Statistical sparsity-based learning for ultra-wideband radar signal reconstruction

Mirbeik, Amir; Lawrence, Victor B.; LaPeruta, Thomas A.; Popescu, Cristian M.; Hohil, Myron E.; Tavassolian, Negar

doi:10.1117/12.2587067

Cited by 2 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The assumption of independent and identically distributed (IID) total variation regularised estimates sets a challenge for choosing the measurement point configuration and phase discrepancy appropriately. We covered two possible strategies for spatial point selection, otherwise known as statistical sparsity-based learning with a suitable prior model [41]. The sample size of the present randomised configuration was found to reduce the phase-dependent fluctuations significantly.…”

Section: Discussionmentioning

confidence: 99%

Investigation of Phase and Wavelength-induced Errors in Full-Wave Radar Tomography of High Contrast Domain: an Application to Small Solar System Bodies

Oluwatoki¹,

Astrid²,

Sorsa³

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper aims to reconstruct the internal structure of a two-dimensional test object via numerically simulated full-wave time domain radar tomography with the presence of carrier wave induced (carrier-induced) uncertainty, following from a complex domain structure, and short wavelength of the signal as compared to the target object diameter. In particular, we consider an application in planetary scientific studies of reconstructing the interior structure of an arbitrary high contrast small Solar System Body (SSSB), i.e., an asteroid, with a probing signal of small wavelength following from the instrument and mission payload limitations. Our uncertainty reduction model is devised from the statistical viewpoint by finding the reconstruction via multiple datasets assuming that the carrier-induced deviations in the reconstruction correspond to random deviations, which are independent and identically distributed (IID) for each dataset. A spatial and frequency-based error marginalisation is implemented, thus inspecting the relationship between the signal baseband frequency and the phase discrepancy of the modelling accuracy, determined by the signal and domain parameters. The numerical experiments are performed for 20 and 60 MHz center frequencies proposed for CubeSat-based radars. Of these, the latter matches the center frequency of the Juventas Radar which will be aboard HERA mission to investigate the interior structure of asteroid Dimorphos, the asteroid moon of the binary system 65803 Didymos.

show abstract

Section: Discussionmentioning

confidence: 99%