Secure and Privacy-Preserving Traffic Flow Prediction in Intelligent Transportation Systems Using Blockchain-Based Federated Learning

Accurate traffic flow prediction is critical for intelligent transportation systems (ITS) but raises privacy concerns in centralized models. We propose a blockchain-based federated learning (BFL) framework for secure, privacy-preserving traffic flow prediction. Using Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models, our approach captures spatiotemporal traffic patterns while using IoT devices to keep raw traffic data local. A consortium blockchain secures model updates, ensuring tamperproof storage and transparency. Evaluated on the Caltrans PeMS dataset, our framework achieves 98.75% accuracy, outperforming centralized models by 12%. Key innovations include a hybrid LSTM-GRU architecture, blockchain-enabled auditing, and differential privacy, reducing prediction error by 18% and detecting poisoning attacks with 99% precision. This scalable, secure solution enables real-time prediction, advancing privacy-aware smart mobility in ITS.