METHOD AND DEVICE FOR SETTING PARAMETER OF RECONFIGURABLE INTELLIGENT SURFACE BASED ON DEEP NURAL NETWORK TO ENHANCE INFORMATION TRANSMISSION RATE OF INDOOR WIRELESS ENVIRONMENT

摘要

Various embodiments relate to an apparatus and method of a base station for setting a phase matrix of a reconfigurable intelligent surface (RIS) device in an indoor space, based on a deep neural network, learning a phase matrix for a RIS device and transmit the learned phase matrix to the RIS device. According to various embodiments, an apparatus and method of a base station randomly generates a channel between a user device and a RIS device in an indoor space, inputs the generated channel to a deep neural network, and detects an optimal phase shift for the generated channel and a learned phase matrix can be constructed with the detected phase shift.