Development of Ka-band Frequency-Modulated Continuous Wave (FMCW) Synthetic Aperture Radar (SAR) On Octorotor Unmanned Aerial Vehicle (UAV)
Prof Zhi Ning Chen
Synthetic aperture radar (SAR) imaging utilizing frequency-modulated continuous wave (FMCW) signals has gained attention in the past few years. Primarily, this is due to the availability of microwave integrated circuits enabling the design of FMCW transceivers with a compact form factor. Moreover, the use of FMCW enables the transmission of wide bandwidth signals while keeping the receiver’s baseband bandwidth narrow. This makes high resolution imaging possible without the high sampling requirements at the receiver.
This talk introduces the development of a high-resolution Ka-band FMCW SAR system developed for a low-flying octorotor unmanned aerial vehicle (UAV) platform by a team from National University of Singapore with its partner. The development focuses on two main aspects: the generation of high-bandwidth FMCW signals and the application of an image reconstruction algorithm considering the motion instability of octorotor platforms.
Two signal generation techniques were employed. The first approach uses the concept of synthetic bandwidth where a narrow band 500-MHz chirp signal is transmitted at 4 adjacent frequencies to obtain time- and frequency- continuous wideband signal with a bandwidth of 2 GHz. Frequency multiplication and upconversion were done to generate a 4-GHz chirp at Ka-band. The second is a more straightforward approach that uses a phase-locked loop (PLL) at X-band to generate a 2-GHz chirp. The frequency output is then multiplied by a factor of 3 to generate a 6-GHz bandwidth at Ka-band. A rail SAR platform was set-up and successfully verified both the designed range resolution of ~3.2-5 cm and angular resolution of 4 cm.
The presentation will also discuss about the octorotor flight trials, wherein a combination of prominent targets, and fused inertial navigation system (INS) and global positioning (GPS) sensor data were used for motion compensation. Uncompensated motion errors are the major cause of the degradation of a reconstructed SAR image on an octorotor platform. It was shown that the data from the on-board low-cost GPS and INS sensors is sufficient for motion correction. The back-projection algorithm is then applied to perform the image reconstruction. To further improve the image quality, phase-gradient autofocus was used. Lastly, we present the successful imaging results achieving azimuth resolution within 4% of the expected value, showing the effectiveness of the motion compensation and the image reconstruction processing. This work demonstrates the highest resolution among the FMCW octorotor SAR systems reported in the literature.