Karol Abratkiewicz - Gdansk University of Technology, Department of Microwave and Antenna Engineering
Adam Popik - Gdansk University of Technology, Department of Microwave and Antenna Engineering
Michał Lewandowski - Gdansk University of Technology, Department of Microwave and Antenna Engineering
Jakub Czerniakowski - Gdansk University of Technology, Department of Microwave and Antenna Engineering
Krzysztof Nyka - Gdansk University of Technology, Department of Microwave and Antenna Engineering
We developed the synthetic aperture radar (SAR) demonstrator within a group project module at the Faculty of Electronics, Telecommunications, and Informatics in the Department of Microwave and Antenna Engineering and the WiComm Center of Excellence. The department conducts both theoretical (analysis, synthesis, simulation, and computer modeling) and practical research in high frequencies ranging from hundreds of megahertz to tens of gigahertz. This covers passive and active microwave and RF circuits, antennas, and subsystems for modern wireless communication and radiolocation.
The goal of our student project was to create a SAR demonstrator for S band to take radar measurements and perform signal processing to create images of surfaces irradiated by the radar’s antenna. We added some functionalities to measure distance and velocity so the device could function as a demonstrator for teaching students. The radar’s open and flexible architecture results in a platform for implementation of additional third-party signal processing algorithms.
SAR emits electromagnetic waves and analyses the echo reflected from irradiated scenes while moving over an investigated surface; thus, creating an image of that surface. In contrast to conventional radar, SAR requires constant, and preferably uniform, movement over observed terrain. With such an antenna motion, we can artificially (through proper signal processing) narrow the antenna beam width in the direction of the movement. This enhances the resolution of the obtained radar images. Typically, SAR is mounted on a plane or spacecraft, making it possible to observe a part of Earth’s surface under investigation. The images produced are practically independent of weather conditions or lighting, unlike optical imaging techniques.
The Radar’s Architecture
The radar system consists of an analog high-frequency (microwave) front-end, local phase-locked loop (PLL) oscillator, intermediate frequency amplifier, power supply block, A/D converter, and DSP software on a PC. Figure 1 shows the radar’s architecture.
The radar operates in continuous wave mode with frequency modulated continuous wave (FMCW) of a triangular (dual ramp) waveform, which is generated by a voltage controlled oscillator (VCO) stabilized in PLL. We can program the entire device from a PC application, which is both portable and scalable. The modulating waveform, produced in the PLL, is applied to VCO, which generates the FM signal in the 2.4 GHz–2.5 GHz band with configurable frequency sweeping time. The user defines the bandwidth and period of the modulating waveform. The generated high-frequency signal is amplified in the power amplifier and then directed to the power divider. One half of the signal directly feeds the transmitting antenna, while the second half pumps the mixer. The receiving antenna delivers the unknown signal to the input of the low-noise amplifier and output is connected to the RF input of the mixer. By mixing the transmitted and received signals, the resulting low-frequency signal reflects all frequency shifts related to the position and motion of objects in the irradiated scene. This low-frequency signal is amplified in the circuit with automatic gain control and applied to the input of the A/D converter. We used the USB-6211 A/D converter with ST-200 extension board. The sampling speed of the converter is 250 kS/s, which combined with a sufficiently high ramp period, is enough for a short-range radar. The A/D converter features a high resolution (16 bits) and can work with symmetrical supply voltages up to ±10 V.
We used two NI products, a USB-6211 converter and LabVIEW software for signal acquisition and processing. We used high-level implementation of reading data from the converter and the user-friendly environment of LabVIEW to ease the development of signal processing, which significantly shortened the time required to realize the project. The algorithms for processing SAR signals and the procedures for calculating distance and velocity are based on built-in, optimized functions available in LabVIEW libraries. Using these, and additional support from NI engineers, we successfully implemented the range Doppler algorithm. We performed most operations in the frequency domain involving one- and two-dimensional FFT, IFFT, and filtration. We sampled the signals and saved them in a two-dimensional array in which rows correspond to the samples in the time domain, and columns correspond to the position on the SAR. We saved the input data in TDMS files, which we then processed in order for imaging. Since LabVIEW software simplifies development of multi-threaded applications, we could develop procedures for real-time calculations of velocity and distance.
We enclosed the completed radar in a case to protect its electronic circuits from mechanical stresses and ensure proper airflow. Figure 2 shows the assembled device with two cylindrical antennas. Figure 3 shows a screenshot of the controlling program.
We tested the radar in field conditions, on the grounds of the Faculty of Electronics, Telecommunications, and Informatics. We wanted to obtain SAR imaging of a radar reflector in an environment free of obstacles. The signal we used for measurement was FMCW with 90 MHz deviation and 10 ms waveform period. Figure 4 shows the area where measurements were performed, the location of the reflector, and the results. The reflector’s radar cross-section for 2.45 GHz center frequency is 1.5 m2. The echo coming from the radar reflector is visualized on the distance of 20 m
We also tested the system on the Energa Arena Stadium, where we placed three reflectors in a free propagation environment. Figure 5 shows the test set and Figure 6 shows the results.
The obtained imaging shows the reflectors in the 14.5, 16, and 19 meter range. In addition, we performed velocity measurements with the radar in Doppler mode and a moving car as a test object. Figure 7 shows those results.
The result of our project is a complete SAR demonstrator for academic purposes. It uses a fast NI USB-6211 converter and LabVIEW software for the implementation of fast signal processing algorithms. The project shows how we can perform complicated signal processing calculations efficiently when we choose a software platform that delivers well-supported tools to help with data acquisition and operations on large sets of data. Further development of the project can focus on optimization of the algorithms and implementation of real-time processing.
Gdansk University of Technology, Department of Microwave and Antenna Engineering