Radar-based passive localization plays an important role in non-cooperative sensing and situational awareness. In angle-of-arrival (AOA) based passive localization, dual-aircraft cooperative positioning offers advantages such as reduced model parameters, high angular accuracy, and good deployment flexibility. However, in multi-target scenarios, conventional approaches usually process the measurements from different platforms independently and then perform target association and position estimation through joint equations, which may suffer from matching errors and localization ambiguity. To address these challenges, this paper proposes a multi-target passive localization algorithm based on joint dual-aircraft MUSIC processing. By constructing a unified covariance matrix from the observations of two platforms, the proposed method performs joint angle estimation and target matching directly in the spectral domain, avoiding additional post-processing association steps. Furthermore, a dimensionality-reduction projection strategy is introduced to significantly reduce the computational burden of multi-dimensional spectrum evaluation and peak searching, while maintaining angular estimation accuracy. In the localization stage, a classical maximum likelihood–based solver is employed as a unified backend to estimate target positions from the estimated bearing information. Simulation results demonstrate that, under quasi-static target conditions, the proposed algorithm achieves reliable localization accuracy and improved robustness in multi-target scenarios, indicating its potential applicability in practical passive sensing systems.
Wang et al. (Sun,) studied this question.