DMM-GPT: A Dual Modeling Mechanism-Based Generic Protocol Translation Framework

With the increasing complexity of heterogeneous communication environments, the problem of data sharing caused by multiple heterogeneous protocols has become the core challenge restricting interoperability. Aiming at the problems of poor scalability and automation ability of existing protocol translation methods, this study proposes a Dual Model Mechanism-based Generic Protocol Translation (DMM-GPT) framework. The framework creatively combines ontology-based semantic modeling with state-machine-based behavioral modeling, enabling automated identification of protocol translation bridging points and autonomous construction of a unified protocol state transition model. Firstly, a protocol ontology model is constructed to model the semantic associations between protocol entities. Based on the ontology, semantic information is extracted. Semantic similarities from message level to field level are computed, thereby sequentially identifying message mapping and field mapping relationships. Secondly, modeling protocol behavior through state machine models. Based on the identified mapping relationships, the heterogeneous protocol bridge points are constructed, and a unified state transition model is synthesized automatically. Thus, interoperable behavior between heterogeneous protocols is implemented. Finally, the complete closed loop of protocol interoperability is implemented by applying semantic alignment results to the state transition model. To validate the effectiveness of DMM-GPT, experiments were conducted in two representative scenarios: service discovery (SLP-Bonjour) and IoT communication (HTTP-CoAP). The results demonstrate its effectiveness in enabling interoperability between heterogeneous protocols.