Ye Yuan

This article addresses weak target detection and tracking in pulsed Doppler (PD) radars plagued by range and Doppler ambiguities, employing track-before-detect methods. Conventional PD radars often resolve ambiguities using the intra-frame multiple PRF (IF-MPRF) mode, which transmits multiple PRF signals within the same beam dwell time. However, this mode faces a trade-off: while longer coherent integration times per PRF improve detection, they reduce the data update rate for tracking. Furthermore, subsequent detection, ambiguity resolution, and tracking methods under this mode suffer from high computational complexity due to processing multi-PRF and multi-frame measurements. To overcome these limitations, we first propose a multi-beam agility PRF (MB-APRF) mode, which transmits a single PRF per beam dwell time and then randomly switches PRFs across beam positions. Compared to IF-MPRF, this mode enables longer coherent integration while maintaining a high tracking data rate. Next, we introduce a multi-frame decoding criterion to distinguish real targets from ambiguous measurements. Building on these advancements, we propose the multi-frame decoding and track-before-detect framework and its fast implementation methods for target trajectory estimation. This framework enhances energy integration gains by optimizing both coherent and non-coherent integration times. Additionally, this framework mitigates the performance degradation caused by the blind range and velocity zones of PD radars through an adaptive detection strategy. Numerical results demonstrate that the proposed technique outperforms its direct competitor, achieving higher detection gain, faster tracking data rates, and lower computational cost.

Distributed hybrid active-passive radars (HAPRs) with separately deployed transmitters and receivers face localization performance limitations due to the inevitable presence of direct-path interference (DPI). To improve target localization accuracy and enhance DPI suppression, we develop a direct position estimator tailored for stationary/low-speed targets in distributed HAPRs, which jointly exploits signals from both active transmitters and non-cooperative illuminators of op portunity (IOs). However, the joint processing of active and passive observations, coupled with the unknown signals from non-cooperative IOs, leads to a intractable high-dimensional optimization problem. The expectation-maximization algorithm is employed to address the optimization problem with low complexity, iteratively suppressing DPI and updating the position estimate. To further reduce computational complexity, we use a sequential estimation method that decomposes the high dimensional problem into a sequence of low-dimensional ones. Wealso derive the Cram´ er–Rao Lower Bound (CRLB) to provide an accurate performance benchmark for the proposed estimator. Extensive numerical analysis shows that the proposed direct estimator achieves estimation accuracy closer to the CRLB than estimators used in only active or passive distributed radars.

To enhance the spatial perception capabilities of distributed radar networks (DRNs) in complex environments, this paper investigates the coordinated optimization of beampattern-aware topology and antenna pointing for coverage maximization. Unlike prior works that focus exclusively on radar node placement, the proposed framework introduces antenna pointing as an additional optimization dimension, enabling cooperative beam direction adjustment and substantially improving regional coverage. A unified system and signal model is first established, encompassing both transmit and receive channels. Based on this model, we derive the antenna gain for the uniform linear array (ULA) and a detection probability expression that explicitly incorporates antenna pointing effects. Furthermore, a beampattern model is constructed to characterize the influence of pointing errors on radar coverage, and a corresponding quantitative metric is defined to evaluate performance. Leveraging this metric, the coverage maximization problem is formulated as a nonconvex joint optimization over both node positions and antenna orientations. To efficiently solve this high-dimensional problem, an enhanced particle swarm optimization algorithm integrated with the butterfly optimization algorithm (BOA-PSO) is adopted, which improves global convergence and search efficiency. Numerical simulations under representative deployment scenarios demonstrate that the proposed method achieves superior coverage and convergence stability, highlighting its potential for practical DRN configuration and resource management.

Hybrid active-passive radars (HAPRs) integrate the sensing advantages of both active and passive radars. Active radars offer high sensing accuracy due to known transmitted waveforms, while passive radars provide improved low probability of interception (LPI) by utilizing non-cooperative illuminators of opportunity (IOs). Although the potential of HAPRs has attracted increasing attention in recent years, testing problems and corresponding detectors for distributed HAPRs remain limited. We formulate a composite binary hypothesis testing problem by incorporating heterogeneous observations from active and passive radars. To address this problem, two detectors are developed following the generalized likelihood ratio test (GLRT) and Wald criterion, respectively. Given the potential differences in observation quality between active and passive radars, arising from differences in system parameters, channel responses, and signal processing mechanisms, traditional global fusion of all active and passive observations may lead to performance degradation. To mitigate this issue, we propose improved GLRT and Wald detectors based on a local fusion strategy that selectively fuses high-quality observations. Simulation results demonstrate that global fusion-based HAPR detectors outperform pure active or pure passive detectors under specific scenarios. Furthermore, the HAPR detectors with local fusion consistently achieve superior detection probabilities and wider coverage, verifying their robustness and effectiveness.