Food quality traceability system based on multi-chain architecture blockchain

Abstract

This study proposes a secure food quality traceability system based on a multi-chain blockchain architecture to improve both data protection and the efficiency of food quality management. The system leverages blockchain technology to monitor and safeguard the entire supply chain—including production, processing, and transportation. To ensure confidentiality and integrity of data, the system integrates the Paillier encryption algorithm and secure hash algorithms for encrypted storage and secure sharing (the Paillier algorithm supports addition and multiplication operations, making it suitable for handling data aggregation and analysis tasks in the supply chain. It has high computational efficiency and can quickly complete data encryption and decryption). Hyperledger Fabric is employed to manage permissioned access, enforce data query control, and mitigate unauthorized operations. Security tests indicate that the system achieves high encryption efficiency—encrypting 2500 kb of data in 22 ms with a 32-bit key and achieving the fastest decryption at 443 ms with a 128-bit key. Compared to traditional systems, it significantly improves throughput (412 TPS vs 305 TPS) and risk interception (92 vs 52). These results demonstrate that the proposed system enhances cybersecurity in traceability applications by offering strong encryption, fine-grained access control, and high performance. It provides a robust and scalable solution for securing sensitive data in food supply chains, with broader implications for blockchain-based cybersecurity applications in industrial networks.

Keywords

traceability system food paillier secure hash algorithm hyperledger fabric encryption

Introduction

Food chain traceability technology is an innovative approach that utilizes blockchain technology for product tracking, ensuring supply chain (SC) transparency, and improving consumer trust.¹ The complexity of SC management has increased with the growth of globalization and information technology. Because of this, it is especially crucial to guarantee authenticity and transparency in all facets of food quality (FQ), from production to consumption.² The untamperability, decentralization and traceability features of blockchain technology make it an ideal technology for achieving effective traceability. It can provide more reliable information about food safety (FS) records for the SC as well as consumers.³ Dong L et al. studied the issue of food SC traceability and developed a SC model containing multiple tiers of suppliers to analyze the impact of traceability technology on the incentives and benefits of SC members. The results showed that while full traceability saves uncontaminated food and increases revenue, it may also lead to impaired member returns due to strategic pricing and even increase the risk of contamination.⁴ Gupta R et al. used a soft systems approach to develop a blockchain-based food traceability system (FTS) for food insecurity in the public distribution system in India. The system provided basic food items to beneficiaries by increasing transparency and helped policy makers to make informed decisions. However, the study did not fully consider the feasibility of the technology in large-scale implementation.⁵ Through service design workshops and trials, Hao F et al. investigated the use of blockchain technology in the food SC of the restaurant business and discovered that it greatly enhances traceability and trust, which in turn raises customer happiness. The study also noted that the effects varied by restaurant type and location. The study provided guidance for technology investments in the restaurant industry, but did not delve into the applicability of blockchain technology in restaurants of different sizes.⁶

A technique called blockchain cryptography makes use of cryptographic concepts to guarantee the security and integrity of data. By using cryptographic techniques like public key encryption and hash functions, it safeguards data and makes sure that only those with the right key can access and change it. Blockchain-based cryptography is extensively utilized in a number of fields, including medical records, finance, and SC management. Hashim A N et al. studied the security issues of blockchain technology and analyzed the architecture, components, and cryptographic algorithms of blockchain. Blockchain ensured secure transmission of information against fraud and hacking through distributed consensus and encryption. The study pointed out that blockchain technology was deficient in encryption and anonymity, which may weaken its ability to resist malicious behavior.⁷ Li X et al. designed a blockchain-based flight operation data sharing scheme for the conflict between data sharing and privacy protection in civil aviation enterprises. The scheme adopted hash anonymous authentication, zero-knowledge concise non-interactive knowledge argumentation, and proxy re-encryption, combined with grouping practical Byzantine fault-tolerant algorithms to reduce the consensus delay. Experiments demonstrated that the data sharing scheme improved the sharing efficiency while protecting data privacy, but its complexity and cost in large-scale practical applications still needed to be optimized.⁸ To optimize key generation, Vatambeti et al. looked at the use of blockchain technology in the healthcare industry and suggested a blockchain-based data broadcasting approach that combines homomorphic encryption with an opposing harmony search algorithm. The scheme securely transmitted data to a cloud server via blockchain and utilized a convolutional neural network for disease classification. Although the method performed well in terms of privacy protection and security, the efficiency and resource consumption in handling large-scale data still need to be improved.⁹ Mahajan H et al. proposed an architecture based on blockchain and lightweight cryptography for secure handling of genetic data in healthcare 4.0. The architecture used elliptic curve cryptography to protect data privacy and stored metadata in blockchain to prevent tampering. The outcomes revealed that the method had significant improvement in encryption and decryption efficiency, but its adaptability and scalability in different medical scenarios still needed to be further verified in practical applications.¹⁰ The above list of blockchain technology related research is shown in Table 1.

Table 1.

List of research related to blockchain technology.

Reference number	Research contents	Result
4 Dong L and others	The impact of traceability technology on incentives for supply chain members	Increase income, but may exacerbate pollution risks
5 Gupta R et al	Indian public distribution system blockchain traceability model	Enhance transparency, large-scale feasibility unverified
6 Hao F and others	The relationship between blockchain application and customer satisfaction in the catering industry	Enhance trust, the effect is affected by the size of the restaurant
7 Hashim A N	Analysis of blockchain architecture security and encryption algorithms	Enhanced security, insufficient anonymity weakens defense
8 Li X and others	Privacy protection scheme for flight data blockchain sharing	Balancing privacy and efficiency, high cost on a large scale
9 Vatambeti et al	Medical data blockchain broadcasting and homomorphic encryption optimization	Security improvement, inefficient large-scale data processing
10 Mahajan H and others	Lightweight encryption blockchain architecture for genetic data	High encryption efficiency, insufficient scalability in medical scenarios

In summary, a crucial component of FS oversight is food traceability. Currently, there are many uses for Internet of Things, blockchain, and encryption technologies in the food traceability space. These technologies also offer crucial technical assistance for FS and quality monitoring. However, the current blockchain-based FTS faces many deficiencies in information encryption, data access, and system wind control capabilities. For example, in the 2013 European horse meat scandal, the lack of transparency in supply chain information made it difficult to trace the source of pollution, resulting in a loss of consumer trust. Secondly, the system stability is poor. The tracing system failed to respond promptly and accurately to the 2018 romaine lettuce and E. coli contamination incident in the United States, resulting in an expanded impact. Furthermore, information integration is difficult. The data standards vary in different stages, such as a multinational food enterprise where it is difficult to achieve full traceability due to inconsistent data formats among suppliers. Finally, the high cost and complexity of technology limit the application of small and medium-sized enterprises.¹¹ At present, the immutability, decentralization, and traceability characteristics of blockchain can effectively ensure data security and privacy, and enhance supply chain transparency. The multi-chain architecture further optimizes system performance by distributing data from different links across multiple chains, achieving efficient storage and fast querying of data, enhancing system stability and scalability. Therefore, the research aims to design a food quality traceability system based on multi-chain architecture blockchain to improve the effectiveness of food quality supervision. One of the research’s two achievements is the development of a blockchain-based food quality traceability system (FQTS), which uses increasingly sophisticated communication and sensor networks to guarantee system stability. Secondly, the research introduces the Hyperledger Fabric network infrastructure framework and adopts data encryption technologies such as InterPlanetary File System (IPFS), Paillier, and Fabric Certificate Authority (Fabric CA) architecture to realize secure access and control of the system. The system can be accessed and controlled securely. The contribution of this research is mainly reflected in two aspects. The first point is that the research has made significant contributions in the field of food quality traceability. By constructing a traceability system based on a multi-chain architecture blockchain, it provides strong technical support for the transparency and traceability of all aspects of food production and consumption, and promotes the healthy development of the food industry. The second point is the contribution of research technology in the application field of blockchain technology, such as the introduction of Hyperledger Fabric network framework and Paillier algorithm, which enriches the application practice of blockchain in data encryption, access control, and other aspects, expands the application scope of blockchain technology in supply chain management and other fields, and provides technical support for information construction in related fields.

Methods and materials

This chapter mainly elaborates on the overall design and related technologies of a blockchain food quality traceability system based on a multi-chain architecture. First, the design of a multi-chain architecture blockchain food quality traceability system was introduced, which clearly includes key components such as service layer, data layer, physical layer, network layer, and interaction layer. The functions and hardware configurations of each layer were explained in detail. Then, the data storage and sharing technology of blockchain-based traceability system was discussed, and how to introduce Paillier algorithm and other methods to construct a data information encryption and sharing model to ensure data security was explained. Finally, the data access and query control technology of blockchain-based traceability system was introduced, and it was proposed to use Hyperledger Fabric framework and Fabric CA to achieve data access control, ensuring the secure and stable operation of the system.

Multi-chain architecture blockchain FQTS design

At present, the issue of FS is of great concern to society. In particular, once FQ problems occur in the production, transportation and wholesale of food, it will pose a great challenge to human health. Therefore, to ensure that the FQ meets the market requirements, good food traceability will be the key. The research is based on blockchain technology to design a traceability system with multi-chain architecture. In a food quality traceability system with a multi-chain architecture, efficient and secure interactions between multiple chains are achieved through cross chain communication tunnels and distributed network architecture. In addition, the zero-trust security gateway constructs a communication tunnel to verify cross chain requests, which can effectively ensure the security of data transmission. Under this system, data collection is encrypted and stored on the blockchain by the service layer. When cross chain synchronization occurs, it is verified through a security gateway. When querying, data is obtained from the corresponding chain according to permissions to ensure the security and effectiveness of information. Figure 1 depicts the system’s general design.

Figure 1.

Overall framework design of FQTS.

In Figure 1, the FQTS consists of five key components, including service layer, data layer, physical layer, network layer, and interaction layer. The physical layer is the basis of the system operation, which is mainly responsible for collecting and transmitting food data in each link through terminal service devices. The physical layer hardware configuration consists of multimodal Internet of Things terminals. It includes a high-precision temperature and humidity sensor (±0.5°C / ±2%RH) that supports the FQ system (hazard analysis and critical control point, HACCP) standard, an industrial-grade read-write with near-field communication as well as dual-frequency recognition of electronic tags, a 4K resolution H.265 encoded vision acquisition device, and an edge computing gateway that supports LoRaWAN/5G dual-mode transmission. In addition, the scale of the enterprise and the cost of hardware procurement for the system were taken into account when constructing the food quality traceability system. For environmental monitoring in the transportation process, medium precision sensors combined with dynamic threshold algorithms can meet most regulatory requirements, while high-precision equipment is retained in high-risk areas of the production process to achieve graded control. In addition, choose hardware that is compatible with open source protocols such as MQTT to avoid vendor binding and reduce long-term operational costs.

The data collected in the physical layer will be input into the service layer, which mainly realizes the automated analysis and encryption services for the traceability food data. Among them, the encryption and sharing algorithm based on blockchain technology is mainly adopted, which is responsible for data query, encryption, decryption and other functions. The data layer mainly adopts SQLite and blockchain-based CouchDBD databases to provide tamper-proof data storage services. The network layer adopts multi-channel network in conjunction with IPFS private network for the network layer design to ensure the security supervision of data as well as effective isolation.¹² In particular, the physical isolation of the blockchain subnet from the IPFS private storage network (Filecoin protocol) is achieved through software-defined wide area network (SD-WAN) technology. In addition, zero-trust security gateways are deployed to build cross chain communication tunnels. The core switching equipment is configured with 100Gbps throughput Clos network topology (CLOS) architecture distributed switches to ensure that the data transmission latency of various links such as production enterprises and logistics nodes is less than 20 ms. The interaction provides food multi-chain batch information query and supervision. The whole system is designed with multi-chain architecture, and the system will accurately trace the whole SC of food. It contains multiple traceability parts such as production enterprises, processing enterprises, logistics enterprises, warehousing enterprises, sales enterprises, and so on.¹³ The structure of multi-chain architecture traceability network is shown in Figure 2.

Figure 2.

Schematic diagram of multi-chain architecture traceability network.

The multi-chain architecture traceability network in Figure 2 shows that the whole traceability network consists of three parties, including regulators, food SC companies, and consumers at the end of the food supply. The traceability network is composed of IPFS private network and blockchain network. Due to the complexity of the internal links of different links, in order to guarantee the accuracy of the traceability service, the FQ traceability service will be realized through the joint participation of the three parties.

Blockchain-based traceability system data storage and sharing technology

Once the FQTS has been designed, the service layer will use a blockchain-based encryption technique to protect the system’s information and secure the data held within. In FQTS, every SC segment has a significant quantity of data. Ensuring that data is not erased or altered during transmission, exchange, or storage is essential. Table 2 displays the data for every connection in the SC.

Table 2.

Data division of each link in the traceability system.

Link	Privacy data	Public data
(1) Production	Manufacturer information, livestock and planting scale, sales information, and product parameters	Batch identification, livestock information, crop information, medication records, fertilizer records, planting environment data, and geographic location
(2) Processing	Purchase object, purchase information, processing worker information, and finished product information	Batch identification, production license, hygiene license, processing manufacturer, processing time, processing environment data, and additive records
(3) Transportation	Both parties’ enterprise information, order information (quantity, weight, type, etc.), transportation vehicle information, and transportation personnel information	Departure location, target location, transportation time, transportation environment data, multimedia recording, batch identification, and transportation company
(4) Storage	Batch storage information quantity and weight	Batch identification, inbound and outbound time, warehouse location, storage environment data, and storage record information
(5) Selling	Purchase batch information and sales information	Batch identification and sales location

In Table 2, the traceability system contains five sections of primary data, involving shareable public data as well as private data. For example, the public searchable data of food production segment data are planting or breeding time and location information, while the private information includes farming information, scale, and quantity sold. To secure the private data information, the research adopts Paillier algorithm as the basis to construct the data information encryption and sharing model. Blockchain data storage and sharing mainly includes three parts: generating secret key, data storage, and secret key acquisition.¹⁴ The whole technical blockchain data security storage and sharing framework is shown in Figure 3.

Figure 3.

Data storage and sharing technology framework.

In Figure 3, within the data IPFS message encryption framework, all nodes on the multi-chain architecture perform the initial step $S p (1^{\partial})$ and generate the global public key $P K$ , which is shown in equation (1).¹⁵

S p (1^{\partial}) \Rightarrow (P K)

(1)

In equation (1), $\partial$ denotes the secret key security parameters. Among them, $P K$ is mainly composed of various types of parameters in multi-chain nodes, while the master secret key under the blockchain is composed of conventions. The expression of global public key $P K$ is shown in equation (2).

P K = {p^{'}, q^{'}, p, q, n, n^{'}, R, g, θ}

(2)

In equation (2), $R \in Z_{n}^{*}$ is a random number. $p$ and $q$ denote two major prime numbers, which are important parameters of the secret key. $θ$ and $g$ are public keys. Parameters $n$ , $p^{'}$ , $q^{'}$ , and $n^{'}$ relations are shown in equation (3).

{\begin{cases} n = p q \\ p = 2 p^{'} + 1 \\ q = 2 q^{'} + 1 \\ n^{'} = p^{'} q^{'} \end{cases}

(3)

M K

is calculated as shown in equation (4).

M K = {R n^{'}}

(4)

Under the blockchain, $P K$ and $M K$ are used as inputs for the node to decrypt the private key $s k_{i}$ , then $s k_{i}$ is generated as shown in equation (5).

K G (P K, M K) \Rightarrow s k_{i}

(5)

Next, the data on the blockchain multi-chain architecture needs to be stored securely by setting the stored data plaintext as $T_{m}$ . Moreover, the Paillier cryptographic algorithm is used to encrypt the information to the ciphertext $C$ . This encryption process is shown in equation (6).¹⁶

E t (T_{m}, P K) \Rightarrow C

(6)

Then $C$ is saved into IPFS again and the hash cipher $H a s h (C)$ corresponding to the data is obtained and saved to the blockchain. This process requires dividing the corresponding ciphertext data in IPFS into k data blocks of size 256 KB, Meanwhile, IPFS utilizes content addressing to achieve data deduplication. When storing files, IPFS automatically divides the data into multiple blocks and calculates a unique hash value for each block. If different files have the same data block, they will share the storage location pointed to by the same hash value, thereby removing duplicate data. This not only reduces storage space usage, but also improves data storage efficiency. Divide k data blocks as shown in equation (7).

T_{m} \Rightarrow {b l o c k_{1}, b l o c k_{2}, b l o c k_{3}, . . ., b l o c k_{k}}

(7)

To secure the data, each data block is processed using 256-bit secure hash algorithm (SHA) for hashing operation. A unique identification code $I D$ is obtained for each data block. Then the secure storage of data is completed by returning the transaction $I D$ , as shown in equation (8).¹⁷

C h a i n (I P F S (C)) \Rightarrow t r a n s I D

(8)

To secure the stored data, the study utilizes threshold decryption $(t, n)$ to decrypt the encrypted stored file. In this case, there are $t$ receivers who get $H a s h (C)$ from within the multi-chain architecture federation. Moreover, the corresponding decrypted plaintext $m$ is obtained by joint decryption, and the process is shown in equation (9).

T_{m} = L (\prod_{i = 1}^{t} C_{i}^{2} \mod n^{2}) \times \frac{1}{4 Δ^{2} θ} \mod n

(9)

In equation (9), $L (\cdot)$ denotes the joint decryption operation. $t$ denotes decrypted secure channel data. $Δ$ and $n$ both denote the contribution parameters obtained through $P K$ . $\mod$ denotes the demodulation operation. In the decryption process, the obtained plaintext $m$ is passed to the $t$ users through the secure channel to obtain the decrypted data.

Blockchain-based traceability system data access and query control technology

In the traceability system of chain architecture blockchain, the access and query of blockchain data often have problems such as permissions and data leakage. To address the above problems, the research proposes a system access and confidentiality control technology. Among them, the traceability system base network is the enterprise-level Hyperledger Fabric framework, and Fabric CA is used to replace the original secret key center. At the same time, proxy re-encryption is introduced into Fabric CA, and data access control is realized according to system user rights. The data access and query sequence of FQTS is shown in Figure 4.

Figure 4.

Data access and query sequence of FQTS.

The timing diagram in Figure 4 shows that the system access and query also contains three main processes, including secret key generation, data storage, and data acquisition. In the system access control, Fabric CA is used as the secret key management center, which performs the first step of the initial operation to generate $P K$ as well as $M K$ . In the access system control, the elliptic curve algorithm is used to pair the access user with a valid public-private key pair $(p k_{i}, s k_{i})$ . Its generation is shown in equation (10).¹⁸

K G (P K, M K) \Rightarrow (p k_{i}, s k_{i})

(10)

The key generation process, defines the elliptic curve on the finite neighborhood $F_{P}$ as $E$ , and one of its upper datums is set to $G$ . The secret key will be released through the public parameters of the system as well as $P K$ jointly, then $P K$ is expressed as shown in equation (11).¹⁹

P K = {E, F_{p}, G}

(11)

Setting a random parameter $S_{K I}$ as the data owner private key $S_{K R}$ , the data owner public key $P_{D O}$ is shown in equation (12).

P_{D O} = S_{K I} \cdot G

(12)

The public key $P_{D R}$ expression of the data receiver is shown in equation (13).

P_{D R} = S_{K R} \cdot G

(13)

In equation (13), $S_{K R}$ denotes the data receiver private key. Then the storage of the data is carried out. The data owner will encrypt the data with the private key $s k_{i}$ to obtain the ciphertext $C$ and generate the corresponding ciphertext identifier $C^{n}$ .²⁰ This encryption process is shown in equation (14).

E t (R, s_{i}) \Rightarrow C

(14)

In equation (14), $R$ denotes the encrypted data. The ciphertext $C$ employs $P_{D O}$ to encrypt the dense plaintext and obtain the ciphertext data $C_{i}$ with the data owner picking a random number $r_{i}$ to generate a random ciphertext identifier $C_{n}^{i}$ . Then the expression of the ciphertext $C$ is shown in equation (15).²¹

C = {\begin{cases} C_{i} = m_{i} + r_{i} \cdot P_{D O} \\ C_{i}^{n} = r_{i} \cdot G \end{cases} i \in (1, 2, 3, . . ., n)

(15)

During the data acquisition process, the system content data storage contract will match a serial number $i$ for the access user ciphertext. This serial number will reflect the current access user ordering and be unique. The ciphertext identifier of the $i$ th access user is identified as $C_{n}^{i}$ .²² In addition, the data owner likewise invokes the contract and reconstructs the encryption key $r k_{a \to b}$ , which serves to delimit the data access privilege zone. The re-encryption is shown in equation (16).

R e K (s k_{a}, p k_{b}) \Rightarrow r k_{a \to b}

(16)

In equation (16), $s k_{a}$ and $p k_{b}$ are re-encryption keys, in which the data $b$ is divided from $a$ block data. When the data receiver wants to access the data, it will send a request to the multi-chain architecture federation node, and the system will verify the access user information with permission. If the user has access requirements, the data owner encrypts all the ciphertext data $C_{i}$ using $r k_{a \to b}$ . Whereas, the data receiver decrypts the data by the corresponding private key $C_{i}^{'}$ to obtain the specified access data $R_{i}$ .^23,24 The access data $R_{i}$ process is shown in equation (17).

D e c r y p t (s k_{b}, {C_{1}, C_{2}, . . . C_{i}}) \Rightarrow {R_{1}, R_{2}, . . ., R_{i}}

(17)

The entire technical process of the food quality traceability system based on multi-chain architecture blockchain is shown in Figure 5.

Figure 5.

Process of food quality traceability technology based on multi-chain architecture and blockchain.

According to Figure 5, this technology is based on the service layer and uses encryption and sharing algorithms based on blockchain technology to automatically analyze and encrypt data. It is responsible for functions such as data query, encryption, and decryption, thereby achieving food quality traceability and information management

Results

This chapter mainly presents and analyzes the security and performance test results of the blockchain food quality traceability system based on multi-chain architecture. Firstly, system security testing and analysis were conducted. By comparing with encryption algorithms such as SM4 and AES, the superiority of the proposed technology was verified in terms of encryption performance, data storage and decryption performance, and ciphertext change rate. This indicates that the new system has more efficient and secure features in data encryption, storage, and decryption. Next, we will conduct performance experiments on the food quality traceability system, comparing the performance of the new and old systems in terms of throughput, latency, risk control capability, error rate, and resource utilization, to verify the practical application effect of the research technology.

System security testing

To verify the feasibility of the proposed technique, the research will next carry out the corresponding experiments. The experiments are conducted using 64-bit Ubuntu 22.04 LTS, and the training framework is Python 3.10. The experimental hardware: the processor is AMD Ryzen 5 5600, the graphics card is NVIDIA GeForce GTX 1650 4 GB, the RAM is 16 GB DDR4, and the hard disk is 1T Samsung 970. The multi-chain architecture traceability system is based on Hyperledger Fabric. In the whole FQ traceability network, four independent node networks are set up in the production chain, logistics and transportation links, warehousing links, and sales links, and at the same time, an additional consumer node network is added. The specific parameters of the multi-chain architecture traceability system are shown in Table 3.

Table 3.

Multi-chain architecture traceability system information.

Index	Parameter
Number of organizational enterprises	16
Number of network links	27
Number of IPFS nodes	7
Number of block transactions	25
Maximum block capacity	20 MB
Block generation time	2s
Number of independent node networks in the system	5

In the FQTS link network, there are four types of enterprises, including agricultural production and processing enterprises, logistics enterprises, warehousing enterprises, and sales enterprises. A total of 27 network chains are set up for effective traceability of FQ. Meanwhile, 7 IPFS nodes are set up in the network to ensure the security of information data. Next, the system security will be tested, and the study introduces Chinese national standard symmetric cipher SM4 (SM4) and advanced encryption standard (AES) as the benchmarks for data encryption experimental testing. In the test, the encryption model Paillier Fabric CA proposed by the research institute was paired with Paillier CA, and the encryption performance of different models was compared as shown in Figure 6.

Figure 6.

Comparison of encryption performance of encryption algorithms under different key lengths. (a) 32-bit key length and (b) 256-bit key length.

Figure 6(a) shows the encryption performance test under 32-bit secret key length. According to the test structure, as the encrypted data size increases, the encryption time of each encryption model is increasing. Among them, the research encryption model has an obvious advantage over the AES model. For example, when the data size is 2500 kb, the encryption time of the research model is 22 ms, while that of AES is 28 ms and that of SM4 is 25 ms. Next, the secret key length is set to 256 bits, and the encryption performance of different algorithms is compared, as shown in Figure 6(b). The overall encryption time of the research model is significantly lower than that of AES and SM4. For example, when the data size is 1000 kb, the encryption time of AES, SM4, and the research model is 2675 ms, 2203 ms, and 1806 ms, respectively. Then, the research compares the data storage and decryption performance of the three algorithms, as shown in Figure 7.

Figure 7.

Data storage and decryption performance test. (a) Data storage time consumption and (b) data decryption performance test.

Figure 7(a) shows the comparison of data storage time consuming. According to the data, the increase of the secret key length makes the storage time of each algorithm expand. Among them, AES and SM4 have close performance when the secret key is 8bit to 32 bit. However, as the secret key length increases, SM4 storage time is lower than AES and performs better. Overall, the research model storage time is significantly better than SM4 and AES. For example, when the secret key is 64 bit, the overall storage time of AES, SM4, and the research model are 128 ms, 112 ms, and 43 ms, respectively. Figure 7(b) shows the data decryption performance test. The decryption performance of the three algorithms is close when the secret key length is in the interval of 8 bit to 32 bit. However, when the secret key length is above 64 bit, the research model has an obvious advantage over other models. For example, when the secret key length is 128 bit, the decryption time is 985 ms for AES, 723 ms for SM4, and 443 ms for the research model. Next, the study introduces the change rate (the ratio of the number of changed words to the total number of characters in the ciphertext) to reflect the actual encryption performance of the different algorithms. The higher the value of change rate, the better the obfuscation of the encryption algorithm. Among them, the key length is set to 32, the experimental plaintext length is set, and the number of plaintext tests under different nodes is 10. Figure 8 displays the specific test findings.

Figure 8.

Results of ciphertext change rate for different algorithms. (a) AES, (b) SM4, and (c) Paillier CA.

Figure 8(a)–(c) show the cipher change rate of AES, SM4, and research models, respectively. The change rate of AES model is in the range of 94.5%–96.7%, and it is unevenly distributed among multiple node intervals. In the SM4 model, the ciphertext change rate ranges from 96.4% to 98.3%. Its distribution is less uniform in node 10 to node 15 and node 25 to node 35. In the experiments with the research model, its change rate is in the range of 96.4%–98.2% and the overall distribution is uniform. This indicates that the research model has better cryptographic obfuscation and is more resistant to means such as brute force decryption.

FQTS performance test

Next, the study compares the effectiveness of FQTS in practical application and compares it with the old system. In particular, the old system uses a centralized database-based system and does not have end-to-end data encryption and advanced access control mechanisms. In addition, the old system uses a traditional distributed sensing network and does not support network standards such as 5G networks and edge computing. The system architecture adopts a single server or cluster to manage data, without distributed node verification mechanism, using conventional relational databases (SQL), and data storage without encryption or only basic access control. Next, the system throughput (transactions per second, TPS) and latency are compared, as shown in Figure 9.

Figure 9.

Comparison of throughput performance and latency performance of the system. (a) System throughput testing and (b) system latency testing.

Figure 9 shows the system throughput test results. In the uplink TPS test, the throughput of both systems gradually increases with the increase of test times in the first 1000 tests. Then the new system maintains a stable state, while the old system shows a downward trend in throughput. Overall, the average value of the uplink throughput of the old system is 453TPS, and the new system is 652TPS. In the query throughput test, the throughput of the new system is also significantly better than the old system, with an average value of 412TPS, while the old system has an average value of 305TPS. Next, the two systems are compared with each other in terms of their ability to control the risk when accessing the data, with the number of risks being set to 100. Figure 10 displays the test’s outcomes.

Figure 10.

Comparison of access control between multi-chain architecture system and old systems.

Figure 10(a) and (b) show the results of access control comparison between the old system and the new system, respectively. The risk interception capability of the new system is significantly better than that of the old system. When the access data is 50, the number of risk interception of the new system is 92, while the old system is only 52. Meanwhile, in the comparison of interception time consuming, the average interception time consuming of the new system is 902 ms, and that of the old system is 1205 ms. Finally, symbol error rate (SER) and resource occupancy rate are introduced to conduct a comprehensive test of the system, as shown in Table 4.

Table 4.

System comprehensive performance test.

Test number	Old system			Multi-chain architecture system
Test number	Risk interception rate (%)	SER (%)	CPU resource consumption rate (%)	Risk interception rate (%)	SER(%)	CPU resource consumption rate (%)
1	72.64	0.0045	72.45	97.45	0.000054	41.58
2	78.67	0.0066	81.41	98.15	0.000059	43.48
3	81.43	0.0057	75.45	97.97	0.000068	40.58
4	75.76	0.0085	76.45	99.64	0.000075	43.45
5	76.38	0.0086	81.45	97.84	0.000067	44.84
6	75.68	0.0047	84.79	98.73	0.000076	39.48
7	79.42	0.0075	84.47	99.18	0.000063	46.87
8	80.72	0.0053	83.48	97.46	0.000075	45.82
9	79.32	0.0024	80.45	98.75	0.000065	46.18
10	74.68	0.0093	76.48	99.18	0.000095	44.78
11	81.24	0.0088	79.78	98.35	0.000073	49.48
12	72.68	0.0075	80.15	99.73	0.000065	50.75
13	71.98	0.0033	79.19	98.13	0.000071	44.85
14	79.26	0.0004	74.98	98.38	0.000073	46.79
15	81.34	0.0018	81.45	97.18	0.000063	48.15
16	79.65	0.0075	80.75	97.86	0.000057	46.84
Mean value	77.55	0.0059	79.57	98.37	0.000069	45.25

Table 4 shows the results of the comprehensive system testing. In this case, in FQ traceability, the data security threshold requires the data SER to be less than 0.0001% while greater than 0.01% is considered as low to medium risk. The research system is superior in risk interception rate, SER, and CPU resource utilization. For example, the mean value of SER of the new system is 0.000069%, which satisfies the system data completeness criteria. In addition, the average value of the risk interception rate of the new system reaches 98.37%, which is better than the old system’s 77.55%. Moreover, the GPU resource of the new system is lower than that of the old system, with an average value of 45.25%. In addition, the average risk interception rate of the new system reached 98.37%, which is better than the 77.55% of the old system. And the GPU resources of the new system are lower than those of the old system, with an average of 45.25%. Finally, comparing the energy consumption of blockchain consensus processes in different systems, the multi-chain architecture system adopts PBFT multi-chain distributed architecture. As shown in Table 5.

Table 5.

Comparison of energy consumption in blockchain consensus processes of different systems.

Test indicators	Multi-chain architecture system (PBFT multi-chain)	Old system (centralized database)	Bitcoin (PoW)	Wave field (DPoS)
Single transaction energy consumption (J)	0.85	0.15	900	0.07
CPU resource ratio (consensus process)	28%	5%	95%	15%
Network message volume (KB/10000 transactions)	1,200	300	50,000	800
Storage growth (MB/10000 transactions)	42	30	1,000	35
Risk interception energy consumption ratio (times/kWh)	102	43	0.5	150

According to the test results in Table 5, the energy consumption of a single transaction in the new system PBFT is 0.85 J, significantly lower than PoW’s 900J, but higher than the old system’s 0.15 J. The main reason is that PBFT requires multi node collaborative verification (pre preparation, preparation, and submission stages), and the computational overhead is concentrated in signature verification, which accounts for 28% of the CPU, and message digest processing. In terms of communication overhead comparison, the O (N²) communication complexity of PBFT results in a message load of 1.2 GB per 10000 transactions, while the old system only has 0.3 GB. However, the multi-chain architecture reduces cross node broadcasting through shard isolation (production/logistics/warehousing independent sub chains), reducing message volume by 40% compared to single chain PBFT. Finally, in the risk interception energy efficiency ratio, the new system can intercept 102 risks per kilowatt hour (43 for the old system), which is supported by PBFT’s Byzantine fault-tolerant mechanism (tolerating ≤1/3 malicious nodes) to achieve high-precision risk control. The cost is an increase in energy consumption, but the safety benefit per unit of energy consumption has increased by 137%.

Discussion and conclusion

In the FQ traceability neighborhood, blockchain technology has attracted much attention because of its tamperability, decentralization, and traceability. To improve the effectiveness of FQ regulation, research is conducted to design a FQTS based on multi-chain architecture blockchain and to improve SC transparency and consumer trust. The system introduced the Hyperledger Fabric network infrastructure framework and used data encryption technology such as IPFS, Paillier, and Fabric CA to realize the secure access and control of the system.

According to the experimental findings, the new system performed better than the conventional system in terms of data encryption, storage, and decryption. For example, under 32-bit key length, the encryption time of the new system was 6 ms faster than that of AES model and 3 ms faster than that of SM4. In the data storage and decryption performance tests, the new system also showed better performance, for example, under 64-bit key length, the storage time of AES, SM4, and the new system was 128 ms, 112 ms, and 43 ms, respectively. The decryption time was 985 ms, 723 ms, and 443 ms, respectively. In addition, the new system also displayed higher cryptographic obfuscation in the change rate test, which indicated that it was more resistant to attacks such as brute force decryption.

In summary, the system has excellent security and system stability in terms of data security and system access control to meet the requirements of more demanding FS traceability scenarios. Although the research has achieved significant results in FQTS security and stability, there are still shortcomings. For example, the network links built are relatively fixed, adding or deleting needs to update the network in a timely manner, and its flexibility needs to be optimized at a later stage. Second, the data encryption process is still cumbersome, and complex scenarios still face high hardware occupancy requirements. Finally, when the amount of data increases sharply, the encryption and decryption processes of the system may become slow, affecting overall performance. Therefore, in the future, it is necessary to further simplify the design of data encryption technology. Meanwhile, optimizing data encryption algorithms, such as introducing more efficient encryption algorithms like lightweight encryption algorithms, can reduce the computational overhead during the encryption and decryption processes. In addition, dynamic network link management mechanisms can also be developed to allow the system to automatically add or remove nodes at runtime without manually updating network configurations.

Footnotes

ORCID iD

Min Li

Funding

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

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Wang

Scrimgeour

. Consumer adoption of blockchain food traceability: effects of innovation-adoption characteristics, expertise in food traceability and blockchain technology, and segmentation. Br Food J 2023; 125(7): 2493–2513.

Singh

Gutub

Nayyar

, et al. Redefining food safety traceability system through blockchain: findings, challenges and open issues. Multimed Tool Appl 2023; 82(14): 21243–21277.

Sugandh

Nigam

Khari

, et al. An approach for risk traceability using blockchain technology for tracking, tracing, and authenticating food products. Information 2023; 14(11): 613–621.

Dong

Jiang

. Impact of traceability technology adoption in food supply chain networks. Manag Sci 2023; 69(3): 1518–1535.

Gupta

Shankar

. Managing food security using blockchain-enabled traceability system. Benchmark Int J 2024; 31(1): 53–74.

Hao

Guo

Zhang

, et al. Blockchain= better food? The adoption of blockchain technology in food supply chain. Int J Contemp Hospit Manag 2024; 36(10): 3340–3360.

Hashim

. Blockchain technology, methodology behind it, and its most extensively used encryption techniques. Al-Salam J Enginee Tech 2023; 2(2): 140–151.

Zhao

Deng

. BFOD: blockchain-based privacy protection and security sharing scheme of flight operation data. IEEE Internet Things J 2023; 11(2): 3392–3401.

Vatambeti

Krishna

ESP

Karthik

, et al. Securing the medical data using enhanced privacy preserving based blockchain technology in Internet of Things. Clust Comput 2024; 27(2): 1625–1637.

10.

Mahajan

Reddy

KTV

. Secure gene profile data processing using lightweight cryptography and blockchain. Clust Comput 2024; 27(3): 2785–2803.

11.

Cao

Johnson

Tulloch

. Exploring blockchain-based traceability for food supply chain sustainability: towards a better way of sustainability communication with consumers. Procedia Comput Sci 2023; 217(5): 1437–1445.

12.

Verma

Kanrar

. Secure document sharing model based on blockchain technology and attribute-based encryption. Multimed Tool Appl 2024; 83(6): 16377–16394.

13.

Aggarwal

Rani

, et al. Revolutionizing agri-food supply chain management with blockchain-based traceability and navigation integration. Clust Comput 2024; 27(9): 12919–12942.

14.

Susanty

Puspitasari

Rosyada

, et al. Design of blockchain-based halal traceability system applications for halal chicken meat-based food supply chain. Int J Inf Technol 2024; 16(3): 1449–1473.

15.

Gousteris

Stamatiou

Halkiopoulos

, et al. Secure distributed cloud storage based on the blockchain technology and smart contracts. Emerg Sci J 2023; 7(2): 469–479.

16.

Ellahi

Wood

Bekhit

AEDA

. Blockchain-based frameworks for food traceability: a systematic review. Foods 2023; 12(16): 3026–3028.

17.

Perumalsamy

Kaliyamurthy

. Blockchain non-fungible token for effective drug traceability system with optimal deep learning on pharmaceutical supply chain management. Eng Technol Appl Sci Res 2025; 15(1): 19261–19266.

18.

Mahalingam

Sharma

. An intelligent blockchain technology for securing an IoT-based agriculture monitoring system. Multimed Tool Appl 2024; 83(4): 10297–10320.

19.

Panghal

Pan

Vern

, et al. Blockchain technology for enhancing sustainable food systems: a consumer perspective. Int J Food Sci Technol 2024; 59(5): 3461–3468.

20.

Shehzad

. Predictive AI models for food spoilage and shelf-life estimation. Glob Trends Food Sci Technol 2025; 1(1): 75–94.

21.

Wang

. Research on the influencing factors of block chain technology adoption in supply chain finance of small and medium-sized enterprises. Adv Manag Sci 2023; 12(1): 1–3.

22.

Chu

. Exploration of new energy vehicle recycling model based on blockchain. Adv Ind Eng Manag 2023; 12(2): 110–113.

23.

Odu

. Blockchain as a disruptive technology in Nigeria’s Fintech ecosystem. Inform manag comp sci 2022; 5(1): 03–10.

24.

Wang

. Integration of big data innovation and traditional marketing theory: an economic analysis of enhancing market competitiveness in the food and beverage industry. Adv Ind Eng Manag 2024; 13(4): 295–298.