T-PRIME: Transformer-based Protocol Identification for Machine-learning at the Edge

Spectrum sharing allows different protocols of the same standard (e.g.,802.11 family) or different standards (e.g., LTE and DVB) to coexist inoverlapping frequency bands. As this paradigm continues to spread, wirelesssystems must also evolve to identify active transmitters and unauthorizedwaveforms in real time under intentional distortion of preambles, extremely lowsignal-to-noise ratios and challenging channel conditions. We overcomelimitations of correlation-based preamble matching methods in such conditionsthrough the design of T-PRIME: a Transformer-based machine learning approach.T-PRIME learns the structural design of transmitted frames through itsattention mechanism, looking at sequence patterns that go beyond the preamblealone. The paper makes three contributions: First, it compares Transformermodels and demonstrates their superiority over traditional methods andstate-of-the-art neural networks. Second, it rigorously analyzes T-PRIME'sreal-time feasibility on DeepWave's AIR-T platform. Third, it utilizes anextensive 66 GB dataset of over-the-air (OTA) WiFi transmissions for training,which is released along with the code for community use. Results reveal nearlyperfect (i.e. $>98\%$) classification accuracy under simulated scenarios,showing $100\%$ detection improvement over legacy methods in low SNR ranges,$97\%$ classification accuracy for OTA single-protocol transmissions and up to$75\%$ double-protocol classification accuracy in interference scenarios.