Global Navigation Satellite System (GNSS)‐based passive bistatic radar (PBR) systems hold promise for use as low‐altitude surveillance mechanisms in critical urban and suburban zones, attributed to their low power consumption, good concealment, and worldwide reach. However, the increasing demand for airspace regulation presents challenges for multi‐target tracking. Additionally, the limited power budget of GNSS signals results in low target SNR, restricting the detection range. Hence, a novel multi‐target track‐before‐detect (TBD) scheme is proposed. This strategy employs a dual‐channel coarse focusing (DCCF)‐based cluster centroid extraction algorithm to identify potential target information in the range Doppler (RD) domain. Subsequently, a modified cardinalised probability hypothesis density (MCPHD) filter is utilised, enhanced with a birth target intensity estimation module assisted by Doppler and a trajectory management module, to accurately track multiple targets under conditions of low SNR. Simulation results and performance analysis using the Optimal Sub‐pattern Assignment (OSPA) metric confirm the effectiveness of our approach. Furthermore, in a real‐world experiment using the GPS L5 signal as an illuminator, the authors successfully processed experimental data to track a civil aircraft over 10 frames, demonstrating the practical applicability of the proposed method in GNSS‐based PBR systems.