Uplink Channel Estimaiton Based On SRS Sequence In 3D MIMO Systems

In recent years,MIMO technology has gained wide attention from scholars at home and abroad.With the development of communication systems,the traditional 2D MIMO channel models are not able to meet the analysis requirements of system performance anymore.In order to better analyze the performance of the communication system,we do our research based on the 3D MIMO channel models.Channel estimation technology plays an important role in the operation of the whole system as it is an important part of 3D MIMO systems.Many follow-up actions require CSI provided by channel estimation such as subsequent receiver equalization detection processing,user group,etc.The performance of MIMO systems is closely related to the accuracy of eatimaiton of the channel impulse response.In the 3D MIMO system,the number of base station antennas is large.In the TDD(Time Division Duplexing)mode,the pilot overhead does not increase with the number of base station antennas due to the reciprocity between the uplink and downlink channels.It makes the TDD mode is currently widely used.Thus,this paper focuses on the channel estimation technology based on SRS in 3D MIMO systems under TDD mode.Firstly,this paper introduces the 3D MIMO system model based on 3GPP TR 36.873 protocol.Then,some existing channel estimation algorithms are introduced and verified by simulation.Then,based on MATLAB,a channel generation platform introduced in Chapter 2 is built.Then,based on this channel platform,we study the statistical characteristics of multipath distribution under different scenarios,different Rice factor K and different antenna spacing,and draw several conclusions about the impact of these characteristics on channel estimation.At the end of this paper,a new channel estimation algorithm based on LOS/NLOS identification is proposed.Firstly,a LOS/NLOS user identification algorithm is proposed.Then we estimate the channel of NLOS users and LOS users respectively according to the results of identificaiton.For NLOS users,we propose an adaptive in-window denoising DFT-based channel eatimation.For LOS users,an novel channel eastimation method based on joint-decision is proposed with the prior knowledge of the 3D MIMO channel characteristics.Simulation results verify the proposed algorithm outperforms traditional DFT-based channel estimation.Both LOS users and NLOS users can accurately estimate the channel impulse response after the LOS/NLOS identificaiton.And it is achieved with additional liner complexity.