Electromagnetic Radiation Source Location in NLOS Environment Based on General Regression Neural Network

The electromagnetic radiation source location in urban environment has always been a problem due to the existence of obstacles. In this case the accuracy of the signal parameters is reduced, and the performance of traditional source location method will decline sharply. In the non-line-of-sight (NLOS) environment, the propagation law of electromagnetic waves is complicated to be characterized by mathematical analysis. In this paper, a general regression neural network (GRNN) was proposed to establish the relationship between the received signal strength (RSS) and the electromagnetic radiation source location parameters. The simulation results showed that the GRNN maintains the favourable locationing accuracy by selecting proper parameter SPREAD than the BP network. Furthermore, the proposed method shows good robustness cross-validation.