Blockchain and Dew Computing-based Model for Secure Energy Trading in Smart Grids

The Peer-to-Peer (P2P) energy trading paradigm allows producers and consumers to buy or sell energy from each other directly. An increasing energy demand requires a robust P2P supply chain driven by digitization and decen-tralization. Blockchain technology offers security and financial transaction benefits for P2P energy trading. However, existing energy trading systems are struggling to preserve privacy along with fair economic profit to the prosumers and consumers. This paper proposes a secure and profitable P2P energy trading mechanism using federated learning, dew computing, and blockchain. Blockchain-enabled Dew Servers (BDS) are employed in microgrids to store the user's data and local management software. A cloud server aggregates local models of individual microgrid models to construct the global federated learning model. The proposed Blockchain-enabled Dew Computing and Federated Learning (BDC-FL) model ensures data privacy for prosumers/consumers and attempts for profit maximization. The proposed system predicts energy production and intelligently flips between energy sharing and trading with the prosumer's consent. The proposed model has been tested using benchmark datasets and demonstrates superior performance compared to state-of-the-art energy trading and sharing mechanisms. Experimental results indicate that our BDC-FL model can increase profits up to 18% for stable participants in the smart microgrid ecosystem.