Soil Moisture Sensing through Phase Measurements in Analog Backscatter Systems

Abstract

Conventional low-cost soil moisture sensors consume milliwatts of power, limiting their scalability for battery free agricultural IoT. To bridge this gap, we propose a narrow band phase-difference approach for soil moisture estimation using ultra-low-power analog backscatter at 868MHz. We develop a geometric electromagnetic model relating spatial phase shifts to soil permittivity, proving differential phase remains a valid observable despite refraction across the layered soil–air interface. We further characterize non-ideal hardware effects, establishing how antenna impedance mismatches and other impairments introduce systematic phase biases that require calibration. Our experiments demonstrate reliable discrimination between dry and wet soils, validating a battery-free sensing architecture that offers a scalable solution for sustainable water management.