APPLICATIONS OF COMPRESSED SENSING TO COHERENT RADAR IMAGING
Open Access
- Author:
- Zhu, Qian
- Graduate Program:
- Electrical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 16, 2016
- Committee Members:
- John D. Mathews, Dissertation Advisor/Co-Advisor
John D. Mathews, Committee Chair/Co-Chair
Vishal Monga, Committee Member
Julio Urbina, Committee Member
Necdet S Aybat, Outside Member - Keywords:
- Compressed Sensing
Coherent Radar Imaging
Meteoroids Fragmentation
Low-altitude Meteors - Abstract:
- Although meteoroids fragmentation has been observed and studied in the optical meteor community since the 1950s, no definitive fragmentation mechanisms for the relatively small meteoroids (mass ≲10−4 kg) have been proposed. This is in part due to the lack of observations to constrain physical mechanisms of the fragmentation process. While it is challenging to record fragmentation in faint optical meteors, observing meteors using HPLA (High-Power, Large-Aperture) radars can yield considerable information especially when employing coherent radar imaging (CRI). CRI can potentially resolve the fragmentation process in three spatial dimensions by monitoring the evolution of the plasma in the meteor head-echo, flare-echo, and trail-echo regions. On the other hand, the emerging field of compressed sensing (CS) provides a novel paradigm for signal acquisition and processing. Furthermore, it has been, and continues to be, applied with great success in radar systems, offering various benefits such as better resolution compared to traditional techniques, reduced resource requirements, and so forth. In this dissertation, we examine how CS can be incorporated to improve the performance of CRI using HPLA radars. We propose a single CS-based formalism that enables the three-dimensions (3D)—the range, Doppler frequency, and cross range (represented by the direction cosines) domain—coherent imaging. We show that the CS-based CRI can not only reduce the system costs and decrease the needed number of baselines by spatial sparse sampling, which can be much less than the number required by Nyquist-Shannon sampling criterion, but also achieve high resolution for target detection. We implement the CS-based CRI for meteor studies with observations conducted at the Jicamarca Radio Observatory (JRO) in Peru. We present the unprecedented resolved details of meteoroids fragmentation, including both along and transverse to the trajectory spreading of the developing plasma, apparently caused by gross fragmentation and plasma diffusion parallel to the geomagnetic field. Moreover, we provide a statistical survey on the properties of low-altitude (≲~90 km) sporadic meteors based on three observation sets conducted in 15,16 April 2010 and 4 & 5 August 2014 at the JRO, respectively. We demonstrate the vast majority (~80%) of these low-altitude meteors exhibit low speed (|V|≲~25 km/s) and a large meteoroid entry angle (across the beam), suggesting that meteoroids fragmentation accompanied by particle flaring plays a significant role in the generation of these deep-in atmospheric echoes in addition to the classical ablation. Meanwhile, it also infers that the aerodynamic loading is inadequate to explain the fragmentation of those meteoroids. Additionally, the time distribution of these echoes shows an occurrence peak near 0300 (Local Time), which is earlier than the usual sunrise peak corresponding to the near-zenith transit of the apex-of-Earth’s way.