Localização coerente passiva de um alvo usando redes neurais feedforward

This work addresses the passive coherent location of a target in a scenario with a single transmitter and several receivers. Firstly, a feedforward neural network estimates the target location based on the time differences of arrival among the target-reflected signals acquired by the receivers, whose positions are fixed and known. In a second case, the neural network also considers the receivers' positions as inputs to compute the target position. In both cases, the transmitter has a fixed location. Problem formulation, methods employed, as well as corresponding simulation results are presented. The simulations consider scenarios with different numbers of receivers and different degrees of errors to determine the network parameters.