关键词: 星基辐射源定位, 直接定位, 多普勒效应, 最大似然估计, 低信噪比, 到达频差

Abstract: To address the degraded geolocation performance of satellite-based emitter localization systems in low signal-to-noise ratio (SNR) environments, a satellite-based direct position determination (DPD) method using the Doppler effect was proposed. The method bypassed the intermediate parameter estimation stage of conventional two-step approaches by exploiting the global constraint that all satellite observations originate from the same emitter, and constructed a location-dependent cost function based on maximum likelihood estimation to achieve direct mapping from raw observation data to emitter location. Two solving algorithms based on eigenvalue decomposition and fast Fourier transform were designed for unknown and known signal scenarios, respectively. The Cramér-Rao lower bound (CRLB) for satellite-based Doppler direct geolocation was derived, revealing the relationships between geolocation accuracy and carrier frequency, observation geometry, and SNR. Simulation results demonstrate that when SNR is below -25dB, the proposed method achieves geolocation accuracy improvements exceeding 55% for unknown signal scenarios and exceeding 92% for known signal scenarios, compared with traditional frequency difference of arrival (FDOA) measurement methods. Geolocation errors remain within 8km even under extremely low SNR conditions of -30dB. Furthermore, the method is capable of resolving multiple co-frequency unknown signal emitters separated by 50km, demonstrating multi-target geolocation capability. The proposed method overcomes the information loss problem inherent in conventional two-step methods under low SNR conditions, providing an effective technical approach for satellite-based emitter geolocation in weak signal environments.

Key words: satellite-based emitter geolocation, direct position determination (DPD), Doppler effect, maximum likelihood estimation, low signal-to-noise ratio, frequency difference of arrival (FDOA)