Feasibility Study of Glucose Concentration Measurement of Aqueous Solution Using Time Domain Reflected Signals

Sensors (Basel). 2022 Feb 3;22(3):1174. doi: 10.3390/s22031174.


Recently, wideband microwave spectroscopy (WBMS) has been applied for material characterization. Blood glucose sensing through microwave spectroscopy is usually done with resonant frequency-domain methods. Time-domain (TD) WBMS is a low-cost and convenient technique that can be used for glucose sensing of the aqueous solution. In this paper, early research for the implementation of a TD dielectric spectroscopy setup for glucose concentration measurement is presented. TD reflected signals from water with different glucose content are calculated using inverse Laplace transform. The proposed setup is a quasi-monostatic setup in which measurements are done with two different devices in the frequency range of 0.1 to 6 GHz to make a comparison between frequency domain (FD) and TD methods. Frequency domain (FD) measurement is performed with VNA and two Vivaldi antennas. Then, TD data is obtained using the transforming option of VNA. Direct TD measurement is operated with a maximum length sequence (m-sequence) transceiver. Measurement and numerical results follow the same trend and show good agreement with each other. A monotonic relation between peaks of TD signals and the corresponding glucose concentration is achieved. The variation of the height of the reflected signal's peak is 0.00002 and 0.0005 for each 50 mg/dL glucose concentration with FD measurements and direct TD measurements, respectively. The glucose concentration range of 25 mg/dL to 400 mg/dL is investigated, and the worst repeatability of this method is 3.65% for 300 mg/dL.

Keywords: dispersive medium analysis; glucose concentration measurement; microwave dielectric spectroscopy; time-domain dielectric spectroscopy; ultra-wideband dielectric spectroscopy.

MeSH terms

  • Dielectric Spectroscopy*
  • Feasibility Studies
  • Glucose
  • Microwaves*
  • Water


  • Water
  • Glucose