1D Localization of Mobile Stochastic EM Sources with Variable Radiated Power Using Two-Stage Neural Model

In this paper, a cascade two-stage neural model is proposed for 1D DoA estimation of mobile stochastic EM sources with variable radiated power. Architecture of the neural model is based on MLP neural networks whose number is determined by the dimension of the linear antenna array at the receiver. Model accuracy is illustrated on the example of two uncorrelated EM sources whose radiated powers are fluctuated during their movement.