3D Indoor Localization of RFID Tags with Phased Array and Deep Neural Network

This paper presents indoor localization of RFID tags with a phased array antenna and a deep neural network (DNN). The antenna can steer the linear null in its radiation pattern crosswise, and the DNN estimates the 3D location of a tag (x, y, z) in a chamber using all received signals back scattered from the tag for all steered radiation patterns. The antenna and indoor radio propagation are simulated at 920 MHz, and the obtained received signals are used as the input of the DNN to train and test itself. Finally, the generalization capability for the polarization and height of a tag, etc. are compared with the conventional MUSIC method.