Research On Multicore DSP Realization Algorithm Of Physical Uplink Shared Channel In LTE-A System

As the smooth evolution of LTE system,LTE-A system plays an important role in the evolution of mobile communication system.As the uplink data channel of the LTE-A system,the physical uplink shared channel plays a key role in the performance of the whole system and the transmission rate of the uplink data.According to the demand of project ?Research on Key Technologies of LTE multi UE base station load and capacity test? and ?FDD-LTE-A signal receiving and processing technology?,this thesis studies the key technologies of the PUSCH channel in LTE-A system,and completes the functional verification of the PUSCH related modules on the multicore DSP chip TMS320C6670.The main contributions are organized as follows:Firstly,based on the physical layer protocol of the 3GPP R12 version,this thesis summarizes the physical layer time-frequency resources and key technologies of LTE-A system,analyses the baseband processing of the PUSCH and introduces the related modules,discusses the generation process of the demodulation reference signal in detail and analyses its characteristics.Secondly,aiming to enhanced MIMO technology,this thesis researches importantly the channel estimation algorithm,introduces the traditional least square(LS)and minimum mean square error(MMSE)algorithms,and then studies the channel estimation algorithm of LS time domain windowing.Based on the characteristics of DMRS,this thesis discusses the linear interpolation algorithm based on LS and improves it using the transform domain interpolation algorithm,proposes the inverse discrete fourier transform/discrete fourier transform interpolation algorithm based on LS and IDFT/DFT interpolation channel estimation algorithm based on linear minimum mean square error(LMMSE).Last,this paper simulate respectively to these channel estimation algorithm in the extended pedestrian A model(EPA)and extended typical urban model(ETU)channel environment,the simulation results show that the IDFT/DFT interpolation channel estimation algorithm based on LMMSE has the best performance.Thirdly,the thesis focuses on the design and implementation of PUSCH transmission module based on DSP,including the analysis of multicore navigation structure and the studies of the bit coprocessor(BCP)and fast fourier transform coprocessor(FFTC)based on multicore navigation.Based on the transmission link process of PUSCH,this thesis realizes CRC addition,code block segmentation,rate matching,scrambling and modulation using BCP,and realizes transmission precoding and generating baseband signal module using FFTC.By comparing processing results between DSP and Matlab,MSE of 10-8 order of magnitude is existed.Besides,the feasibility of the application of FFTC and BCP in PUSCH is analyzed from the point of view of processing time.Finally,this thesis studies the design and implementation of PUSCH receiver module based on DSP,researches design and implementation of LS time domain windowing channel estimation algorithm on DSP and optimizes the algorithm using FFTC.This thesis also studies the algorithms of ZF,MMSE and QR,the QR algorithm is selected as the PUSCH signal detection algorithm and implemented on DSP through the performance simulation and the complexity comparison.Through testing with simulation link,obtaining the result that the accuracy of TMS320C6670 is lower than Matlab,but the error is relatively small,thus can meet the requirements of LTE-A system.