AN ADAPTIVE HYPERGRAPH-ENHANCED SEMANTIC GOSSIP PROTOCOL FOR EFFICIENT BLOCKCHAIN CONSENSUS

Abstract

Blockchain (BC) technology has gained much interest due to its inherent qualities of security, immutability and decentralization. Despite these benefits, standard BC networks continue to face a number of difficulties, including the vulnerability of conventional leader-based consensus procedures to targeted leader assaults, high block propagation delay, and duplicate network traffic. To overcome these shortcomings, this paper proposes a unique unified framework that combines a hyper-delegated-leader based byzantine fault tolerant (HD-LBFT) protocol with a semantic-aware gradient-driven gossip (SAGD-Gossip). HD-LBFT improves performance scalability by streamlining leader selection and communication through the use of Hypergraph structures. The SAGD-Gossip protocol replaces blind flooding with an intelligent on-demand mechanism that optimizes data transfer by routing pull requests among the effective channels in terms of semantic role, latency and criticality. This technique significantly lowers propagation delays and duplicate network broadcasts while maintaining data integrity among peers. Experimental evaluation indicates that the proposed model enhances BC performance in terms of transaction confirmation latency obtained at node 10. The proposed model achieved a lower latency of 0.25s, block propagation latency of 261.5712ms with a block size of 0.1 MB, and block reception latency attained 217.1126ms with a block size of 0.1 MB. The proposed techniques provide a scalable, high performance and secure decentralized BC network, making it suitable for use in other real-world applications such as distributed IoT systems, supply chain management and financial transactions.