ISAR imaging with random missing observations based on non-iterative signal reconstruction algorithm

Random missing observations in real-world inverse synthetic aperture radar (ISAR) imaging may appear due to the instability of the radar system. Under this circumstance, the signal of one range bin over the slow-time has limited samples. Using traditional range-Doppler algorithm, high-quality ISAR images for random missing observations cannot be obtained. Recently, a new non-iterative algorithm based on the combined robust statistics and compressive sensing (CS) theory has been proposed to efficiently recover a complete signal from a small random set of samples, showing robustness in the presence of noise [16]. It is also important to emphasize that non-iterative method is computationally simpler than the iterative signal reconstruction solutions, and thus more amenable to practical applications. Therefore, based on the non-iterative robust signal reconstruction method, a new algorithm for ISAR imaging with random missing observations is proposed in this paper. The efficiency of the proposed approach is demonstrated on the examples with simulated and real data.