YAMAZAKI Takato (OHNO Kohei laboratory), a second-year student of the master’s course in the Network Design Program, Graduate School of Advanced Mathematical Sciences, received the Student Encouragement Award at the Oral Session of the December Research Meeting of the Technical Committee on Wideband System (December 21, 2023). This award is given to the speaker who makes an excellent presentation among those who wrote a paper as the first author and made a presentation on the day at the Oral Session of the Technical Committee on Wideband System.
The title of the awarded presentation is "A Study on Prediction of Power Arrival at Ahead Point during Indoor Walking Using LSTM." This paper proposes the prediction of the received power of first arrival wave of indoor millimeter waves. The channel is measured assuming that a user walks indoors, and the received power of the first arrival wave is obtained from the impulse response. LSTM (Long Short-Term Memory) is used to predict the received power. On the other hand, the received power obtained in the experiment includes errors of measurement. Therefore, in order to reduce the influence of measurement errors on learning and to improve prediction accuracy, prediction accuracy is improved by preprocessing the learning data. Furthermore, it was shown that the application of this method and the closed-loop processing enables the prediction of the received power at a point ahead. Machine learning enables the prediction of millimeter-wave power, achieving beamforming of antennas and switching of communication systems, resulting in more reliable millimeter-wave communication.
≪
Japanese version≫