Research On FBMC Transmission System And Channel Estimation

The fifth-generation(5G)mobile communication puts forward higher requirements for the spectrum utilization and flexibility of new modulation waveforms.Compared with orthogonal frequency division multiplexing(OFDM),the time-domain and frequency-domain energy of the prototype filter used by the filter bank multi-carrier(FBMC)is concentrated on a narrower bandwidth,which greatly reduces the side effects,the offset quadrature amplitude modulation(OQAM)mapping is used to split the complex data into two real data in one symbol period to overlap and send,which not only ensures the transmission efficiency but also suppresses the inter-carrier interference(ICI).Since each sub-carrier of FBMC is only orthogonal in the real number domain,the inherent interference caused by the phase rotation factor and the transmission multiplexing response of the filter at the transceiver end makes FBMC more effective than OFDM in terms of channel estimation and the combination of multiple input multiple output(MIMO)systems.Pilot or send data needs additional processing.In this paper,after studying the relevant contents of the FBMC transmission system,in the neighborhood zeroing interference elimination method,the time-frequency resource grid points in the first-order neighborhood around the inserted pilot are filled with data 0,and the auxiliary pilot is designed as the ratio of the pilot's imaginary part interference to the interference weight at the auxiliary pilot position to eliminate the imaginary part interference.The ratio of the received signal data observed on the receiving antenna to the pilot data can be used to obtain the complex channel fading of the pilot grid points,the complex channel fading at other data symbols is obtained through linear interpolation,and the estimated value of the transmitted signal is obtained through zero-forcing equalization and OQAM de-mapping.In order to make the ICM and the IAM suitable for the time-selective channel of the FBMC system,the secondary pilots are filled to form a comb-like pilot structure,and the LS is used to calculate the feedback interference to obtain better performance;and the ICI is used by extending the block pilot to the comb pilot in the time domain.Although the estimation accuracy is reduced,the fast-changing channel can be estimated,and the extended Kalman filter is used to improve the performance.When estimating the fading and frequency offset,according to the known pilot symbols,use the second-order autoregressive model for the real and imaginary parts of the complex channel fading,and build the frequency offset into a first-order model,that is,the equation of state.The observation equation uses the predicted value at the last moment to do a first-order Taylor series expansion to estimate the parameters.In the SISO system,to estimate the multipath time-varying channel of FBMC,an extended Kalman filter algorithm based on auxiliary pilot is given,while in the MIMO system,the frequency domain conjugate data block is sent to make the imaginary part interference between data blocks opposite to each other,improve the accuracy of FBMC channel estimation,and combine this structure with the extended Kalman filter to estimate the channel parameters.