Signal models for TDOA/FDOA estimation

Emitter position and velocity estimation given time and frequency differences of arrival

“…Other covariance matrices for different models, which are obtained from analyzing the CRLB, can also be used instead. For example, the covariance matrix given in [13] assumes that the transmitted signal and the attenuations to the sensors are known, and the covariance matrix given in [14] assumes that the signal is deterministic but unknown and also the attenuations to the sensors are unknown. We evaluate the position RMSE and velocity RMSE versus the parameter γ for a sensor array with a circular configuration.…”

Multistatic pseudolinear target motion analysis using hybrid measurements

“…For stationary Gaussian signals in acoustic sonar applications, the TDOA and FDOA measurements are asymptotically uncorrelated and the CRLB matrix of joint TDOA/FDOA measurements is in block-diagonal form under relatively mild conditions such as sufficiently large signal time-bandwidth product and moderately high signal-to-noise ratio [31][32][33]. In contrast, the TDOA and FDOA measurements are strongly correlated for deterministic non-stationary non-Gaussian signals, especially chirp-like signals in radar applications [32,33]. The covariance matrices for the TDOA measurements, εε K k , and the FDOA measurements, ϱϱ K k , are modeled at each time instant by [34] i k …”

“…Nguyen, K. Doğançay / Signal Processing 130 (2017)[22][23][24][25][26][27][28][29][30][31][32][33][34][35][36] …”

Two‐step estimator for moving‐emitter geolocation using time difference of arrival/frequency‐difference of arrival measurements

“…Since the signal models for TDOA/FDOA are assumed to be deterministic in the case of passively locating an electromagnetic emitter [23], the observed TDOA/FDOA measurements u m corrupted by the additive noise are represented by …”