Tek Kullacılı Çok Girdi Çok Çıktı Akıllı Yansıtıcı Yüzey Sistemi için Kanal Kestirimi

Abstract

There is a remarkable increase in the number of communication devices and mobile connections incident to the new and advanced technologies brought by the fifth-generation of communication. In order to reduce the high complexity, energy consumption, and hardware costs caused by this huge increase in the communication area, researches are already focusing on fifth-generation technologies. Recently, the technology called intelligent reflecting surface or reconfigurable intelligent surface is one of the technologies that has become the subject of research. Intelligent reflecting surfaces are obtained by placing a large number of passive reflective elements on a planar surface. Each element can independently control the amplitude and phase of the reflected signal. With this technology, it is aimed to recreate the wireless propagation environment by using passive elements on the smart reflective surface. Among the uses of smart reflective surfaces, it is considered to be used in situations where the clear line of sight between the transmitting and receiving equipment is blocked, the signal to the receiver is very weak, the channel conditions are not conducive to communication, or to reduce co-channel interference.

In this study, channels of smart reflective surface supported systems, in which there are many antennas in both transmitter and receiver equipment, are discussed. These communication channels are modeled with channel parameters between each antenna and reflector unit. Knowing these channel parameters is necessary to determine the beamforming and intelligent reflecting surfaces reflector parameters of the system. In this thesis, the problem of estimating the channel parameters in such a system is investigated. Three different channel estimation methods in the literature have been discussed and these have been examined and compared. The methods discussed are named as leasr squares Khatri-Rao factorizarion, biliner alternatin least squares and conjugate gradient on manifold optimization.