Linear Bayesian Approaches for Low-Oversampled Stepwise Staggered SAR Data

Abstract

Staggered synthetic aperture radar (SAR) is an innovative concept to obtain an ultrawide continuous swath with fine azimuth resolution using multiple elevation beams and pulse repetition interval (PRI) variation. Conventionally, the interpolation of the nonuniform data with gaps (due to the interruption during transmission, i.e., blockage) requires a high oversampling ratio to avoid image quality degradation, which leads to increased range ambiguities and high downlink data rates. In recent years, there has been a growing interest in improving the image quality for low-oversampled staggered mode. This study suggests a new concept, stepwise staggered SAR, which employs stepwise staggered PRIs to facilitate the use of novel linear Bayesian approaches to achieve high-quality recovery of blockage in low-oversampled data. The so-called step in “stepwise” refers to a period that contains multiple constant PRIs in a PRI staggering cycle, ensuring part of the data uniformly sampled to reduce the error propagation during resampling. In addition to the no-consecutive gap condition, the stepwise PRI staggering strategy meets the requirement of no more than one gap at any location within the slant range of interest in each PRI staggering cycle, which makes it possible to easily extract prior information from the consecutive known samples on both sides of each blockage sample to realize the proposed linear Bayesian approaches. Because the prior information adapts to characteristics of backscatter, the linear Bayesian estimation has an excellent performance in blockage recovery, which has been validated by simulations with different scenarios.