TDOA and FDOA Estimation Methods for Signals with Known Waveforms and Their Performance Analyses

‍ ‍In-depth analysis and long-term accumulation can be used by non-cooperative positioning systems to obtain rich prior information about the waveforms of various electromagnetic signals such as those of television， radio， and navigation systems. The exploitation of this information is expected to significantly improve the signal-parameter estimation accuracy. This study focused on the time-difference-of-arrival （TDOA） problem and frequency-difference-of-arrival （FDOA） estimation of known waveform signals in the context of relative motion between the observation stations and emitter， aiming at providing a solution for the passive location problem of deception jamming for navigation and communication sources， as well as of illegal users. First， the data received at an observation station was modeled by introducing the known signal waveform information， and the relationships between the observed data and the time delays and frequency shifts of the signals of different observation stations relative to the original signal were visually presented. Then， taking full advantage of the known waveform of the incident signal， the TDOA and FDOA of two stations were effectively estimated by estimating the time delays and frequency shifts of the two stations， respectively， and then differentiating the estimates on different stations. On this basis， this study analyzed the theoretical lower bound of the estimation accuracy of the TDOA and FDOA parameters in this scenario， which revealed the influence of factors such as the received signal amplitude on the TDOA and FDOA estimation accuracies. The relationship between the estimation accuracies for the dual-station TDOA and FDOA values was also analyzed， along with the accuracies of the single-channel time delay and frequency shift estimates. Finally， the parameter estimation performances of the new method in different environments were verified using simulations. The simulation results showed that in the given signal environments， the adaptability of the new method to a low signal-to-noise ratio was approximately 15 dB higher than that of the traditional method based on the cross-ambiguity function； and when the parameter estimation performance of the new method reached convergence， its TDOA and FDOA estimation accuracies were very consistent with the theoretical lower bound given here.