Model-Driven Full Life-Cycle Support Reference Framework for Blockchain-Based Application Development

Blockchain technology has reshaped multiple industries by providing decentralization, immutability, and data transparency. These characteristics, however, make the development of blockchain-based applications demanding and costly. Model-Driven Development (MDD) offers a promising paradigm by abstracting low-level details into higher-level models, thereby enhancing efficiency and reducing errors. Yet, existing MDD-based solutions are fragmented and fail to support the software development lifecycle (SDLC) in a consistent manner. Moreover, the intrinsic complexity of blockchain and its heterogeneous infrastructures introduces further challenges across SDLC stages, underscoring the need for a unified framework that systematically integrates MDD with blockchain-specific development. This research proposes a MDD-based reference framework that supports a consistent modeling process across the entire SDLC of blockchain-based applications, from requirements to deployment. The framework adapts various modeling techniques to address blockchain-specific features such as decentralization, one-time deployment, and multi-party collaboration. The model transformation methods are improved to support the seamless full SDLC development process of blockchain-based systems. The proposed reference framework is platform-independent at modeling level, enabling compatibility with heterogeneous blockchain platforms. Its adaptability was validated through a prototype implementation supporting both Ethereum and Hyperledger Fabric. Further, a comparative case study was conducted on a cross-border food supply chain. The results demonstrate the feasibility and adaptability of the MDD-based reference framework, as well as improved efficiency for blockchain-based system development, which promotes a broader adoption of blockchain technology.