This paper presents the design of a high-voltage pulse-based radar and a supervised data processing method for soil moisture estimation. The goal of this research was to design a pulse-based radar to detect changes in soil moisture using a cross-hole approach. The pulse-based radar with three transmitting antennas was placed into a 12 m deep hole, and a receiver with three receive antennas was placed into a different hole separated by 100 m from the transmitter. The pulse generator was based on a Marx generator with an LC filter, and for the receiver, the high-frequency data acquisition card was used, which can acquire signals using 3 Gigabytes per second. Used borehole antennas were designed to operate in the wide frequency band to ensure signal propagation through the soil. A deep regression convolutional network is proposed in this paper to estimate volumetric soil moisture using time-sampled signals. A regression convolutional network is extended to three dimensions to model changes in wave propagation between the transmitted and received signals. The training dataset was acquired during the period of 73 days of acquisition between two boreholes separated by 100 m. The soil moisture measurements were acquired at three points 25 m apart to provide ground truth data. Additionally, water was poured into several specially prepared boreholes between transmitter and receiver antennas to acquire additional dataset for training, validation, and testing of convolutional neural networks. Experimental results showed that the proposed system is able to detect changes in the volumetric soil moisture using Tx and Rx antennas.

This paper proposes an improved design of a pulse-based radar. An improved design of a pulse generator is presented using step recovery diodes and a signal mixer for the received signal. Two-step recovery diodes produce pulses of 120 ps in duration. A pulse generator is improved by removing the negative power supply, resulting in a reduced number of electronic pulses. A sampling mixer at the receiver’s site receives the generated signal and stretches it from picoseconds into microseconds. The improved pulse generator is also used in the sampling mixer as a strobe pulse generator, which makes the sampling mixer much simpler. The stretched signal is then sampled by a low sample rate using an analog to digital converter. The proposed radar design achieves up to 8 GHz bandwidth and an equivalent receiving sample rate of about 100 GSa/s. The radar is controlled using a software-defined radio called Red Pitaya, which is also used for data acquisition. The proposed radar design uses widely available commercial components, which makes radar design widely available with low cost implementation.