Design and Analysis of an RCSRR-based Microwave Sensor for Dielectric Characterization of Liquid Analytes

This paper presents the design and empirical modeling of a high-sensitivity microwave sensor based on a Rectangular Complementary Split Ring Resonator (RCSRR) integrated into a microstrip structure for dielectric sensing applications. Operating at a resonant frequency of 5.11 GHz, the sensor is capable of detecting subtle variations in the dielectric constants of liquid analytes such as ethanol, glycerin, benzene, NaCl, and carbon disulfide. The proposed design achieves a sensitivity of 35 MHz per dielectric unit (DU), making it suitable for precise material characterization. To enhance analytical accuracy, multiple regression models—linear and polynomial (degrees 2 and 3) were applied to predict the dielectric constant from frequency shifts. Among these, the cubic polynomial regression model demonstrated superior performance with a coefficient of determination (R²) of 0.998 and a root mean squared error (RMSE) of 0.00064, significantly outperforming linear and quadratic models. This integration of metamaterial-based microwave design and machine learning regression provides a compact, non-invasive, and accurate sensing solution for real-time applications in biomedical diagnostics, industrial quality control, and environmental monitoring.