Real time non-invasive monitoring of glucose and nitrogen sources with a novel window sliding based algorithm

The development of fast and cost-effective methods for measuring biological molecules has many advantages over conventional methods. However, these methods, which are used for monitoring biological molecules, have some drawbacks, such as high cost, time consumption, or labor intensity. On the other hand, microwaves are interacted with sample which can be calculated easily. Thus, microwaves provide compact, uncomplicated, non-invasive, and continuous monitoring of various critical substances such as glucose and nitrogen sources. Here, we show that a new algorithm, based on a sliding window approach, which effectively identifies the optimum operating point using Vector Network Analyzer (VNA) measurements of biological macromolecules including glucose, ammonium sulfate, and yeast extract. Moreover, the effect of container type (mica glass and urine container) on microwave sensing, using a VNA connected with a WR-28 adapter, was investigated. The experimental results confirmed that, mica glass resulted in better differentiation than urine container for glucose estimation. Furthermore, glucose, ammonium sulfate, and yeast extract amounts were effectively determined with novel algorithm. Reflection coefficients ($S_{11}$) of glucose, yeast extract, and ammonium sulfate ranged between −14.14 dB, −14.41 dB, −10.65 dB, −10.85 dB, and −13.84 dB, −14.16 dB, respectively at optimal operation points when macromolecule concentrations were between 20–80 g/L. In addition, the time complexity of the proposed algorithm was performed and the overall time complexity is $O\left(n\right)$ (linear time) and the time complexity per incoming update is $O\left(1\right)$ (constant time). In this context, the algorithm is also suitable for online applications. The current study proposed a promising approach for cost effective and rapid estimation of biological substances.