Research And Implementation Of PDSCH Resource Extraction And Channel Estimation For 5G Road Tester

With the further development of the 5th Generation(5G),5G has entered into a scale development stage.The complex network deployment and diverse application scenarios have created a huge challenge for the network test of 5G.In order to better process and analyze data,5G road tester has great advantages for 5G network testing and large-scale user behavior testing.This thesis focuses on resource extraction and Channel estimation of Physical Downlink Shared Channel(PDSCH)in 5G system,which is based on the major theme project of technological innovation and application development of "R&D and application of 5G road tester" in Chongqing.In this thesis,the resource allocation in time-frequency domain,reference signal extraction and resource mapping process of PDSCH are studied.In view of the change of PDSCH channel estimation that the demodulation reference signal replaces the cell reference signal in 5G system,the thesis introduces the generation process of the modulation reference signal,and proposes a modularized pseudo-random sequence generation method.Firstly,an integer multiple of 12 state offsets is introduced to solve the problem of time wasting,which is caused by the fixed state offsets.Then,the sequence which is required by the system is divided according to the new state offset.Finally,the pseudo-random sequence of the initial position resource block is generated to provide initial parameters for obtaining the demodulation reference signal.Comparing with traditional sequence generation,the result indicates that the new method consumes less time and improves the efficiency of reference signal generation.Channel estimation performance in 5G PDSCH is an important factor,which affects the overall performance of the system.Based on pilot channel estimation,this thesis studies channel estimation algorithm from pilot and non-pilot respectively.Finally,The Least Square algorithm is used at the pilot.At non-pilot,obtaining the Signal to Noise Ratio bases on the characteristic of channel energy concentrated distribution in time domain,Linear Least Mean Square Error(LMMSE)interpolation is completed in frequency domain with sliding window,while linear interpolation is used in time domain.Simulation results demonstrates that this design scheme can take a good compromise between system performance and computational complexity.Based on the research of PDSCH resource extraction and channel estimation algorithm,the design of FPGA is carried out by using the top-down modular design idea.The resource extraction module uses bitmap to improve the operation rate,and the channel estimation uses parallel pipeline to reduce the delay.Then,simulating the waveform of the module by using the Model Sim software.Comparing the output results of Matlab with Model Sim,the correctness of the scheme is verified.While the thesis analyzes the system resource occupation and time consumption.Finally,the whole machine test of 5G road tester is carried out to verify the feasibility of the design scheme.