The paper is organized as follows: Sect. 2 presents the literature review, Sect. 3 presents the methodologies used in the work, Sect. 4 offers the proposed model, Sect. 5 analyses the result using different experiments, and Sect. 6 concludes the work.
A lightweight BC has been developed for secure industrial operations and data regulation, using real-time cryptographic algorithms on an ARM Cortex-M4 processor and an efficient consensus mechanism proof of authentication (PoAh).
The study tested a design for smart industries, transforming a fruit processing plant into a secure industrial platform. It proposed a decentralized, secure access management method for IoT, combining multi-agent techniques with BCT and creating a BC manager to enhance control and secure connections. The study introduces a healthcare security model using BCT data encoding to encrypt sensitive data on cloud servers. It employs a 128-bit Advanced Encryption Standard (AES) key and SHA-256 hashing algorithm to prevent unauthorized access.
The model’s real-world applicability is demonstrated through an interface-based system, outperforming Rivest-Shamir-Adleman (RSA) and Digital Signature Algorithm (DSA) in metrics. The study introduces a secure BCT-based data storage system that combines user authentication and health condition prediction. It uses reversed public keys (PuK) and Private Keys (PrK), and the reversed PuK-PrK combine Rivest-Shamir-Adleman. (RP2-RSA) for enhanced security, and the Feature Selection (FS) correlation factor-induced Salp Swarm Optimization Algorithm (SSOA). The study explores the integration of BCT, Smart Contracts (SC), and IoT devices in agricultural processes, focusing on pre-and post-harvest stages. It presents a model where BC is the basis, IoT devices collect data, and SC manages interactions.
A survey that verifies experts’ BCT measures BCT’s maturity in the agricultural supply chain. SC, IoT, and transaction records emphasize BCTs’ importance in agriculture and provide a maturity model. Sensor data prediction, indoor parameter computation, and actuator regulation were introduced in a BCT-Enabled Greenhouse System Optimization Method (BCT-EGSOM). The system achieved 19% energy savings and 41% advantage over baseline methods. Using the Hyperledger Fabric network, a proof-of-concept showed Network Throughput (NT), End-to-End Delay (EED), and Resource Utilization (RU). The integration of IoT sensing with BCT will enhance SAP. It uses controlled sensing devices as reliable data sources for permissionless BCs, demonstrating its efficiency and practicality across multiple platforms. The study proposes a dual-BCT using the InterPlanetary File System (IPFS) to secure agricultural data in IoT networks.
The study employs a vision mechanism to retrieve data segments and a consortium BCT, the Agricultural Sample Data Chain (ASDC), based on Ethereum. This model provides faster and greater resilience than solo BCT storage. The model presented a significant advancement in the IoT domain, particularly in the agricultural sector. Their research addresses the challenges posed by the big data generated by a network of agricultural machines and service centres. Recognizing the strain this data can place on network traffic and storage systems, especially in low-latency applications like health monitoring of agricultural machinery, the study proposes an innovative solution. The following Table 1 presents the overview of the above-discussed literature.
The study invented a novel BCT-based decentralized model for cloud resource management in Edge Computing (EC). This approach addresses the inherent trade-offs between consistency and availability in controller design patterns, as the Consistency, Availability, and Partition Tolerance (CAP) theorem dictates. Deploying Byzantine-resistant consensus algorithms, their model demonstrates improved system availability during network partitions, maintaining configuration consistency.
The article introduces an advanced lightweight authentication system for IoT devices in healthcare networks, using Physically Unclonable Functions (PUF). Addressing the vulnerability of these devices to attacks, their system ensures robust authentication without relying on stored sensitive data and enhances device privacy using temporary identities. This method significantly improves IoT device security, providing a more effective and secure alternative to traditional security measures.
The authors present Subspace Clustering using an Evolutionary algorithm, Off-Spring generation, and Multi-Objective Optimization (SCEOMOO), targeting challenges in subspace clustering in high-dimensional data sets. Focusing on determining cluster numbers and optimizing subspaces, SCEOMOO shows considerable improvements in performance compared to existing methods, validated on six standard data sets.
The model investigates security and privacy challenges in green Agri-IoT. They propose a four-tier architecture and classify threats into five types: privacy, authentication, confidentiality, availability, and integrity. The study also reviews and compares current secure and privacy-preserving IoT, focusing on their adaptation for Agri applications. Additionally, it explores BC solutions and consensus algorithms tailored for green Agri-IoT, identifying future research directions in this domain.
Authors address security threats in IoT-based manufacturing. The study underscores how IoT developments in manufacturing introduce new risks, particularly in cyber-physical systems. It presents two classifications to order cyber-physical attacks and quality check measures, linking manufacturing process variations with potential security attacks. The research highlights the importance of understanding and mitigating these emerging vulnerabilities in IoT-enabled manufacturing.
Recent advancements in the convergence of Federated Learning (FL), BC, and Reinforcement Learning (RL) have significantly enhanced privacy preservation in the Industrial Internet of Things (IIoT) and intelligent systems.
The hierarchical FL for anomaly detection in IIoT environments, leveraging Deep Reinforcement Learning (DRL) to optimize local models without requiring raw data sharing. By quantifying privacy leakage degrees and incorporating adaptive learning policies, the model achieved high detection accuracy, low latency, and improved privacy robustness across distributed devices.
In a complementary direction, introduced a Blockchain-based Secure Data Aggregation (BSDA) for edge computing in IoT, where task security levels were integrated into block headers. Using a self-adaptive double bootstrapped Deep Deterministic Policy Gradient (DDPG), BSDA successfully preserved task-level privacy while optimizing energy-efficient data collection and task delegation. In the context of 5G-enabled IIoT and Intelligent Transportation Systems (ITS), privacy-aware routing and trust management have emerged as key focus areas.
The model developed Quantitative Structure-Property Relationships (QoSPR), a Quality of service (QoS)- and privacy-aware routing protocol using FL-RL to develop optimal gateway deployment methods that balance latency, load distribution, and data privacy during inter-device routing.
Developing further on BC’s immutability and traceability, advised BHTE, a heterogeneous BC-based hierarchical trust evaluation method for 5G-ITS. By integrating federated Deep Learning (DL) with multi-level trust assessment and BC-backed trust storage, BHTE effectively resists trust-based manipulation while maintaining high system throughput. Collectively, these works highlight the emerging trend of privacy-aware, decentralized intelligence in complex IoT ecosystems — a method that inspires further enhancements to the privacy and trust mechanisms in the proposed Agri-IoT.
The model presents a bibliometric analysis of the integration of IoT + BC in agriculture, highlighting a significant annual growth rate of 47.58% in related research. It underscores the potential of these technologies in enhancing SAP and environmental monitoring.
The research introduces a BC designed to enhance land management systems by addressing problems such as data fragmentation and lack of transparency, thereby improving efficiency and trust among stakeholders.
Authors examine the innovative integration of BC within agricultural and livestock IoT, discussing how this convergence enhances operational security and transparency. The study proposes a secure, privacy-preserving IoT communication model for precision agriculture by integrating BC + FL. The method aims to enhance data security and privacy in agricultural IoT. The research [38] surveys the convergence of AI, BC, and quantum cryptography to enhance security in IIoT, which can be extended to agricultural applications.
From the above literature survey, while many works have explored BC applications across various sectors, they frequently lack a comprehensive method personalized to the specific convolutions of Agri-IoT. This research gap has motivated the proposed model to introduce a distinctive multi-tiered BCT, uniquely integrated with a QNN + BO, coupled with an ECC-Chaotic Map (EC³) algorithm. This model explicitly addresses the failed requirements in Agri-IoT, such as advanced data encryption, efficient processing capabilities, and scalability, which are crucial for modern agricultural tasks. By focusing on these key areas, the proposed QNN + BO enhances data security and management in agriculture and improves operational efficiency, filling a significant gap in the existing research and contributing to a novel contribution to Agri-IoT.
