Ilmu Komputer & AI editorial
Open AccessOA2023
Block Chain for IOT Security Using Consensus Algorithms
A Double-Chain Framework with Enhanced PBFT and Partial Blind Signatures for Secure IoT Data Exchange
P. Kalpana; I. Anusha Prem· Data Analytics and Artificial Intelligence· 2023· DOI 10.46632/daai/3/2/16
The core problem
The Internet of Things (IoT) has rapidly expanded, connecting billions of devices that generate and exchange vast amounts of data. However, this growth introduces significant security and privacy challenges, particularly in decentralized environments where trust is not inherently established. Blockchain technology, originally designed for cryptocurrency, offers a promising solution due to its decentralized, tamper-proof, and transparent nature. The first distributed recordkeeping system with a built-in trust structure is the block chain. It creates a dependable architecture for decentralized control through information redundancy across multiple nodes. This paper addresses the need for a secure IoT information exchange framework by leveraging blockchain and consensus algorithms. The authors propose a minimal blockchain-based framework that ensures data integrity, privacy, and efficient resource transfer. The core research question focuses on evaluating the performance of consensus algorithms in IoT contexts, specifically their time to achieve consensus, which must be minimal for practical deployment. The study assesses three popular consensus algorithms—modified Proof of Work (PoW)
Innovation
The simulations on Contiki demonstrate that the time to achieve consensus is less than seconds for all three algorithms under various conditions. Specifically, modified PoW shows higher latency due to its computational intensity, but remains feasible for low-power IoT devices when difficulty is adjusted. Realistic Byzantine fault tolerance (PBFT variant) achieves consensus faster than modified PoW, with times often under 1 second for networks up to 50 nodes. Binary consensus exhibits the lowest latency, making it suitable for resource-constrained devices, but it may require additional assumptions about network synchrony. The results indicate that the number of participating nodes significantly impacts consensus time: as increases, time grows polynomially for PBFT and binary consensus, while modified PoW grows exponentially without difficulty adjustment. The radio propagation model also affects performance; log-normal shadowing introduces more packet loss, increasing consensus time compared to the unit disk graph model. Mote type plays a role: more powerful motes (e.g., Tmote Sky) achieve faster consensus than constrained ones (e.g., MicaZ). The improved PBFT mechanism reduces c
The Internet of Things (IoT) has rapidly expanded, connecting billions of devices that generate and exchange vast amounts of data. However, this growth introduces significant security and privacy challenges, particularly in decentralized environments where trust is not inherently established. Blockchain technology, originally designed for cryptocurrency, offers a promising solution due to its decentralized, tamper-proof, and transparent nature. The first distributed recordkeeping system with a built-in trust structure is the block chain. It creates a dependable architecture for decentralized control through information redundancy across multiple nodes. This paper addresses the need for a secure IoT information exchange framework by leveraging blockchain and consensus algorithms. The authors propose a minimal blockchain-based framework that ensures data integrity, privacy, and efficient resource transfer. The core research question focuses on evaluating the performance of consensus algorithms in IoT contexts, specifically their time to achieve consensus, which must be minimal for practical deployment. The study assesses three popular consensus algorithms—modified Proof of Work (PoW), Practical Byzantine Fault Tolerance (PBFT), and binary consensus—under various conditions including mote type, number of participating nodes, and radio propagation model. A comprehensive solution is put forward to enable an IoT node to switch between different consensus algorithms dynamically, with simulations conducted on the Contiki IoT operating system demonstrating strong performance (time to achieve consensus less than seconds).
The proposed framework employs a double-chain approach combining a data blockchain and a transaction blockchain. The data blockchain handles distributed storage and tamper-proof data, while the transaction blockchain manages resource and data transfers, as well as privacy protection. To enhance consensus efficiency, the authors improve the practical Byzantine fault-tolerant (PBFT) mechanism. Additionally, better partial blind signature-based algorithms are integrated into the transaction blockchain to enhance data registration efficiency and privacy protection. The performance evaluation focuses on three consensus algorithms: modified PoW, realistic Byzantine fault tolerance (a variant of PBFT), and binary consensus. These are assessed under varying conditions: mote type (e.g., Tmote Sky, MicaZ), number of participating nodes (ranging from small to large networks), and radio propagation models (e.g., unit disk graph, log-normal shadowing). Simulations are conducted using the Contiki IoT operating system, a popular open-source OS for IoT devices. The key metric is the time to achieve consensus, which must be minimal for IoT applications. The study also proposes a comprehensive solution that allows an IoT node to switch between consensus algorithms based on network conditions, ensuring adaptability and efficiency. The mathematical model for consensus time can be expressed as:
Why it matters
The study highlights the trade-offs between consensus algorithms for IoT security. Modified PoW, while secure, is energy-intensive and may not be suitable for battery-powered devices unless combined with lightweight variants. PBFT offers a good balance between security and efficiency, but its quadratic message complexity can limit scalability. Binary consensus is fastest but may sacrifice fault tolerance in asynchronous networks. The proposed double-chain architecture effectively separates data storage from transaction processing, enhancing privacy and efficiency. The improved PBFT and partial blind signatures contribute to better performance and privacy protection. The ability for an IoT node to switch between consensus algorithms based on network conditions is a significant contribution, enabling adaptability. However, the study has limitations: simulations are based on Contiki and may not fully capture real-world deployment challenges such as mobility and dynamic network topology. Future work could explore hybrid consensus mechanisms and integration with edge computing. The taxonomy candidates for this work include Architecture (double-chain), Cybersecurity (privacy, tamper-proof), Network (IoT, radio propagation), and Cryptography (blind signatures, consensus). The findings suggest that a one-size-fits-all consensus algorithm is not optimal for IoT; instead, a dynamic selection mechanism is recommended. The time to achieve consensus remains a critical metric, and the study demonstrates that with proper optimization, blockchain can be viable for IoT security.
Who should read this
CS practitioners and researchers
Opening member content…