A Sparse TDOA Estimation Method for LPI Source Localization Using Distributed Sensors

In this paper, we propose a novel sparse time difference of arrival (S-TDOA) estimation method for the localization of the low probability of intercept (LPI) signals using distributed sensors. The proposed method uses the phase linearity of the wideband spatial covariances between the sensors obtained using sparse data. In the conventional distributed sensor systems, whole the data between the sensors is required to be transmitted to a central processor with a high computational load. In this work, the TDOA estimation is performed by using sparse sensor data in order to reduce computational load and high data sharing rate, which is more desirable for the practical implementations. It is shown in various simulations that the proposed method can effectively estimate the TDOA between the distributed sensors using the sparse multi-sensor data even in low signal-to-noise ratio. It is also shown that the TDOA estimation performance of the proposed method follows to Cramer Rao lower bound. Moreover, it is also shown that the data sharing rate and computation time are quite reduced without a significant performance loss with the sparsity process.