Introduction to SAR and ISAR
Synthetic aperture radar (SAR) uses radio waves to illuminate a target area to receive, record, and process reflected signals. Resulting echoes are further processed to obtain a 2D image of the landscape.
Range resolution depends on the bandwidth of the signal while cross-range (azimuth) resolution depends on the length of radar antenna. Synthetic aperture means that synthetic antenna is created during the time of the target observation, which reduces the physical dimension of the radar antenna. Typically, an airplane, satellite, or drone carries such radar antenna. Inverse SAR (ISAR) relies on reflected waves from moving objects. Both SAR and ISAR technologies can make images of the targets in almost all weather conditions (for example, fog, rain, clouds, and snow) during the day and night.
The classical SAR and ISAR radars are active radars, which means they come equipped with transmitters to illuminate targets. Imaging using passive radar technology is a new trend in radiolocation. The passive radars do not transmit their own energy. They rely on the existing sources of energy to illuminate the potential targets. Passive radars can use commercially available transmitters such as FM radio, digital radio, television (digital audio broadcasting or digital video broadcasting-terrestrial), GSM, WiFi, or other radars as illuminators of opportunity. Due to the lack of their own transmitters, the passive radars allow for so called “silent operations.”
We used the NI USRP platform with an external frequency multiplier to design an active SAR system. We could reach close to 1 GHz bandwidth—40 MHz generated by USRP then multiplied by 24. Also, our radar worked in the band around 5.5 GHz, which meant we could use commercially available antennas (WiFi). Since we could not transmit at 5.5 GHz with the USRPs used, we mixed the signal up to 5.5 GHz band using electronic components from another supplier. The proposed active SAR radar demonstrator using NI USRP hardware is a frequency modulated continuous wave (FMCW) radar. The FMCW principles at the receiver side of mixing transmitted and received signals resulted in the beat signal, which generated a frequency corresponding to the target range. The further the object is, the higher the differential frequency is (and consequently, the higher the mixed frequency). We can perform Fourier transform on this mixed signal to determine the distance to the object.
When we mount a radar on an autonomous vehicle such as a drone, size and weight are of the foremost importance. We removed all chassis from the systems and kept only electronics, which helped us reduce the total system weight to below 5 kg and keep power consumption at around 70 W. For passive SAR/ISAR imaging, we used radar with HDD-8264 for data storage and PXIe-5663E for signal receiving. NI HDD-8264 is a RAID hard drive that can store large amounts of continuously streamed data. Since it uses RAID architecture, it can maintain high-throughput streaming. In particular, HDD-8264 allows for 600 MB/s continuous streaming, while newer solutions such as HDD-8266 (SSD version) can reach 3.6 GB/s. We did all signal processing for passive SAR/ISAR imaging offline.
Additionally, the hardware we used made multistatic operations (the passive radar receiver is dislocated and the receiver nodes are deployed in different geographical positions) possible. In such cases, we synchronized the recording PXI devices precisely using dedicated GPS PXI-6682 modules and PXIe-6674T timing and synchronization modules. We used this hardware so we could record the RF signals of interest truthfully from any place covered by GPS signal on the planet. We used LabVIEW to write record and playback software. Additionally, we rewrote our record and playback software so that we can use it for any application, not only radar applications.
Digital Signal Processing
One of the most challenging parts of developing any radar system is digital signal processing. In our applications we used real-time SAR processing for an active FMCW radar system and offline processing for passive SAR/ISAR imaging implemented with The MathWorks, Inc. MATLAB® software.
Business Results and Next Steps
Using COTS equipment greatly reduced our efforts to design complex radar systems. We could focus more on DSP rather than on designing RF I/O over and over again.
Regarding hardware needs for future applications, the higher the bandwidth of the generated signal, the higher the resolution. Therefore, we are currently looking for the solution in the higher bandwidth. In addition, we plan to move to the higher bands such as X band. Obviously, for UAV-type applications we always need more power efficient and smaller form factor solutions.
Dr. Piotr Samczyński
Warsaw University of Technology Institute of Electronic Systems
Warsaw 15/19, 00-665