Low-Complexity PRF-Agile Radar Maneuvering Target Detection via ANS-ATRT-VEM

Abstract

This paper presents a robust and low-complexity method for maneuvering target detection in pulse repetition frequency agile radar systems. We introduce a new paradigm—a conditional dependency-based sequential decoupling framework. In contrast to conventional paradigms that either fragment the problem (divide-and-conquer) or confront it with exhaustive search (global optimization), our framework explicitly leverages the inherent hierarchical dependencies within the signal model. It establishes a strictly ordered processing chain: the ambiguity number search first compensates for the dominant range migration and achieves envelope alignment directly in the time domain. This, in turn, enables the augmented time‑reversal transform to separate velocity and acceleration phases analytically. Finally, with the decoupled signal structure, the virtual echo modeling synthesizes an expanded virtual array aperture from very few pulses. This sequential dependency transforms a high‑dimensional, coupled estimation problem into a series of tractable low‑dimensional sub‑tasks. The framework is supported by comprehensive theoretical analysis, including closed‑form detection probability in the virtual domain, verification of its constant‑false‑alarm‑rate property, and derivation of the Cramér–Rao lower bound under non‑uniform sampling. Simulations under low signal‑to‑noise ratio and severely limited pulse conditions demonstrate that the proposed method significantly outperforms state‑of‑the‑art techniques in estimation accuracy, computational efficiency, and robustness, providing a principled and practical solution for high‑speed maneuvering target detection in agile radar systems.