Energy efficient transactions for blockchain networks using adaptive global best–worst particle swarm optimization

Blockchain technology offers decentralized and secure transaction processing but suffers from critical limitations in scalability, energy efficiency, and latency, hindering its adoption in real-time high-throughput applications. This study proposes a novel Adaptive Global Best–Worst Particle Swarm Optimization (AGBWPSO) algorithm integrated with dynamic sharding to address these challenges effectively. Unlike traditional GBWPSO, the proposed AGBWPSO employs a dual-extremum influence mechanism that combines both global best and worst positions, along with adaptive nonlinear parameter adjustment strategies for the inertia weight, cognitive, and social coefficients. This enhances exploration–exploitation balance, prevents premature convergence, and ensures efficient shard reallocation under dynamic transaction loads. The integration with dynamic sharding enables parallel transaction processing across optimally configured shards, significantly improving blockchain performance metrics. Extensive simulations conducted on Ethereum, Bitcoin, Hyperledger Fabric, financial, and IoT transaction datasets demonstrate that the proposed AGBWPSO achieves up to 5.88% improvement in transaction throughput (TPS), 14.3% reduction in latency, and 20% reduction in energy consumption per transaction compared to existing optimization methods. These results establish AGBWPSO as a robust and scalable solution for enhancing the operational efficiency and sustainability of blockchain networks in real-world applications.