In this paper, a horn-shaped strip antenna exponentially tapered carved on a multilayer dielectric substrate for an indoor body position tracking system is proposed. The performance of the proposed antenna was verified by testing it as a tracking state of an indoor resting body position. Among different feeding techniques, the uniplanar T-junction power divider approach is used. The performance verification of the proposed antenna is explained through its compact size and 3D shape, along with a performance comparison of the return loss radiation pattern and the realized gain. The suggested antenna has an 88.88% fractional bandwidth and a return loss between 6 and 15.6 GHz, with a maximum gain of 9.46 dBi in the 9.5 GHz region. Within the intended band, the radiation pattern had an excellent directivity characteristics. The proposed antenna was connected to an NVA-R661 module of Xethru Inc. for sleeping body position tracking. The performance of the antenna is measured through microwave imagining of the state of the resting body in various sleeping positions on the bed using a Recurrent Neural Network (RNN). The predicted outcomes clearly define the antenna's performance and could be used for sensing and prediction purposes.
Keywords: Recurrent Neural Network; TEM horn; UBW; artificial intelligence; microstrip power divider; radar sensors.