计算机科学
可扩展性
无线传感器网络
智能对象
分布式计算
计算机网络
方案(数学)
块(置换群论)
计算机安全
数据库
物联网
数学分析
数学
几何学
作者
Pronaya Bhattacharya,Amod Kumar Tiwari,Ashwin Verma,Abdulatif Alabdulatif,Sudeep Tanwar,Ravi Sharma
标识
DOI:10.1016/j.csi.2023.103785
摘要
Industrial Internet-of-Things (IIoT) plays a vital cog in Industry 4.0 design principles and has revolutionized the manufacturing process by shifting from traditional mechanical systems to smart automation. In IIoT, smart embedded sensors are placed strategically on machine parts, components, and devices. These sensors exchange data over open public channels (Internet), which poses high security and privacy risks to the operational processes, and could lead to machinery failure. The problem increases fourfold in distributed systems, where trust and security are crucial factors. Thus, recent schemes have combined blockchain (BC) with crypto-based primitives in IIoT, but less focus is placed on the resource requirements of such systems. Thus, this paper proposes a scheme named LightBlocks that combines lightweight crypto-primitives with scalable BC operations and presents a low-powered consensus approach, which is appropriate for IIoT ecosystems. The scheme allows the local sensor nodes (LSN) to identify a cluster head (CH), which forwards data to gateway nodes (GN). At GN, a low-powered signcryption scheme is proposed for sensor data. Once the data is signed and energy-efficient consensus Proof-of-Light (PoL) is proposed, block validations are done through selected verifier nodes in a defined consensus period only. Due to this, the transactional throughput improves significantly, and mining overheads are also reduced. We present the attacker model and propose a formal security model to evaluate the security strength of the scheme. The simulation compares the scheme against traditional IIoT frameworks for parameters like mining latency, node commits, signing latency, and energy dissipation. The improvements indicate the resilience of the scheme for practical IIoT setups.
科研通智能强力驱动
Strongly Powered by AbleSci AI