UHF RFID Microwave Sensor Tag Design for an RSSI-Based Machine Learning Assisted Binary Ethanol–Water Mixture Characterization

This article proposes a novel microwavesensor-based UHF radio frequency identification (RFID) tagdesign for the dielectric parameter characterization of binaryethanol-water liquid samples with different concentrations.The sensor tag is designed on an FR-4 substrate with theImpinj M6 Dura chip. The fabricated prototype has an over-all physical size of 82.1x19.2 mm. The mixture samplesdropped into the middle part of the microwave sensor tagchange the RSSI value from the RFID system, the mainparameter to be associated with the water ratio and dielectricparameters in the binary mixture to be used in machine learn-ing (ML). Several ML regression methods have been used forethanol-water characterization. The GPR model obtained apromising result with R-2=0.98, RMSE=4.08, and MAE=0.85 for the water ratio in the mixture,R-2=0.98, RMSE=3.02,and MAE=0.66 for the dielectric constant real part, and R-2=0.99, RMSE=0.26, and MAE=0.05 for the dielectric constantimaginary part of the mixture, and the model yielded thehighest predictive performance among the four ML models.According to the obtained ML metrics, the water ratio in the binary mixture, the dielectric constant real and imaginaryparts of the mixture using the RSSI received over the RFID system, the distance, and the RFID frequency have beensuccessfully characterized. The proposed design has the technical potential to be used as characterization equipmentfor binary ethanol-water samples with low cost, high precision, and reusable features with ML assistance