Research On Two-port Network Channel Estimation For 5G Mobile Communication

Orthogonal frequency division multiplexing(OFDM)communication system has been officially adopted as the physical layer transmission technique of the fifth-generation mobile communication(5G)new radio(NR)due to its high data transmission efficiency and effective resistance to multipath fading channels.In order to improve the signal distortion caused by wireless channels,channel estimation has become a key technology for receiver design in communication systems.Facing the needs of future massive multiple-input multiple-output(MIMO)communication systems,5G NR,the new air interface standard specifies the concept of signal transmission and reception for multiple antenna ports.Therefore,conventional channel estimation algorithms of OFDM systems is no longer applicable due to lower efficiency.The channel estimation for the 5G NR standard is usually based on the demodulation reference signal(DMRS).For a multi-antenna port network,the DMRS sequence generally corresponds to the antenna port one by one to support the receiver to recover the transmitted data from different ports.In this paper,channel estimation technique of the two-port OFDM system is studied.In order to implement the identification of antenna ports at the receiver,DMRS sequences based on cyclic shift has been introduced at the transmitter.Moreover,the channel impulse response(CIR)of each port will be extracted through the time-domain windowing after channel estimation technique based on discrete Fourier transform(DFT).In response to the above situation,this paper first proposed the DFT-LS channel estimation algorithm based on the least square(LS)algorithm.However,due to the requirements of 5G NR terminal broadband adaptive technology to support the access of terminal devices working in bandwidth part(BWP),and BWP will cause the channel impulse response estimation to appear Gibbs phenomenon,edge distortion effects will appear,which deteriorates the performance of the channel estimation algorithm.Based on the method of extending effective bandwidth,this paper proposes a BWP-DFT-LS channel estimation algorithm without edge distortion.Simulation results show that the BWP-DFTLS algorithm is more suitable for the channel estimation in BWP scenarios than the DFTLS algorithm.Finally,this paper proposes the DFT-MMSE channel estimation algorithm based on the minimum mean square error(MMSE)algorithm and gives the autocorrelation matrix calculation method for the multipath Rayleigh fading channel.To satisfy the BWP requirement,the BWP-DFT-MMSE channel estimation algorithm without edge distortion is proposed.Simulation results show that the BWP-DFT-MMSE algorithm is more suitable for channel estimation in BWP scenarios than the DFT-MMSE algorithm and can provide significantly better system performance than the BWP-DFT-LS algorithm.