Development of DroneSAR
My primiary Ph.D. project is developing the DroneSAR system. Funded by NSF IoT4Ag, Digital Forestry and Indiana Corn Marketing Concil. The system aims to become the world's first Drone based SAR system to measure soil moisture and biomass with high resolution for agricultural applications.
As the leading engineer of this project, my work include hardware design, digital signal processing algorithm development and science algorithm development. My work also extent to managing a team of 6 engineers for field campaign, data processing and hardware/software development.
The DroneSAR System
DroneSAR is a laboratory-developed, fully polarimetric synthetic aperture radar (SAR) system designed for deployment on small unmanned aerial vehicles. With a total system weight of less than 2.5 kg, DroneSAR can be integrated onto lightweight drone platforms without compromising flight performance. The radar operates in the S-band using open frequency allocations, eliminating regulatory constraints associated with licensed spectrum use. To ensure high-quality measurements, digital filtering techniques are applied to mitigate interference from nearby communication signals and other sources of electromagnetic noise.
The prototype of DroneSAR was initially developed in Cape Town University, South Africa by Dr. Michael Ingg's group also as known as the MiloSAR. It was brought to Purdue in 2022, and refined it's design. Featuring 2 transmission channels and 2 receving channals in FMCW sawtooth waveform with a bandwidth of 175MHz and a pulse repetition frequency of 1250. The system is powered by a Red Pitaya FPGA board enabling high speed reading and writing to a SD card for further processing.
System Integration
The radar is integrated with a ALTA-X Drone System. To avoid RFI with the remote controller, a 900 MHz radio is used between the drone and the controller. A SBUS PWM channel signal is used to "Trigger" the radar for start and stop recording. The Radar directly draw power from the UAV with a internal voltage regulator to ensure the consistant power. To ensure the repeatability and the accuracy of the radar measurement, a high accuracy Trimble APX GNSS system is integrated with the radar system. A motion conpensation algorithm is developed to remove the random motion caused by the Drone Platform.
Image Generated By the SAR System
System Calibration
Calibration algorithm is developed to perform radiometric and polarimetric calibration on the radar system. 8 corner refelectors are depolyed at a airport runway for radiometric calibration. A forested area is used for polarimetric calibration. The result demonstred a radiometric accuracy of < 1 dB and channel cross talk <-10dB.
Field Campaign
The DroneSAR field campaign was conducted to evaluate system performance and collect high-quality radar measurements under controlled yet realistic conditions. Flights were carried out over agricultural test fields with well-characterized surface properties, allowing repeated observations under varying soil moisture and surface roughness conditions. The UAV followed predefined flight trajectories to ensure consistent spatial sampling and precise geolocation. Ground truth data, including in situ soil moisture measurements and surface characterization, were collected concurrently to support calibration and validation of the radar observations. This coordinated field campaign enabled a comprehensive assessment of system stability, data quality, and retrieval performance in real-world operating conditions.
Soil Moisture Retrieval Algorithm
