Researches And Improvement On LTE Downlink Channel Estimation Algorithm

LTE (Long Term Evolution) is a transition technology of3G to4G, which usesOFDM and MIMO technology as the sole criterion for the evolution of its wirelessnetwork. The OFDM technology is a multi-carrier modulation technique, it canreduces the effection of inter-symbol interference (ISI) and inter-symbol interference(ICI) on the signal, and has a high spectrum efficiency, which can effectively combatmultipath delay spread. MIMO technology based on the the original transmissionbandwidth and power consumption, can be multiplied to improve the spectrumutilization. Combing the MIMO technology and OFDM technology can significantlyimprove the network coverage and transmission reliability. And accurate channelestimation,which gives full play to the superiority of the key, is to ensure that thetransmission quality of the MIMO-OFDM system,.Based on the basic principles of MIMO and OFDM,the two key technologies ofLTE, this paper expounds the necessity combination of MIMO-OFDM,then makes adetailed introduction of LTE wireless channel model and radio channel estimationtechniques; focuses on the MIMO-OFDM channel estimation techniques based ontraining sequence and pilot.The channel estimation based on training sequence is suitable for slowly varyingchannel.firstly,the choice of auxiliary information pattern is described in detail, andthe applicable environmental of different auxiliary information programs is describedin theory, which is verified by simulation. Then for each channel estimation algorithmbased on training sequence is studied respectively, putting forward the MIMO-OFDMchannel estimated could use SISO-OFDM channel estimation algorithm.Finally,thevarious algorithm were analyzed Through the simulation test.Pilot-based channel estimation is suitable for fast-varying channel，which obtainsthe channel information of pilot position firstly, then to do the interpolation to get thewhole channel response. Each channel estimation algorithm (ie, a variety ofconventional interpolation algorithm) of the data position were studied, the lastsimulation results are compared and analyzed, and made a detailed analysis of theperformance of various interpolation algorithm.Finally, based on LS algorithm a new channel estimation algorithm is proposed.Because of its simple algorithm, LS algorithm has been widely used in the OFDM system.But the general LS algorithm is to ignore the effects of noise, so the estimationerror will be relatively large. In order to reduce the LS estimation error, the articleconcludes with optimal FIR filter channel estimation algorithm based on LS algorithm.In adding noise environment, compared with the traditional windowing algorithm, thisalgorithm can guarantee local frequency performance is optimal better, and thuseffectively reduces the mean square error (MSE). Simulation results demonstrate itseffectiveness.