Efficient 2D localization of a number of mutually arbitrary positioned stochastic EM sources in far-field using neural model

In this paper, an architecture of a model for 2D space localization of sources of stochastic radiation is presented. Model represents the connection between planar antenna array for signal sampling and multilayer perceptron network for 2D Direction of Arrival (DOA) estimation. Proposed model is applied for the case of two stochastic sources that are moving independently in a plane parallel to a plane of antenna array.