Blockchain-Based Mobile IoT System with Configurable Sensor Modules

Abstract

In this study, a Multi-Sensor IoT Device (MSID) is developed that is designed to collect various environmental data and interconnect with the cloud and blockchain to ensure reliable data management. The MSID is designed with a flexible, modular structure that supports a variety of sensor configurations and is easily expandable with 3D-printed components. The system performance was monitored in real-time, with a high cloud upload success rate of 98.35% and an average transmission delay of only 0.64 s, confirming stable data collection every minute. Blockchain-based sensor data storage ensured data integrity and tamper-proofness, with all transactions successfully recorded and verified via smart contract. The proposed Blockchain-based Mobile IoT System (BMIS) has shown strong potential for use in environmental monitoring, industrial asset management, and other areas that require reliable data collection and long-term preservation.

1. Introduction

The Internet of Things (IoT) encompasses a comprehensive network of sensing and actuating devices that collect, analyze, and exchange data across various platforms and applications. As one of the most significant technological advancements in recent years, the IoT has catalyzed an unprecedented information revolution [1]. The IoT extends beyond mere devices to include the supporting architecture, and as device connectivity expands, an increasing number of application reference architecture services will be deployed on these platforms [2]. Although the IoT fundamentally represents the concept of machines and objects connected to the internet, it is susceptible to certain limitations, such as applications that fail to perform as intended, necessitating careful optimization of all connected devices [3]. According to market research, the global IoT market was valued at USD 1.99 billion in 2018 and is projected to reach USD 11.03 billion by 2026 [4]. Zikria et al. (2021) [5] forecasted continuous growth in internet-connected devices, with predictions that 90% of vehicles will incorporate IoT connectivity by 2030, and individuals will possess an average of 15 connected devices per person. Following more than two decades of development, IoT has evolved into a sophisticated network of interconnected devices that facilitate automation across diverse sectors, including remote healthcare systems [6], precision agriculture [7], intelligent supply chain management [8], advanced transportation networks [9], and smart urban infrastructure [10]. Consequently, the IoT paradigm has gained widespread recognition and implementation across numerous domains [11,12,13].

The Wireless Sensor Network (WSN) constitutes a fundamental component of the IoT architecture and is rapidly expanding into numerous real-time applications. Both the IoT and WSN now encompass a diverse range of critical and non-critical applications that influence virtually every aspect of daily human activities [14]. WSN technology is extensively deployed for various purposes including tracking, video surveillance, remote monitoring, positioning, event reporting, industrial automation, disaster detection, and health monitoring systems. The proliferation of the IoT has stimulated research focused on energy efficiency to minimize memory requirements and protocol design complexity in WSN implementations, while simultaneously reducing operational costs and extending network lifespans [15,16]. Nevertheless, the paramount concern in WSN deployment for IoT applications relates to security and privacy vulnerabilities stemming from the heterogeneity of requirements and attributes; consequently, robust protection mechanisms against unauthorized access are essential to safeguard user data integrity and confidentiality [17]. To address these challenges, innovative solutions are being explored through the integration of advanced digital technologies. For instance, supply chain tracking systems that leverage blockchain technology in conjunction with IoT capabilities can significantly enhance transparency and operational efficiency [18].

The blockchain, pioneered by Satoshi Nakamoto in 2009, represents a system of sequential cryptographically secured blocks designed to prevent data manipulation [19]. This technology functions as a digital, distributed ledger shared across a network, offering enhanced security as records, once added, cannot be altered without consensus among participating entities. The applications of the blockchain extend to diverse domains, including smart contract implementation for financial fraud detection, secure medical record exchange, and numerous other use cases [20]. The cornerstone of blockchain’s decentralized methodology lies in extensive client participation, which renders data manipulation virtually impossible [21]. Within a blockchain-based decentralized Peer-to-Peer (P2P) trading network, recipient nodes validate incoming messages and, upon confirmation of authenticity, incorporate them into blocks. Subsequently, a consensus algorithm—known as “Proof of Work”—verifies the integrity of data within each block. Following successful execution of the consensus algorithm, the block is appended to the chain and universally acknowledged by all network nodes, facilitating continuous chain propagation [22]. Blockchains are classified into three primary categories—public, private, and consortium—based on their utilization patterns and inherent characteristics. Public blockchains maintain complete decentralization and unrestricted access, whereas private blockchains limit management capabilities to users specifically invited by a single organization. Consortium blockchains, alternatively termed “federated blockchains”, represent an intermediate framework between public and private models, with management responsibilities distributed among multiple organizations [23]. The structural attributes of blockchain technology have generated significant interest regarding its diverse application potential and operational methodology innovations. The integration of the blockchain with the IoT represents one of the most promising applications of this technology [24].

2. Related Works

WSN-IoT and blockchain technologies have been extensively investigated across diverse sectors, including content delivery networks, smart grid systems, and data management infrastructures, owing to their distinctive characteristics of auditability, decentralization, persistence, and anonymity [25].

A WSN-IoT-based wellness monitoring system for implementation in smart homes and intelligent buildings was developed by Ghayvat et al. (2020) [26], with evaluations conducted across various residential environments. Solutions were also proposed to address wireless network interference and signal attenuation challenges, thereby enhancing system reliability and operational efficiency.

The design and implementation of an IoT-enabled smart healthcare system integrating multiple healthcare services (including business analytics, oncological care, emergency response, and operational services) was presented by Onasanya et al. (2019) [27]. The Saskatchewan Health Authority was used as a case study, with the system aimed at optimizing healthcare delivery efficiency and improving patient quality of life.

A comprehensive analysis of blockchain technology’s influence on the development of sustainable IoT ecosystems was conducted by Sharma et al. (2020) [28]. Critical factors were identified, and future research trajectories were delineated to address technical challenges and facilitate environmentally sustainable IoT implementations.

A modified blockchain framework specifically tailored for IoT environments to strengthen security and privacy in medical big data applications was introduced by Dwivedi et al. (2019) [29]. Advanced encryption methodologies were incorporated to ensure secure data management while maintaining anonymization protocols.

BCTLF, a hierarchical framework integrating IoT and blockchain technologies to enhance security in transportation and logistics sectors, was introduced by Humayun et al. (2020) [30]. Several key benefits were highlighted in the analysis, including smart payment systems, fault tolerance mechanisms, real-time information exchange capabilities, and significant improvements in privacy and security protocols.

Challenges within industrial processes were identified by Khan et al. (2022) [31], and secure, efficient implementation strategies utilizing IoT and blockchain technologies were proposed. A comprehensive review of the existing research encompassing IoT, Industrial Internet of Things (IIoT), blockchain, and hyperledger technologies was undertaken. A distributed consortium framework based on hyperledger cogs was designed to facilitate secure transaction processing through both on-chain and off-chain communication channels. Chain code and consensus policies were also developed to optimize resource efficiency while maintaining system reliability in blockchain and IIoT environments, resulting in a versatile solution applicable across industrial and manufacturing contexts.

The application of blockchain technology to enhance node authentication in WSN and strengthen overall security infrastructure was investigated by Verma et al. (2020) [32]. Limitations in existing methodologies were critically analyzed, and implications of blockchain integration for sensor network security were examined. A foundation for future developments in WSN network security enhancement was established through this investigation.

The implementation of IOTA blockchain technology to address privacy and security concerns in digital environmental monitoring systems was advocated by Gangwani et al. (2021) [33]. IOTA’s Directed Acyclic Graph (DAG) architecture and its Masked Authenticated Messaging (MAM) protocol were specifically focused on as mechanisms for the secure management of IoT-generated data.

As research integrating blockchain and IoT technologies proliferates across diverse domains including healthcare, logistics, and energy sectors, innovative frameworks and application scenarios continue to emerge for enhancing security protocols and optimizing operational efficiency. Despite the recent surge in studies leveraging blockchain and WSN-IoT technologies, there remains a significant research gap regarding the direct acquisition of IoT sensor data and its subsequent management through blockchain storage mechanisms. Furthermore, there is insufficient research addressing the assurance of data integrity and reliability within these systems.

The primary objective of this study is to develop an IoT data collection apparatus that securely acquires and manages sensor data within WSN-IoT environments. To achieve this goal, the research implements a microcontroller-based multi-sensor network architecture and designs a Multi-Sensor IoT Device (MSID) configured for data storage and analysis in integration with cloud computing platforms.

This investigation culminates in the development of a Blockchain-based Mobile IoT System (BMIS) that incorporates transaction-based data storage and management capabilities through blockchain technology implementation. The system addresses critical challenges in WSN and IoT applications by preventing sensor data tampering, ensuring data transparency, and enhancing network security protocols. This research aims to provide robust and efficient IoT data management technologies and organizational frameworks applicable across various industrial sectors.

3. Materials and Methods

3.1. System Architecture and Components

3.1.1. Overall System Design and Hardware Configuration of MSID and BMIS

In this study, to establish a WSN-IoT system for IoT data collection and blockchain storage, a MSID was designed with multiple sensors and modules integrated around a central MicroController Unit (MCU). The MSID was engineered based on the NodeMCU ESP8266 (SZH-EK051, Espressif Systems, Shanghai, China) MCU, which functions as the primary component responsible for raw data acquisition from each connected sensor.

Table 1. Specifications of the sensors and MCU configured in the MSID.

Sensor	Measurement		Sensitivity	Interface
DHT-22	Temperature/Humidity		±0.2 °C/±0.1% RH	Digital
DS18B20	Temperature		±0.5 °C	1-Wire
BH1750	Light intensity		±20% LUX	I2C
BMP180	Barometric pressure		±0.03 hPa (0.25 m)	I2C
SSD1306	Display		100 kHz	I2C
MicroController	CPU	Clock Speed	Flash Memory/SRAM	Wi-Fi Built-In	Output Voltage
NodeMCU ESP8266	Tensilica LX106 32-bit RISC	80–160 MHZ	4 MB/64 KB	802.11 b/g/n	2.7–3.6 V

presents the detailed specifications and performance metrics of the NodeMCU ESP8266 and each sensor integrated into the MSID architecture. The MSID comprises several sensors and modules, including the following: a DHT-22 (AM2302, AOSONG Electronics Co., Ltd., Guangzhou, China) temperature and humidity measurement sensor module connected via a digital interface; a waterproof DS18B20 (SEN050007, Maxim Integrated, San Jose, CA, USA) temperature sensor module utilizing a 1-wire interface protocol; a BH1750 (SEN0097, ROHM Semiconductor, Kyoto, Japan) light intensity measurement sensor module interfaced through an Inter-Integrated Circuit (I²C); a BMP180 (P0000JNL, Bosch Sensortec GmbH, Kusterdingen, Germany) barometric pressure measurement sensor module; and a 0.96-inch SSD1306 (HAM5629, Solomon Systech Limited, Hong Kong, China) OLED display module featuring 128 × 64 pixel resolution. The I²C interface employed in this system represents a synchronous serial communication protocol between master and slave devices, enabling connectivity of up to 127 slave devices via two transmission lines: SDA (data) and SCL (clock). In this configuration, the master device generates clock signals to which slave devices synchronize their data transmission and reception operations [34].

Figure 1 illustrates the architectural block diagram of the BMIS, which integrates the MSID with an IoT cloud platform and blockchain technology to ensure secure data processing and management. The system architecture is specifically designed to collect and store real-time data from both integrated and expandable sensors and modules connected to the NodeMCU ESP8266 MCU.

The system has been engineered to maintain continuous operation through dual power options: C-type USB serial communication or an 18,650 lithium-ion battery. Data collected via the Wi-Fi-enabled NodeMCU ESP8266 is transmitted to the IoT cloud platform for processing and storage. The stored data can be accessed and visualized through Comma-Separated Values (CSV) files and interactive dashboards, facilitating real-time monitoring and comprehensive data analysis. Additionally, the SSD1306 display module has been configured to provide immediate visualization of sensor data updates directly from the MSID hardware. To strengthen data integrity and enhance security protocols, all processed data are recorded on a public blockchain network. This integrated approach enables users to monitor and analyze data in real time through multiple interfaces: the Workstation, the IoT cloud platform, and blockchain transaction status verification systems.

This study implemented the ThingSpeak (MathWorks, Natick, MA, USA) software platform, which operates compatibly with the NodeMCU ESP8266 to facilitate real-time data transmission and storage for the MSID. The integration of IoT technology with ThingSpeak provides a unified platform that enhances visualization and management capabilities across multiple operational dimensions [35]. The system utilizes channel identification numbers and API key configurations supplied by ThingSpeak to transmit sensor data to designated channels. Communication with the ThingSpeak server is established through the NodeMCU ESP8266’s integrated Wi-Fi functionality, with automated reconnection protocols implemented to address potential connection failures.

The ThingSpeak platform has been configured with dedicated channels for the systematic storage and organization of sensor data, with fields structured according to the quantity of sensor data parameters. Users can access and analyze this data in real time through multiple interfaces, including web-based dashboards, mobile applications, and programmatic data retrieval via an Application Programming Interface (API). Furthermore, the platform incorporates CSV export functionality, enabling users to download time-zone-organized data files to local environments for offline analysis.

The structural design of the MSID was developed using Autodesk Fusion (version 16.12.0.2384, Autodesk, Inc., San Rafael, CA, USA), a specialized 3D CAD software platform for product design applications. The MSID’s structural components were precisely modeled according to physical measurements of the incorporated sensors and electronic components. Following the detailed design of each component, the resultant model was exported in Standard Tessellation Language (STL) format. The STL files were subsequently processed through Creality Print (version 5.1.6) slicer software to generate g-code files for fabrication. These converted g-code instructions facilitated the construction of each component using a Creality Ender-3 V3 Plus 3D printer (Shenzhen Creality 3D Technology Co., Ltd., Shenzhen, China) with Hyper PLA_1.75 filament (Shenzhen Creality 3D Technology Co., Ltd., Shenzhen, China).

Figure 2 presents both exploded and assembled views of the MSID design developed in this study. All structural components were fabricated using consistent printing parameters.

Table 2. 3D printing parameters and slicing data for MISD components.

Parameter		Setting
Resolution		0.20 mm
Infill density		15%
Sparse infill pattern		Rectilinear pattern
Internal solid infill pattern		Monotonic pattern
Nozzle temperature/Bed temperature		220 °C/65 °C
Part Name	Printing Time (H:M:S)	Material Weight (g)	Material Length (m)
Master_Case	02:05:31	87.29	29.27
Master_Top	01:11:43	31.86	10.68
Slave_Case_1	00:44:19	17.8	5.97
Slave_Top_1	00:18:13	6.05	2.03
Slave_Case_2	01:03:43	36.28	12.16
Slave_Top_2	00:36:25	13.11	4.39

documents the comprehensive 3D printing parameters and slicing specifications implemented for the MSID components. The fabrication process utilized a print resolution of 0.20 mm with a fill density of 15%. Rectilinear patterns were applied for the Sparse Infill Pattern, while Monotonic patterns were employed for the Internal Solid Infill Pattern. Throughout the printing process, the nozzle and bed temperatures were maintained at constant values of 220 °C and 65 °C, respectively, to ensure optimal material extrusion and adhesion characteristics.

3.1.2. Electronic Circuitry and Firmware Configuration of the MSID

Each sensor incorporated in this study transmits measured data to the NodeMCU ESP8266 through its respective interface communication protocol. Figure 3 illustrates the comprehensive circuit configuration of the MSID, depicting the connectivity of all sensors and modules within the system. The MSID’s power supply architecture is based on an 18,650 lithium-ion battery, with voltage stabilization achieved through a regulator (LM7085, Texas Instruments Inc., Dallas, TX, USA) that modulates and reduces the voltage to a consistent 5V level. This regulated voltage is directed to the VIN pin of the NodeMCU ESP8266, ensuring operational stability throughout the device while maintaining compatibility with 5V motion detection sensors. Additionally, the MSID incorporates a manual power switch mechanism, enabling users to control the circuit’s power state according to operational requirements.

The MSID configuration depicted in Figure 3 is engineered to acquire data through two distinct sensor frameworks: built-in sensor modules and expandable I²C sensor modules. The built-in sensor modules establish direct connections to the NodeMCU ESP8266 (functioning as the I²C Master) for real-time data acquisition, while expandable sensor modules (operating as I²C Slaves) utilizing I²C communication protocols are implemented to facilitate supplementary data collection capabilities and functional extensibility.

In this investigation, two distinct methodologies were implemented for the firmware configuration of the MSID. The primary approach involved updating the NodeMCU ESP8266 with Tasmota firmware (version 14.5.0), while the alternative method entailed direct firmware development and deployment using the Arduino Integrated Development Environment (IDE) (version 2.3.4). These methodological approaches were selectively applied according to specific research objectives, with their respective characteristics and implementations described in the following sections.

• Tasmota Firmware

Tasmota firmware was selected for implementation based on the network connectivity requirements and wireless update capabilities essential for the MSID. Tasmota represents an open-source firmware solution compatible with ESP8266 and ESP32-based MCUs, providing robust wireless network connectivity and comprehensive remote-control functionalities [36]. The implementation process involved flashing a precompiled binary file to the NodeMCU ESP8266, followed by configuration procedures executed through the web-based interface. This systematic approach facilitated the establishment of appropriate network environment parameters and module configurations necessary for bidirectional sensor data transmission and reception.

• Arduino IDE

For sensor data acquisition and network connectivity implementation, C++-based firmware was developed using the Arduino IDE and deployed directly to the NodeMCU ESP8266 via serial communication protocols. The firmware was programmed to execute real-time measurements of multiple environmental parameters, including temperature, humidity, light intensity, barometric pressure, and additional configured sensor values. Furthermore, the firmware incorporates diagnostic functionality to generate warning messages upon detection of anomalous sensor readings or when sensor data transmission failures occur.

3.1.3. Sensor Communication and Data Protocol

The engineered MSID implements an I²C-based master–slave architecture for data transmission, configured to receive sensor data through standardized 20-byte data packets. Figure 4 illustrates the comprehensive data packet structure and data protocol specifications for the expandable slave sensor module, specifically the CO₂ sensor RX-9M (EX-NN-20123VN5KA, EXSEN Inc., Daejeon-si, Republic of Korea). The data packet architecture comprises a total of 20 bytes and encapsulates multiple data elements: data quantity indicators, sensor measurement values, data unit specifications, sensor enumeration, checktime values for monitoring update cycles, and checksum verification. All data within the packet is transmitted and received in ASCII format, with the checktime parameter functioning as a mechanism to track the data update frequency. The checksum value, which verifies data packet integrity, is calculated according to the following methodology:

$$Checksum=\left(\sum _{i=1}^{n-1}{Byte}_i\right) \oplus 0x55$$

(1)

In Equation (1), $n$ represents the total length (number of bytes) of the packet, and ${Byte}_1$ through ${Byte}_{n-1}$ represent all valid data bytes within the data packet. The calculation implements error detection methodology by summing all data elements within the packet and subsequently performing an XOR (exclusive OR) operation with 0x55. If the calculated checksum value deviates from the expected value, this discrepancy indicates the presence of anomalies in the sensor readings. This verification mechanism serves as a reliable indicator for data transmission errors or potential measurement irregularities.

The automated sensor detection process on connected slave devices is executed through systematic address scanning performed by the master (MSID) device. In environments conforming to the I2C protocol, addressing is implemented using a 7-bit addressing scheme, establishing a valid address range from 0 to 127. Notable exceptions to this scanning range include the BH1750 light sensor (0x23), SSD1306 display module (0x3C), and BMP180 barometric pressure sensor (0x77), which are excluded from the scanning process due to their predefined address assignments. The formula governing the sensor address scanning procedure is defined as follows:

$$A=\left\{S_{a }\in \mathbb{Z}| 0 \le S_{a }<128, S_{a }\notin \left\{0x23, 0x3C, 0x77\right\}\right\}$$

(2)

In Equation (2), $A$ is the set of valid sensor addresses, and $S_{a }$ is the address of the individual sensor to be discovered. After the address scan is complete, the master requests data from a specific address and checks for a response. If the slave sensor responds normally, it stores the data in a buffer and validates the data with a calculated checksum. Once the data are validated, the process of isolating and analyzing individual sensor data is performed. An algorithm is applied to determine the location of the sensor data within the packet, and the master device reads a 10-byte string based on the start of the sensor data block. The unit $U_k$ of data of sensor $k$ defined as follows:

$$U_k=\left\{{Byte}_j| j=S_k+1, j<S_k+11 \right\}$$

(3)

In Equation (3), $S_k$ is defined as a byte containing the amount of data of sensor $k$, and $S_k+1$ means the starting position of the sensor data unit byte. The sensor data value $V_k$ is defined as follows:

$$V_k=\left\{{Byte}_j| j=1, j<S_k+1\right\}$$

(4)

In this process, a technique was implemented whereby the master device automatically distinguishes data from the slave sensors and processes them separately.

3.2. Blockchain System

Figure 5 illustrates the structure of the blockchain, highlighting the hash-based linkage between blocks. The blockchain system comprises a hierarchical architecture that encompasses data storage, network organization, consensus processes, smart contract execution, and application service delivery. The fundamental unit of a blockchain is a transaction, and each transaction is transformed into a fixed-length hash value through a hash function to ensure data integrity. Multiple transactions are aggregated into a Merkle tree structure to generate a Merkle root hash, which is stored in the block header and serves as the foundation for verifying the integrity of all transactions within the block. Blocks are sequentially linked by incorporating the hash value of the preceding block. The block header contains several essential elements: block version, previous block hash, Merkle root hash, timestamp, nonce, and target difficulty. The definitive block hash is generated based on these components [37,38].

The data layer of a blockchain system provides the fundamental data structures and security techniques, including data blocks, chain structure, timestamps, hash functions, Merkle trees, and asymmetric encryption. The network layer performs data transmission, block propagation, and verification functions between nodes based on a P2P network architecture. The consensus layer ensures block validation and network-wide state synchronization through various consensus algorithms, including Proof of Work (PoW), Proof of Stake (PoS), Delegated Proof of Stake (DPoS), and Byzantine Fault Tolerance (BFT). The extension layer provides an environment for executing smart contracts and custom algorithms, while the application layer enables the implementation of diverse services such as supply chain management, IoT data storage, security and privacy enhancements, and financial technology services [39].

3.2.1. Local Blockchain Simulation

To conduct simulations for data storage and management on a local blockchain environment, an Ethereum blockchain model was implemented using the Truffle framework (version 5.11.5) in conjunction with Ganache (version 2.7.1), an Ethereum Test Network. The Ethereum blockchain, which commenced operations on 30 July 2015, was initially secured by the PoW protocol for more than seven years. However, due to the energy-intensive nature of PoW, the network subsequently transitioned to the PoS consensus mechanism [40]. For the simulation environment, ten unique addresses were randomly generated, and a smart contract was developed using the Solidity programming language (version 0.8.21). To facilitate interaction with the deployed smart contract, an Instance was created utilizing the Web3 library.

3.2.2. Blockchain Network

In this study, a smart contract was deployed to establish a sensor data storage system based on the Ethereum blockchain. The smart contract was implemented using the Solidity programming language (version 0.8.0) and executed on the Ethereum Sepolia Testnet (EST) network.

The EST is a public test network that provides the same features and operations as the main network but runs using ETH for testing purposes [41]. Ethereum, including Sepolia Testnet, currently uses the PoS consensus mechanism, which offers higher energy efficiency and faster finality compared to the previous PoW algorithm. These characteristics make it more suitable for handling frequent and lightweight transactions generated by IoT devices in near real-time scenarios.

MetaMask was utilized for executing blockchain transactions. MetaMask functions as a self-hosted wallet that enables users to store, transmit, and receive ETH and ERC20 tokens while maintaining direct control over their digital assets [42]. For testing purposes, 0.05 ETH was acquired at no cost through the Ethereum Sepolia Faucet supported by Google Cloud, facilitating access to the network’s functionality without incurring actual financial expenditure.

3.2.3. Network Connection and Smart Contract

The Infura platform, an API infrastructure service, was utilized for seamless interaction with the EST. Infura provides developer-friendly API endpoints, comprehensive monitoring capabilities, and multi-network connectivity to facilitate the development of scalable and reliable blockchain systems [43]. For smart contract development and deployment, the Hardhat framework (version 2.22.18) was employed to compile and deploy the smart contract to the EST. During this process, connection to the EST was established via Infura’s Remote Procedure Call (RPC) Endpoint, and the smart contract deployment transaction was signed using the configured private key. A smart contract functions as executable code that operates on a blockchain to facilitate, execute, and enforce agreements between untrusted parties without requiring the intervention of a trusted third party [35].

3.2.4. Sensor Data Storage and Retrieval System

Upon successful completion of the smart contract deployment process, the contract address is registered on the Sepolia Etherscan platform, enabling implementation of functionality to search for the corresponding contract address for data storage and retrieval operations. The sensor data from the MSID is acquired in real time through Hyper Text Transfer Protocol (HTTP) requests, with data from both built-in sensor modules and expandable sensor modules being automatically detected, converted into JavaScript Object Notation (JSON) objects, and processed for blockchain storage. When a transaction is executed, the data are stored in the EST, and a transaction hash is returned to verify successful storage completion.

3.3. IoT-Based Remote Monitoring and Blockchain Data Integrity System

3.3.1. System Architecture for Dual-Path Data Transmission

This study proposes a dual-path data transmission system that simultaneously enables real-time monitoring and ensures tamper-proof storage of sensor measurements. In our approach, sensor data are captured directly from IoT devices (MSID) and are transmitted along two distinct channels:

• ThingSpeak platform: for real-time monitoring, the original sensor data are transmitted directly to the ThingSpeak platform via an HTTP POST method using its API. This supports immediate visualization and analysis on the cloud-based platform provided by MathWorks;
• Ethereum Sepolia Testnet: in parallel, the collected sensor data are prepared for secure storage on the blockchain. First, the sensor data are combined with their corresponding timestamp. Instead of using traditional AES, our implementation utilizes the following blockchain-standard procedures:

Figure 6 illustrates the overall system architecture, emphasizing the dual transmission paths for real-time visualization on ThingSpeak and secure, immutable storage on the EST.

3.3.2. Data Security and Integrity Assurance Mechanisms

To ensure the integrity and authenticity of the sensor data during transmission and storage, the system leverages the following mechanisms:

(1)

Keccak-256 Hash Function for Data Integrity Verification:

The system computes a Keccak-256 hash over the combination of sensor data and its associated timestamp. This hash acts as a unique digital fingerprint, providing evidence that the original data have remained unaltered once recorded on the blockchain. The Keccak function (also known as SHA-3) provides robust protection against collision attacks and cryptographic vulnerabilities, serving as a highly secure hashing standard. This widely implemented algorithm plays a critical role in ensuring data privacy and maintaining integrity across diverse computing systems and digital environments [44].

(2)

RLP Encoding for Transaction Formatting:

The sensor data (or their hashed representation) are encoded using Recursive Length Prefix (RLP) encoding. RLP encoding standardizes the transaction data format, ensuring compatibility with the Ethereum network and facilitating secure transaction validation.

(3)

HTTPS Protocol-Based Secure Transmission:

HTTPS (i.e., HTTP/1.1 over TLS) is used specifically for transmitting data from the Node.js server to the EST via the JSON-RPC endpoint. This ensures that blockchain transaction data are securely encrypted during transit over public networks. Note that for real-time visualization, sensor data are transmitted to the ThingSpeak platform using HTTP, as the primary focus is on immediate data display rather than on cryptographic communication security.

(4)

ECDSA Signature-Based Sender Authentication:

The prepared blockchain transaction is signed using the Elliptic Curve Digital Signature Algorithm (ECDSA) with the IoT device’s private key. This signature mechanism not only authenticates the sender but also provides non-repudiation, ensuring future traceability of the transaction. ECDSA is a standardized derivative of the Digital Signature Algorithm that employs elliptic curve operations over finite fields, offering greater efficiency with significantly shorter key lengths for equivalent security levels compared to its predecessors [45].

The security architecture is summarized in

Table 3. Security architecture implementation.

Category	Applied Technology	Purpose
Data hash generation	Keccak-256	Data integrity verification
Data encoding	RLP	Standardized formatting of transaction data
Transmission protocol	HTTPS (TLS)	Network segment encryption
Sender authentication	ECDSA	Prevention of forgery and ensuring traceability

below:

3.3.3. Implementation and Integration with BMIS

The proposed security architecture was integrated with the BMIS described in Section 3.1. The NodeMCU ESP8266 was programmed to implement the Keccak-256 hashing function, combined with a timestamp, and to format the sensor data for blockchain transactions using RLP encoding. Additionally, each transaction was digitally signed using ECDSA with the device’s private key to ensure data authenticity and integrity. To optimize the transaction costs associated with blockchain storage, batch processing was implemented, wherein data points collected at one-minute intervals were aggregated and committed to the blockchain at ten-minute intervals.

3.4. Experimental Setup

In this study, an experiment was conducted to verify the process of collecting, storing, and retrieving environmental data on the blockchain using the developed MSID. The experiment was conducted in an environment where the latitude and longitude were specified in the ThingSpeak platform by setting the Cheonan, Chungcheongnam-do region as the operating location, and the actual data collection and blockchain storage process was performed in a room temperature environment in the laboratory. This experiment focused on evaluating the continuity and stability of data collection and blockchain storage.

The implementation and evaluation of the proposed method was carried out on a Lenovo IdeaPad Gaming 3 15IHU6 (Windows 11 22H2) with an Intel Core i5 3.11 GHz CPU and 32 GB RAM, and Node.js (version 18.20.6) was used for data processing and blockchain integration. The MSID collected temperature, humidity, light intensity, and barometric pressure data continuously for two hours at 1 min intervals using DHT-22, BH1750, BMP180, and DS18B20 sensors, and the collected data were uploaded to the ThingSpeak cloud platform and monitored in real time. The data collected were also processed in 10 min intervals and stored on the EST via a smart contract.

4. Results and Discussion

4.1. Manufactured MSID Master and Slave Modules with 3D Printing Process

The MSID designed and manufactured in this study consists of a master module and two types of slave modules. The printed and assembled MSID is shown in Figure 7. The MSID master is a module with a built-in main control unit and is designed with three pin header sockets on the top for slotting in Slave_1 modules. Sensor modules can be added and replaced as needed. In addition, an additional pin header socket can be configured on the side to slot into the Slave_2 module, and Slave_2 has an additional socket on the opposite side for expansion to accommodate additional slave modules. Each part was manufactured using 3D printing to ensure precise joining and stable fixation, and finally, the top and side slot structures enabled stable connection and assembly between the master and slave modules. Ultimately, a modular system was completed that can flexibly respond to various sensor configurations and data collection requirements.

4.2. Analyze System Performance and Assess Data Collection Reliability

The device continuously acquired data at 1 min intervals for two hours and uploaded them to the ThingSpeak cloud in real time. The summarized system performance of the MSID is shown in

Table 4. System performance evaluation results of MSID.

Category	Result	Note
Total data collected/Duration	121 points/120 min	100% data acquisition success
Data acquisition interval (Mean ± SD)	60.15 ± 0.82 s	Close to the 1 min setting
Total uploads/Success rate	119 uploads/98.35%	Minor data loss detected
Upload delay time (Mean ± SD)	0.64 ± 0.21 s	minimal delay in ThingSpeak upload
Wi-Fi connection time	approximately 10 s	initial connection only

. The data acquisition interval was 60.15 s, which was almost consistent with the set value of 1 min, and only a small error of ±0.82 s occurred, confirming the high accuracy of the data acquisition timing control. In addition, 119 out of 121 total data points were successfully saved to the ThingSpeak server, indicating an upload success rate of about 98.35%. Although some data were missing due to network environment and server response delays, overall stable data collection and upload performance was confirmed. As shown in Figure 8, the temporal variation of each sensor’s data could be visually monitored through the ThingSpeak dashboard. Once the data were acquired, the upload time was very short, averaging 0.64 s, with a standard deviation of ±0.21 s, indicating stable upload performance. The Wi-Fi connection time was about 10 s for the first connection, and the performance of maintaining the connection afterward was stable, so it is expected to operate reliably in a data collection environment for a long time.

4.3. Analyze Sensor Data Storage Interactions on a Local Blockchain

The following process was performed to successfully deploy the smart contract. First, the contract for migration was deployed in block number 1, and the transaction hash was confirmed as 0x0782a383ad0d782af683461e9db0d99ee37b560a5902a90e01531bb8c02da711. A custom smart contract was then deployed for the purpose of this study. The contract was recorded at block number 3, and the transaction hash was verified as 0x231741b2c4ce683bff87e60669b9d155ad1d8d1516a2c79628ddedad1557c3da. A total of 366,419 gas was consumed during the deployment, with the gas price set at 3.178 Gwei. The total cost of deploying this contract was calculated to be approximately 0.0011646 ETH. This contract is designed as a smart contract that provides basic data management functionality to store string data and retrieve stored data.

A total of 13 transactions were made to store sensor data in the local blockchain environment. Each transaction was performed through the same sender address (0xd38d2909729e8CDe2d137F1034B20a47BD0EaDdA), and the receiver address of all transactions was recorded as the deployed smart contract address (0x6a7cfF4389D1Df0162788e10A39e9e8367c4A643). Gas usage per transaction ranged from a minimum of 59,033 gas to a maximum of 78,657 gas, with an average gas usage calculated to be approximately 72,308 gas. The gas price fluctuated slightly from transaction to transaction but remained around 2.68 Gwei on average. The block hash and transaction hash of each transaction confirmed that all data storage was correctly recorded on the blockchain.

4.4. Smart Contract Deployment and Functionality

The smart contract is designed to store time-series sensor data, where each entry consists of a timestamp and a string of sensor data. The code of the smart contract deployed on the EST is shown in Figure 9. The SensorData contract defines a structure (SensorEntry) for storing sensor data, which contains the timestamp of when the data were collected and the sensor data value. Sensor data are stored via the storeData function, which automatically records the block timestamp of the blockchain network to ensure data timeliness. Additionally, when storing data, a DataStored event is generated so that an external monitoring system can check in real time whether data have been stored. The saved data are stored sequentially in the contract’s entries array, and the getLatestData function is provided to retrieve data. This function returns the most recent data and timestamp of the stored information and includes exception handling to prevent lookups when no data are stored.

4.5. Analysis of Blockchain-Based Sensor Data Transactions

The experimental result summarizing the sensor data stored on the blockchain at 10 min intervals over a 2 h period in the EST environment is shown in

Table 5. Experimental results of blockchain-based sensor data storage.

Category	Result
Total number of transactions/Transactions success rate	13 transactions/100%
Processing Time (Mean ± SD)	16.38 ± 11.68 s
Transaction fee (ETH, Mean ± SD)	8.61 × 10−5 ± 5.1 × 10−6 ETH
Gas price (Gwei 1, Mean ± SD)	0.309 ± 0.030 Gwei

¹ Gwei = 10⁻⁹ ETH (Ethereum gas price unit)

. A total of 13 transactions occurred, and all transactions were successfully recorded in the block, resulting in a 100% success rate. The detailed transaction history recorded during the period is shown in Figure 10. Transaction processing time averaged 16.38 s, with a standard deviation of 11.68 s. This variability can be attributed to block generation intervals and network congestion. The average fee per transaction was 8.61 × 10⁻⁵ ETH, with a standard deviation of 5.1 × 10⁻⁶ ETH. The average gas price per transaction was 0.309 ± 0.030 Gwei and remained relatively stable throughout the data storage process. This stability is likely due to the small size of the hash-based data saved in this study. The results confirmed that a blockchain-based sensor data storage system implemented in an EST environment can operate with a high success rate, stable gas price, and low fee level.

Figure 9. Solidity code for sensor data management.

Figure 9. Solidity code for sensor data management.

The event logs and data recorded on the blockchain when storing sensor data are shown in Figure 11. Each transaction was stored as a timestamp, a JSON string of sensor data, and the original binary form (HEX) of that data. This storage structure enables verification of data originality and integrity, allowing for the development of a reliable data management system based on blockchain-stored data.

5. Conclusions

In this study, various environmental data are collected, and an MSID is developed to reliably manage the collected data. BMIS, a cloud and blockchain-based data storage system, is proposed to monitor the data in real time and ensure its integrity. The modular design of the master–slave structure enables flexible responses to various sensor configurations, while the precise part manufacturing based on 3D printing and stable bonding structure allows the device to replace and expand sensor modules.

As a result of the system performance evaluation, the MSID stably collected data at the set interval of one minute and recorded a high success rate of 98.35% for uploading to the cloud, confirming its reliability in long-term continuous data collection and transmission. Data upload latency was also very low, averaging 0.64 s, which is ideal for real-time monitoring and data flow management. In the blockchain-based data storage experiment, all 13 transactions were successfully processed, showing a 100% storage success rate, and both average gas consumption and fees were maintained at low levels, proving this to be a suitable storage method for environmental data management requiring data integrity and originality. This study confirmed that managing stored data event logs through smart contracts effectively guarantees storage history and prevents data tampering on the blockchain.

The BMIS, the MSID-based cloud-blockchain-connected data management system proposed in this study, demonstrates high potential as an effective data management platform that simultaneously meets requirements for real-time collection and monitoring of sensor data, integrity assurance, and long-term storage. In the future, this system can be utilized in various fields where data reliability is crucial, such as environmental monitoring, industrial facility management, and smart farms. Further research, including the application of data encryption and privacy protection technologies, as well as the integration of AI-based data analysis and prediction functions, will enhance both the reliability and utility of the system.