Research On Mobility In Millimeter Wave Communication

The current 4G(the forth generation mobile communication technology)is based on low-frequency(below 6GHz)signal carrier operation.The available spectrum resources are small and the frequency band is crowded,while have long been unable to meet the explosive growth of needs of the multimedia data communication.To solve the current problems,researchers have begun to use the millimeter-wave band(6GHz~300GHz)with rich spectrum resources for communication.However,the path loss of the millimeter-wave band in free space is large,and the penetrating power is lower than that of the low frequency electromagnetic wave used by 4G system.Especially when the millimeter-wave encounters the reflection and refraction of an obstacle or rainfall during the propagation process,the energy loss will be more serious,resulting in the propagation path existing in the millimeter-wave transmission system is mainly as the LoS(Line of Sight)path.Beacause the millimeter-wave has a small wavelength,we can use the huge array gain producted by the large directivity beam formed by a large-scale antenna array to compensate for the path loss.So far,many researchers on millimeter-wave have been carried out based on static channels.However,many scenes with relative motion,such as highways and high-speed rails,require mobile communication services.In addition to the angle of arrival(AoAs,Angles Of Arrive)and the angle of departure(AoDs,Angles Of Depature)constantly changing,the time-varying millimeter wave channel is also affected by the Doppler shift.A large amount of pilot overhead cannot meet the needs of millimeter wave channels.Therefore,it is very urgent and valuable to study the millimeter wave channel in the mobile state.This paper mainly studies the problem of channel estimation and beam tracking.First,we study the problem of millimeter-wave channel estimation in the mobile state.Using a digital-analog mixed multiple input multiple output(MIMO)system model,a suitable frame structure is designed to divide the channel estimation into two stages: angle estimation and path gain estimation.The angle estimation phase uses an improved block sparse compressed sensing algorithm for estimation.Compared with the traditional compressed sensing algorithm and the Least Square(LS)algorithm,it can obtain better performance.In the gain estimation phase,a linear fitting of the path gain in each frame will obtain better performance than other literatures.Then,we study the problem of millimeter-wave beam tracking in the mobile state.In order to obtain the starting point of beam tracking,we use the received signal to train the starting angle of the beam at the beginning of the algorithm,and then use probability model tracking to obtain the angle information of the transmitting and receiving end during the movement.Through simulation analysis,it can be found that the above algorithm has better performance than the traditional beam training and beam tracking algorithms.Based on the above ideas,we further propose a codebook-based beam tracking algorithm to be applied to communication scenarios with different requirements.Using different algorithms for different scenarios will achieve better communication results.