Spatiotemporal Blockchain Networks for AI-driven IoT-Enabled Multi-Layer Urban Energy Optimization in Smart Cities

Effective energy management is essential for the sustainability of rapidly growing smart cities. Traditional centralized systems, which often rely on Artificial Neural Networks (ANNs), struggle with scalability and vulnerability to cybersecurity threats for IoT-enabled networks. To remedy these concerns, this study introduced a Spatiotemporal Blockchain Network (SBN) and a blockchain-integrated Transformer-based Energy Prediction Model (TEPM). The SBN provides transparency, immutability and added protection for sensitive energy transactions and the TEPM utilizes sophisticated attention mechanisms to accurately predict energy using spatiotemporally dynamic characteristics. Overall, the future incorporation of the SBN and TEPM would allow for real-time energy optimization and corresponding reduction in wasted energy and operational efficiency maximization. The framework was implemented and validated as a case study in MATLAB/Simulink with Hyperledger Fabric, achieving a 22% improvement in energy efficiency, a forecasting error of 3.2% (MAPE), a throughput rate of 450 transactions per second and supported a scale of over 15,000 IoT nodes. These results indicate the system’s ability to revolutionize urban energy management by improving the capacities for secure, scalable and adaptive energy distribution, and therefore support sustainable and resilient smart city infrastructures.