Research On Channel Estimation Technique For 5G FBMC Systems

In modern communication systems,Orthogonal Frequency Division Multiplexing(OFDM)technology is one of the most widely used multi-carrier transmission technologies,but it has the disadvantages of low spectrum utilization,sensitivity to frequency offset and slow outof-band attenuation.Filter Bank Multicarrier(FBMC)does not require cyclic prefix(CP)and has higher spectrum efficiency and good out-of-band suppression.Besides,FBMC does not need to maintain orthogonality between subcarriers,and can use the non-continuous spectrum resources flexibly.These characteristics make FBMC technology become a research hotspot in 5G.However,the system structure of FBMC is more complicated because of pulse shapes with good time and frequency localization characteristics being introduced.Meanwhile,FBMC must use Offset Quadrature Amplitude Modulation(OQAM)method to ensure the orthogonality between subcarriers,but the orthogonality is only in the real value field.There is inherent imaginary interference in the complex value field.This interference makes channel estimation in FBMC systems face great challenges.Channel estimation method of FBMC system is researched in this paper.Firstly,the channel models in wireless communication systems are introduced and the fading characteristics of wireless channel are analyzed.The basic theory and implementation methods of FBMC are introduced,including the design of prototype filter,the method of guaranteeing orthogonality between subcarriers and the implementation of filter bank.And the advantages and disadvantages of FBMC are analyzed.Then,channel estimation of FBMC system is addressed in this paper.The research contents include the pilot structure,channel estimation method and estimation interpolation algorithm.The existing scattered pilot-based channel estimation methods often have problems such as large power overhead and slow convergence of the iteration.A new joint channel estimation and iterative detection method is proposed,which first uses the auxiliary pilot to cancel the interference in the first-order neighborhood to complete the initial channel estimate,and then approximate the interference outside the first-order neighborhood through iterative detection,and feeds it back to correct the initial channel estimation.Simulation results demonstrate that the proposed method has fast convergence and high estimation accuracy.Finally,the channel estimation for channels of frequency selectivity in FBMC system is researched.The existing channel estimation always relies on the assumption of mind frequency selective channels.However,in the future mobile communications,the frequency selectivity of channels makes the performance of the existing channel estimation characterized by severe error floors at medium to high SNR.The channel estimation for channels of frequency selectivity relies on the Taylor series approximation of the received signal around the pilot frequencies,which is able to reduce the signal model distortion error.Based on the signal model,a new pilot structure and a channel estimation method are researched.Simulation results demonstrate that the channel estimation method is effective in estimating channels of frequency selectivity.