Abstract: The rapid development of satellite internet constellations in low Earth orbit （LEO） has provided new opportunities for global communication and positioning services. In particular， SpaceX’s Starlink constellation has emerged as one of the most representative LEO satellite networks because of its size and global coverage. In addition to communication services， Starlink’s downlink signals are considered as potential Signals of Opportunity （SoOP） that can be utilized to enhance or supplement traditional satellite navigation systems such as GPS and BeiDou. Given that the constellation comprises thousands of satellites， these signals are characterized by high stability， wide bandwidth， and global accessibility， which makes them particularly suitable for use in navigation. In this study， we propose a Doppler-based positioning method that utilizes Starlink’s downlink data frame signals to improve the accuracy of opportunistic positioning via the Starlink constellation. This method combines the primary synchronization symbols with user data symbols that contain pilot sequences in the downlink signal by employing a two-step estimation algorithm to measure Doppler frequency shifts. The measurements are then fed into an extended Kalman filter to compute the receiver’s position. The results of simulations based on Starlink’s real orbital parameters demonstrate that the proposed Doppler shift estimation algorithm outperforms traditional methods such as phase-locked loops， maximum likelihood estimation， and segmented differential cross-correlation in terms of noise resistance and accuracy. At a signal-to-noise ratio of 0 dB， the root mean square error （RMSE） of the estimation was 3.7 Hz， while the RMSE during satellite overpasses was 2.1 Hz. Furthermore， positioning simulations using five satellites showed a three-dimensional positioning error of 13.5 m and a horizontal error of 6.1 m. Compared to existing beacon-based positioning methods that require six satellites， this represents an improvement of 9.4 and 0.4 m， respectively. These results indicate that the proposed data frame-based positioning algorithm surpasses traditional methods in terms of both noise robustness and positioning accuracy. Thus， the proposed method enhances the performance of opportunistic positioning techniques based on Starlink satellites.