RSHS: A Blockchain Consensus Mechanism for Edge Computing-Supported Agri-IoT Systems

The Agricultural Internet of Things (Agri-IoT) has emerged to boost food safety and agricultural efficiency, but its integration with centralized servers, even when supplemented by edge computing, presents risks. These include challenges in device security, data privacy, and scalability. In this paper, we integrate blockchain with Agri-IoT and introduce reputation-based secure HotStuff (RSHS), aiming to address these security and scalability issues. Firstly, RSHS calculates the comprehensive reputation of nodes by incorporating both consensus and execution reputations. This facilitates a classification threshold for nodes, enhancing the reliability of nodes in consensus. Secondly, RSHS integrates peer evaluation scores with reputation and a verifiable random function for primary node selection, thereby balancing security and unpredictability. Building on this approach, the phased selection strategy further reduces communication overhead by progressively eliminating low-reputation nodes. Finally, the simulation results reveal that with an increase in both the network size and the proportion of Byzantine nodes, RSHS, in comparison to HotStuff, diminishes the average delay by 68.89% and yields an average throughput that is 3.82 times higher. In terms of fault tolerance, RSHS improves the chain inclusion rate by 39.84% under fork attacks and sees a 17.85% increase under silence attacks. These findings demonstrate that RSHS consistently exhibits superior scalability and good fault tolerance.