The Research Of Channel Estimation Algorithm For LTE Downlink

In order to satisfy the requirement of high speed data transmission and multi service integration,the long term evolution proposed by 3GPP has emerged as the times require,and LTE technology has been regarded as a key point of the research by many experts and scholars with its unique attraction.OFDM technology,as the core technology of LTE downlink,has the advantages of high transmission rate and high spectrum utilization.Channel estimation is an important part of OFDM technology,the purpose is to enable the receiver to recover the channel state information accurately,and to improve the performance of the system.In this paper,we mainly study the channel estimation based on pilot in LTE downlink.The main research work and innovative points are as follows:(1)The paper researches the characteristics of the LTE downlink physical layer,and mainly describes the frame structure,resources structure,physical channel and the reference signal,and the paper focus on the basic principle and system model of the OFDM;Because of the multipath fading in the wireless channel,in view of this,the paper focus on the statistical characteristics of multipath fading,and the channel impulse response model is derived.(2)In the light of the channel estimation algorithms based on pilot,three pilot structures are studied: block type pilot,comb type pilot,and lattice type pilot.The paper focuses on the three classical channel estimation algorithms about pilot: LS algorithm,MMSE algorithm,and SVD-MMSE algorithm,and the performance comparison of the simulation experiment is also given.At the same time,the problem of how to get the whole channel response from t he channel response at the pilot position is discussed,and several interpolation algorithms are discussed and the complexity of the algorithm is compared.(3)The paper studies the channel estimation algorithm based on DFT.When the channel is the non sample spaced channel,it will cause the energy leakage of channel impulse response,to solve this problem,this paper put forward the method by Hamming window processing in the frequency domain.The DFT algorithm only eliminates the noise outside the cyclic prefix,and the internal noise of the cyclic prefix is not eliminated,according to this problem,the paper puts forward the method of setting optimal threshold for de-noising in the time domain,especially for the optimal threshold,the paper proposes the improved selection method,and gives the design of simulation experiment.The experiment results show that,the estimation performance of the improved DFT algorithm is improved.