The integration of blockchain-based federated learning (BFL) and Industry 4.0 utilizes intermediate models to execute task deployment and result acceptance, effectively solving the problems of data barriers and data resource waste in Industry 4.0. However, the BFL ecosystem is susceptible to poisoning and inference attacks that undermine data privacy and model security. In this paper, we propose PSFL, a robust FL framework that guarantees both data privacy and model security. Specifically, we design a cross-validation algorithm where numerous participants conduct a thorough assessment of the user’s contribution growth rate in the current round. This approach proves effective in identifying Byzantine attackers engaged in malicious activities within the system. Furthermore, we propose a lightweight multi-receiver signcryption mechanism employing secure key distribution, which significantly minimizes resource overhead. Finally, the security of PSFL is proved based on the random oracle model. Empirical assessment affirms the effectiveness and practicality of PSFL, even with different proportions of malicious users, PSFL’s performance is 10%20% higher than Trimmed Mean and M-Krum. In summary, PSFL improves the model accuracy and the security of the model transmission process in scenarios involving edge node poisoning, which demonstrates that PSFL can be well adapted to Industry 4.0.

Asynchronous distributed key generation (ADKG) is a strong-robustness key management technology to bootstrap threshold cryptosystems without a global clock, which can enable decentralized security management for threshold digital wallets in blockchain network. However, the high-computational cost of existing ADKG protocols makes it difficult to remove the “slow” connotation from the word “asynchronous” in a high-threshold security context. In this article, we propose a simpler “two-high” ADKG protocol (H2CT) for blockchain to improve the computational efficiency of asynchronous communications while balancing it with threshold security. Concretely, we first construct a computationally efficient asynchronous complete secret sharing (ACSS) scheme using number theory transformation, reducing the computational complexity of share evaluation from$O(n^{2})$to$O(n\log n)$in the dealing phase. To eliminate the negative impact of up to t biased secrets brought by the implicate messages in the agreement phase, we extend the verifiable ACSS scheme to a publicly verifiable ACSS scheme (pvACSS) using Feldman polynomial commitment. Leveraging this enhanced randomness, the matrix computation cost and message size in the remaining phases are reduced to about half that of the existing works, i.e.,$O(2n^{2})$. Finally, considering the high-scalability requirements, the network size n is increased to up to 1024 nodes and the results show that our pvACSS and the derived H2CT reduce the runtime by approximately 33.96 s and 824.46 s, respectively, over the state-of-the-art. Moreover, we perform simulations on an open-source blockchain testbed, fully demonstrating the efficiency advantages of our H2CT protocol.