In precision agriculture, information technology is used to improve farm management practices. Thereby, productivity can be increased and challenges with overfertilization and water consumption can be addressed. This requires low-power and wireless underground sensor nodes for monitoring the physical, chemical and biological soil parameters at the position of the plant roots. Three ESP32-based nodes with these capabilities have been designed to measure soil moisture and temperature. A system has been developed to collect the measurement data from the sensor nodes with a drone and forward the data to a ground station, using the LoRa transmission standard. In the investigations of the deployed system, an increase in the communication range between the sensor node and the ground station, from 300 m to 1000 m by using a drone, was demonstrated. Further, the decrease in the signal strength with the increasing sensor node depth and flight height of the drone was characterized. The maximum readout distance of 550 m between the sensor node and drone was determined. From this, it was estimated that the system enables the readout of the sensor nodes distributed over an area of 470 hectares. Additionally, analysis showed that the antenna orientation at the sensor node and the drone influenced the signal strength distribution around the node due to the antenna radiation pattern. The reproducibility of the LoRa signal strength measurements was demonstrated to support the validity of the results presented. It is concluded that the system design is suitable for collecting the data of distributed sensor nodes in agriculture.

The composition of materials generally has an influence on its dielectric properties. Hence it may be possible to determine their composition using microwave dielectric spectra. However with complex materials like foodstuffs it is difficult to devise a suitable dielectric model. Therefore multivariate calibration methods are discussed here. With these calibrations it is also possible to process the measured scattering parameters directly. In this case it is not necessary to calculate the permittivity and more degrees of freedom are available for the design of a sensor. The direct processing of the S-parameters is demonstrated with a new transmissionline sensor.

An open resonator for mm-waves is presented in this paper. It is used for the measurement of the complex permittivity of materials. For the fabrication there must be compromises because the structures are relatively small. During the development of the resonator the limitations of the manufacturing are taken into account. After simulations and a calibration with artificial neural networks the resonator is built and first measurements will be carried out.