The space-air-ground-sea three-dimensional (3D) network is a comprehensive communication network system. This 3D network combines extensive coverage of satellite communications, adaptability of unmanned aerial vehicle (UAV) communications, reliability of terrestrial communications, and the necessity for maritime communications. These networks generate enormous amounts of data, and training machine learning (ML) models on this data will have a significant impact on industry. At the same time, the availability of such data poses numerous security threats, which can be overcome using Federated Learning (FL). Decentralized training in FL can provide a universal model from local data generated by 3D networks. However, most existing FL frameworks have a centralized server, which questions credibility, single-point failure, and global confidence. To solve these problems, industrial blockchain technology has received a lot of attention by replacing centralized servers in traditional FL, which offers a promising approach to address key issues like data privacy and security. In a blockchain-based system, digital signatures are a core component for ensuring data integrity and system security, however, private key disclosure can pose significant risks. Security can be enhanced by using threshold signatures which provide a more reliable foundation for FL by storing keys in multiple nodes and requiring multiple nodes to collaborate to generate signatures. In this paper, we propose a Threshold signing scheme for ISO/IEC Digital Signature Standards (TDSS) in an industrial blockchain. The TDSS scheme helps FL to achieve truly distributed decentralization for unified space-air-ground-sea model training. The TDSS scheme exploits the SM-2 digital signature algorithm in the ISO/IEC standard when t out of n nodes in an industrial blockchain interact with each other to calculate the signature. The experimental results and analysis show that the TDSS scheme has provable security and efficient against security attacks, which can be applied to large-scale threshold signing scenarios.