New Approaches for TDOA Source Localization: Selection Based LS and Deep Learning TDOA Konum Bulma için Yeni Yaklaşimlar: Seçim Tabanli EKK ve Derin Ö?grenme
34th Signal Processing and Communications Applications Conference, SIU 2026, İstanbul, Turkey, 7 - 10 July 2026, (Full Text)
- Publication Type: Conference Paper / Full Text
- Doi Number: 10.1109/siu71813.2026.11637042
- City: İstanbul
- Country: Turkey
- Keywords: analytical solution, artificial neural network, localization, TDOA
- Middle East Technical University Affiliated: Yes
Abstract
In electronic warfare systems, the position of a target radar can be estimated using time difference of arrival (TDOA) measurements. This study focuses on a location estimation scenario with three receivers, where position uncertainty is significant. Two methods are proposed for this problem that eliminate ambiguity by narrowing the solution space. The first is the least-squares (LS) being an analytical solution-based, intelligent selective estimator. This estimation method offers lower computational complexity while performing similarly to maximum likelihood methods. The second method is an artificial neural network-based estimator. Although this method lags behind maximum likelihood methods in terms of position accuracy, it is considered a good alternative when an analytical solution is not available.