Super-Resolution Spectral Approach for the Accuracy Enhancement of Biomedical Resonant Microwave Sensors

Super-resolution spectral methods are applied and compared to improve the estimation result provided by biomedical microwave resonant sensors. In particular, the resolution of resonant sensors is revealed to be significantly improved, despite their intrinsic low quality factor. Excellent robustness against noise is also demonstrated. Algorithms are first validated on ad hoc synthetic data mimicking the response of a resonant sensor. Additionally, experimental validation is carried out by using data coming from a microwave resonant sensor, which is specifically designed for blood-glucose monitoring.