A Robust LPNN Technique for Target Localization Under Hybrid TOA/AOA Measurements

This paper presents an approach based on the Lagrange programming neural network (LPNN) framework for target localization under the outlier situation. The problem is formulated as a minimization problem of the mixture of \(l_1\) and \(l_2\) norms of the measurement errors with time of arrival (TOA) and angle of arrival (AOA) measurements. In our approach, we introduce an approximation function for the \(l_1\) norm to avoid non-differentiable points of the \(l_1\) norm. Moreover, we study the network stability of the proposed approach. We carry out simulations with various settings to validate the performance of the presented approach. In addition, the proposed method is compared with the existing linear and nonlinear target localization techniques. Based on the mean square error (MSE) values obtained under different settings, the proposed approach is superior to the other comparison algorithms.