Adaptive Range-Doppler Distributed Target Detection Under Structured Interference in Non-Gaussian Clutter: A Sparse Recovery Perspective

With the increase of urgent demand for the emergent distributed target (e.g., collaborative drone swarms, ships, and car groups) detection in complicated environments, traditional adaptive detection methods face many challenges: 1) The distributed targets often appear as a set of multiple targets’ scattering centers (SCs) in range-Doppler domain with unknown Doppler frequencies, dispersing the target energy and making it more difficult to detect effectively. 2) The electromagnetic interference and non-Gaussian clutter raise the false alarm of radar and/or reduce the detection probability of the targets. In this article, we consider range-Doppler distributed target detection under structured interference and non-Gaussian clutter. To overcome the above challenges, we first assume that the targets’ SCs are sparsely distributed in the Doppler domain. Then, a sparse recovery method is utilized to extract the target component from the echo data. To mitigate the influence of interference, two strategies are adopted to design detectors: detecting while canceling interference strategy and interference canceling before detecting strategy. Moreover, the clutter is modeled as a spherically invariant random vector with an unknown texture component and an unknown covariance matrix (CM). A set of training data is adopted to estimate the CM. The computational complexity of the detectors is analyzed theoretically. Numerical experiments based on the simulated and measured data show that the proposed detectors outperform the existing ones.