Research On Multi-core Dsp Realization Of Physical Downlink Shared Channel In LTE-A

Long term evolution-advanced(LTE-A)system is an important turning point in the development of wireless communication industry,and the research and implementation of LTE-A system play a pivotal role in the progress of human society.With the increasing transmission rate of the communicatio n system,the baseband processing based on single-core digital signal processor(DSP)has not been able to meet the real-time requirement,and the multi-core parallel mode and wireless communication coprocessor are widely used.According to the project development requirements of “FDD-LTE-A signal receiving and processing technology” and “Research on Key Technologies of LTE multi UE base station load and capacity test”,in the thesis,the realization of the physical downlink shared channel(PDSCH)in LTE-A system on multi-core DSP is researched.The main work and innovations of this thesis are as follows:1.For the different time-frequency resource mapping between UE-specific reference signal and cell-specific reference signal in the LTE-A system,the frequency division multiplexing and code division multiplexing are used between the ports,and the frequency division multiplexing technology is only used by the cell reference signal.Based on the adequate research of the characteristic and traditional channel estimation algorithms,two channel estimation algorithms based on the dedicated pilot information for the eight antennas are designed,which are LS-UE and LMMSE-UE algorithms.Through the simulation,the feasibility of the two algorithms is obtained,and then considering the performance and complexity of algorithm,the LS-UE algorithm is realized on the multi-core DSP,which proves the practicability of the algorithm.2.According to the complexity of the traditional soft sphere decoding algorithm which is related to the initialization of the symbol distance and the Q R decomposition algorithm which does not consider noise.An improved soft sphere decoding algorithm is presented,which is based on the QR decomposition detection algorithm considering noise and soft sphere decoding algorithm.The simulation results show that the performance of the new algorithm and traditional algorithm is basically the same,while the complexity is decreased to a certain extent.When the signal-to-noise ratio is 0,the average number of search points can be decresed by 5% to 10%.The result indicates that the new algorithm is more suitable for the realization of baseband system.Finally,the feasibility and efficiency of the new detection algorithm are verified by multi-core DSP.3.The fast fourier transform coprocessor(FFTC)and bit rate coprocessor(BCP),which are provided by multi-core DSP,are used to implement the sender of PDSCH.And compared with the realization of single-core DSP,the processing time shows that the processing performance of coprocessor is significantly better than the single-core DSP,and with the increasing of data,the processing of coprocessor will reach the pipeline operation,and the processing time of single-core DSP is proportional to the amount of data.4.Turbo decoding is the most time-consuming in the implementation module of PDSCH receiver,and in this thesis,the turbo decoder coprocessor(TCP3d)is used to realize the turbo decoding.Similarly,compared with the realization of single-core DSP,the processing time shows that the coprocessor can evidently increase the processing efficiency of the turbo decoding.