Research And Implementation Of Channel Estimation Techniques In 60GHz Millimeter Wave Communication System

There are abundant certification-free spectrum resources in 60 GHz band,which brings the pos-sibility to further improve the transmission rate of communication systems.With the development of RF technology,60 GHz millimeter wave communication has gradually become a research hotspot for next-generation wireless communication.From the perspective of propagation characteristics,60 GHz millimeter wave signals have large free space loss,with large signal attenuation caused by oxygen,rainfall,foliage and obstacles,making them suitable for short-range communication with high security and spatial reusability.Once the channel estimates are obtained by the channel estima-tion algorithm,most of the channel interference can be eliminated by using equalization techniques.The accuracy of the channel estimation affects the performance of the communication system,so the channel estimation module is an important part of the communication system.Therefore,the channel estimation algorithm and the high speed FPGA implementation of the channel estimation module for60 GHz millimeter wave communication system are investigated.After introducing the research background and related studies,the channel model of 60 GHz mil-limeter wave and the communication system architecture are presented.According to the propagation characteristics of 60 GHz millimeter wave,the channel model of the simulation system is established with reference to the S-V and UWB channel models,and the SC-FDE structure is used to design the 60 GHz millimeter wave simulation system framework and data frame structure as the basis of the channel estimation algorithm research,taking into account the effects of phase noise and carrier frequency offset.Then some channel estimation algorithms are studied and simulated for comparison based on the frame structure in the 60 GHz millimeter wave communication system.Both LS and DFT-based channel estimation algorithms are implemented in the system.Two channel estimation algorithms,Golay128 and Golay256,which utilize Golay complementary sequences,and the Gu512 algorithm based on sequence correlation are designed.These algorithms are also optimized based on the prin-ciple of DFT-based algorithm,and the MSE performance can be improved by 3 d B.The Gu512-NC algorithm,further using the noise threshold for noise reduction,is proposed to improve the MSE per-formance by 0.5 d B at low signal-to-noise ratio.Except for the LS algorithm,the BER performance of the other channel estimation algorithms are close to the ideal channel.Next,the overall hardware implementation scheme of the system is described in detail.The configurations of the ADC/DAC module and high-speed serial interface in the system,as well as the receiving and transmitting data processing procedures of the FPGA baseband processing system,are investigated as references for the design of the channel estimation module.Finally,the Gu512 channel estimation algorithm with lower complexity is selected to implement the channel estimation module in the FPGA system,and the 8-way parallel implementation method of I/Q separately correlated estimation is proposed for the system design,which can solve the ?/2modulation problem of the channel estimation sequence.The synthesis results in Quartus and the simulation results in Modelsim show that the channel estimation module can perform the channel estimation function at the maximum clock frequency of 495.79 MHz using 4% of the logic utilization and the timing is in accordance with the design expectation.