ReFormer: Generating Radio Fakes from the Learned Channel Prior

We introduce ReFormer, a generative AI (GAI) model designed to efficiently learn wireless channel prior models and generate synthetic radio-frequency (RF) data. These RF fakes closely mimic the statistical properties of the original training data or can be produced with altered statistics to augment datasets gathered from real-world experiments. For such applications, adaptability and scalability are important issues. This is why ReFormer leverages transformer-based autoregressive generation, trained on learned discrete representations of RF signals. Moreover, this architecture allows the model to generate data under specific constraints or conditions through the use of prompts, which is particularly beneficial for tasks like channel estimation and modeling. It may also leverage the data from a source system to generate training data for a target system. We show how different transformer architectures and other design choices affect the quality of generated RF fakes, evaluated using metrics such as precision and recall, classification accuracy and signal constellation diagrams.