Rich by Coin

Blockchain-based, privacy-preserving information sharing is increasingly critical for modern supply chains. This study proposes a method that fuses a blockchain consensus algorithm with federated learning to secure cross-organizational data exchange. It designs a blockchain-based verifiable consensus mechanism and couples it with federated learning to build an encryption model that protects data while enabling collaboration. To mitigate privacy leakage inherent in federated learning, the approach introduces causal pseudo-random functions and cuckoo hashing to process data and reduce communication overhead, preventing hash conflicts.

The encryption model is then deployed in the data transmission system, complemented by multi-factor authentication to bolster security. In experiments, the consensus algorithm exhibited an average node-election latency of 115.20 ms and 8.56 ms for node replacement, while the method achieved 92.48% accuracy in data generation after processing and 98.87% data-tampering detection, with privacy breaches at 0.1% and an average response time of 1.0 s. Overall, the framework balances privacy protection with efficient sharing, lowering system overhead and delays through a tighter integration of the consensus mechanism and federated learning, and thereby enhancing the efficiency of multi-party information sharing in complex supply networks. By optimizing interactions between the consensus process and federated learning, the approach improves operational efficiency while maintaining verifiable, auditable data sharing.