechanism.

In contrast, existing methods introduce additional computational overhead. (Cai et al., 2023), which integrates graph neural networks, involves pairwise node interactions and feature propagation, leading to a complexity of approximately: C C a i = O n 2 . Similarly, (Zhonghua et al., 2024) employs attribute-based access control using smart contracts, resulting in: C Z h o n g h u a = O n ⁡ log ⁡ n due to multiple validation and policy evaluation steps. (Hasan et al., 2022) also incurs higher overhead due to blockchain and IPFS interactions. The linear complexity of the proposed framework ensures that the system scales efficiently with increasing number of IoT devices. This theoretical analysis supports the experimental results presented in Section 5, where the proposed method demonstrates lower computational overhead, latency, and energy consumption.

5 Results and discussions

5.1 Computational overhead and latency analysis

The computational efficiency of our proposed approach is experimented on varying densities of IoT nodes and the results are presented in Figure 3. The computational time curve as proposed has a linear relationship between the number of nodes and the computational time with around 5 ms at 50 nodes up to 20 ms at 500 nodes. In contrast, (Hasan et al., 2022) shows an increase from 8 ms to 35 ms, (Zhonghua et al., 2024) from 10 ms to 45 ms, and (Cai et al., 2023) from 15 ms to 70 ms over the same range. This shows that the proposed model is able to reduce the computational complexity by about 42% compared to (Hasan et al., 2022), 55% compared to (Zhonghua et al., 2024), and 71% compared to (Cai et al., 2023) at higher node density. The simplicity of the proposed model makes it achieve this improvement. Graph neural networks, which consist of various layers (layers of feature extraction and propagation) that are increasingly more complex, are used by (Cai et al., 2023). The attribute-based access control to smart contracts is implemented in the article by (Zhonghua et al., 2024) and involves multiple verifications and evaluation of policies. The work by (Hasan et al., 2022) is simpler and yet has to interact with IPFS, which would lead to more communication and processing cost.

To further analyse system stability, the variation in computational overhead across different node densities is observed to be gradual and consistent. The proposed method shows a near-linear increase with an average increment of approximately 3–5 ms per scaling step, indicating stable performance. In contrast, existing methods exhibit larger fluctuations, particularly (Cai et al., 2023), where the increase exceeds 15 ms at higher node densities. This confirms that the proposed framework maintains better computational stability. On the other hand, the proposed approach involves only simple hashing and edge-based verification, which can lower the processing overhead. This is further supported by the results in Figure 4.

Our approach has an end-to-end latency of around 30 ms, whereas (Hasan et al., 2022) has 55 ms, (Zhonghua et al., 2024) has 65 ms, and (Cai et al., 2023) has 90 ms. This equates to latency improvements of 45%, 53% and 66%, respectively. This performance gain is mainly attributed to removing the heavy consensus mechanisms and the use of edge-assisted preprocessing to reduce both processing time for transactions verification and network delay. Hence, the proposed framework is highly suitable for the real-time IoT applications involving low latency.

Figure 5 shows that latency increases with block size (1, 2, 4, 8 and 16 KB) for all methods. However, the proposed framework demonstrates slower growth due to lightweight validation and edge preprocessing. At higher block sizes (16 KB), the proposed method achieves ∼47% lower latency compared to (Hasan et al., 2022) and ∼70% lower than (Cai et al., 2023)

5.2 Storage and energy efficiency analysis

Figure 6 presents the evaluation of the storage efficiency of the proposed framework, from which it is clear that the storage size per transaction is much smaller than other approaches. Our method requires only 1 KB of storage per transaction, compared with 12 kB, 8 KB and 15 KB required in (Hasan et al., 2022), (Zhonghua et al., 2024) and (Cai et al., 2023), respectively. This translates into a 91% reduction in storage compared to (Hasan et al., 2022), 87% reduction compared to (Zhonghua et al., 2024), and 93% reduction compared to (Cai et al., 2023). This significant reduction is mainly due to the use of hash-based data representation, which only stores a lightweight hash of the data on the blockchain rather than the data itself or additional metadata. (Hasan et al., 2022) uses IPFS to provide distributed storage, but still requires storing data references and metadata, which contributes to storage costs. (Cai et al., 2023) adds extra structures for graph representation learning, leading to increased storage. (Zhonghua et al., 2024) requires storing access policies and smart contract states, which add to moderate storage costs.

The proposed method is much simpler and efficient as it stores only necessary cryptographic hashes. The energy consumption (Figure 7) exhibits a similar trend. At 500 nodes, the proposed method uses around 45 J compared to 80 J used by (Hasan et al., 2022), 95 J by (Zhonghua et al., 2024), and 140 J by (Cai et al., 2023). This represents energy savings of 44%, 53% and 68% respectively. The energy savings are due to the reduction in the number of computational operations and communication. The fact that IoT devices do not need to perform complicated operations and that the packet size is reduced by applying hashing techniques, reduces the total energy consumption. Hence, the proposed framework is ideal for IoT devices and wireless sensor networks (WSNs) with a battery power source.

The energy consumption trend also demonstrates consistent behavior across varying network sizes. The proposed method demonstrates a controlled growth in energy consumption, whereby, the average growth is about 10–15 J across the levels of scaling. Comparatively, current approaches have more abrupt increments, especially where energy consumption increases high because of the computational overhead (Cai et al., 2023). This implies that the proposed framework offers a more predictable and stable energy performance.

5.3 Scalability and throughput analysis with numerical comparison

The throughput analysis, as shown in Figure 8, is used to analyse the scalability of the proposed framework. The throughput of the proposed framework is around 105 transactions per second even with 500 nodes. By contrast, (Hasan et al., 2022) decreases to 70 transactions per second, (Zhonghua et al., 2024) decreases to 60 transactions per second and (Cai et al., 2023) further decreases to 40 transactions per second. This suggests that the proposed system exhibits a throughput that is 50% better than (Hasan et al., 2022), 75% better than (Zhonghua et al., 2024), and 160% better than (Cai et al., 2023). The drop in throughput of existing methods can be explained by the additional overheads in terms of computational load and network congestion. (Cai et al., 2023) is limited by processing delays due to GNN computations and (Zhonghua et al., 2024) by bottlenecks due to smart contract execution. (Hasan et al., 2022) is relatively good but suffers from IPFS communication and storage delays. On the other hand, the proposed network achieves load balancing at the edge nodes and has a simple consensus algorithm, which ensures consistent performance even with increased load.

Scalability wise, the proposed framework sustains a constant throughput with slight fluctuations in the throughput with increasing node densities. This much less than 20% reduction with higher loads even compared with existing methods which have shown reductions more than 40%–60%. This demonstrates the strength and stability of the suggested solution in large-scale settings. Figure 9 demonstrates how malicious nodes can affect the system throughput. The higher the percentage of malicious nodes, the lower the throughput of all the methods. Nevertheless, the suggested framework is more stable, because of majority-based LVC mechanism. The proposed method still maintains 85 tx/sec even at 40% malicious nodes, which is approximately 102% higher than (Hasan et al., 2022) and 325% higher than (Cai et al., 2023). Although throughput decreases with increasing malicious node density, the proposed LVC mechanism maintains stable transaction validation due to majority-based validator agreement. The percentage improvements are computed relative to the baseline throughput under identical malicious node density.

5.4 Ablation analysis

Figure 10 shows the ablation analysis of the proposed framework. As it can be seen, the processing of edges can be removed at the cost of 30 ms latency, which is reduced to 55 ms with the removal of edge processing. On the same note, the removal of hashing has the maximum latency (60 ms) because the removal of hashing leads to large data transmission, which wastes time. The removal of LVC is the worst in terms of latency (60 ms), since the large data transmission, caused by the removal of LVC, wastes time. This proves that every part has its own substantial contribution to the performance of the system.

The quantitative comparison of all performance metrics is summarized in Table 3, which highlights the superiority of the proposed framework over existing methods (Hasan et al., 2022; Zhonghua et al., 2024; Cai et al., 2023). It can be observed that the proposed approach consistently achieves lower computational overhead, reduced latency, minimal storage requirements, and improved energy efficiency while maintaining higher throughput. Specifically, the computational overhead of the proposed method is reduced to 20 ms, compared to 35 ms, 45 ms, and 70 ms for (Hasan et al., 2022; Zhonghua et al., 2024; Cai et al., 2023), respectively, resulting in improvements of 42%, 55%, and 71%. Similarly, the latency is reduced to 30 ms, achieving improvements of 45%, 53%, and 66% over the respective methods. In terms of storage efficiency, the proposed framework requires only 1 KB per transaction, which is significantly lower than 12 kB, 8 kB, and 15 KB required by (Hasan et al., 2022; Zhonghua et al., 2024; Cai et al., 2023), leading to a maximum reduction of 93%. The energy consumption is also reduced to 45 J, offering savings of 44%, 53%, and 68%, respectively.

TABLE 3. Quantitative performance comparison of proposed framework with existing methods.

Furthermore, the proposed system demonstrates strong scalability by maintaining a throughput of 105 transactions per second, which is 50% higher than (Hasan et al., 2022), 75% higher than (Zhonghua et al., 2024), and up to 160% higher than (Cai et al., 2023) under high node density. The proposed framework achieves a balanced improvement across all performance metrics, with particularly significant gains in storage efficiency and throughput, making it highly suitable for large-scale and real-time IoT applications. Figure 11 is a comparison of all the performance measures of the various methods. It is evident that the proposed framework outperforms existing approaches by a huge margin in terms of computation, latency, storage, and energy consumption and achieves much higher throughput.

6 Conclusion

This paper proposed a lightweight blockchain approach to solve the problem of efficient and secure communication in IoT systems. The proposed framework, in contrast to traditional blockchain approaches that impose high computational and storage costs, harnesses edge computing and a hash-based data model to reduce resource requirements without compromising security. Through offloading processing to edge nodes and simplifying consensus, the framework enables quick and efficient communication for real-time IoT applications. The results of the performance analysis indicate the superior performance of the proposed system compared to the state of the art. The proposed system achieves about 42%–71% reduction in the computation cost, hence reducing the processing time. It also delivers up to 66% lower latency, allowing for quicker data transmission and verification. The system is scalable for large-scale IoT due to the nearly 93% reduction in storage using hash-based data storage. In addition, energy efficiency is improved by up to 68%, essential for energy-constrained IoT devices. The proposed system also has a high degree of scalability with throughput of 105 transactions per second, up to 160% better than the existing methods under high network traffic conditions. The proposed approach shows that it is possible to enable secure and efficient IoT communications without employing advanced blockchain techniques. The findings demonstrate it is ready for large-scale, real-time IoT deployments.

Future work will investigate integration with distributed storage frameworks such as IPFS for fully decentralized IoT data persistence.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Nasurudeen Ahamed N, United Arab Emirates University, United Arab Emirates

Reviewed by: Tamara Saad Mohamed, University of Kut, Iraq

Zeyad Abdulhameed Taha Ahmed, INTI International University, Malaysia