Quantification of Honey Adulteration Using a Planar Microstrip Metamaterial-Motivated Sensor and Software-Defined Radio

This work presents the application of a planar microstrip metamaterial (MM)-motivated sensor in the form of a split-ring resonator (SRR) for the detection and quantification of water adulteration in honey samples. The proposed sensor utilizes inexpensive software-defined radio (SDR) measurements, offering a cost-effective solution. Unlike previous MM sensors employing SRR configuration, our proposed MM-motivated sensor is integrated with a microstrip feedline and an additional vertical bar, enhancing its sensing capability. To enable accurate measurements, a simple calibration procedure based on baseline normalization is implemented, allowing for amplitude-only transmission measurements ( $|{S}_{{21}}|$ ) using the SDR. A special sample holder was designed to increase the repeatability of measurements. We separately employ two postprocessing techniques, namely, the rolling average (RA) approach and the Savitzky-Golay (SG) filter, to effectively eliminate ripples in the measured $|{S}_{{21}}|$ data obtained from the SDR. For validation and evaluation of the proposed sensor and postprocessing techniques, measurements of flower honey samples with water adulteration levels ( $\delta $ ) up to 10% (mass-to-mass basis) were also performed using a vector network analyzer (VNA). The resonance frequency of $|{S}_{{21}}|$ is utilized as the basis for analysis. We establish a metric function (linear function) that correlates the shift in resonance frequency with $\delta $ . It is observed that this function fits to measured resonance frequency values with an ${R}^{{2}}$ value greater than 0.96 for both postprocessing techniques with or without the sample holder. By inverting this function, we can predict $\delta $ with considerable accuracy for a given resonance frequency by using another linear function. It is observed that percentage variations between predicted and measured $\delta $ values are between ${1.54}\%$ and ${6.46}\%$ for testing samples with $\delta = {5}\%$ and $\delta = {7}\%$ .