Research On PAPR Performance Optimization Algorithms In GFDM Wireless Communication System

At present,the mobile communication network is evolving from the fourth generation mobile communication technology(4G)to the fifth generation mobile communication technology(5G),and 5G has higher requirements in terms of speed,energy consumption,delay and user experience.Orthogonal frequency division multiplexing(OFDM)is not well suited to meet the new features of 5G.Generalized frequency division multiplexing(GFDM)as an alternative waveform of 5G,once raised has attracted wide attention.However,as a multi-carrier technology,GFDM system has a considerably obvious defects in high peak-to-average power ratio(PAPR),even higher than the PAPR of OFDM system.This paper carries out the research in the PAPR performance optimization algorithm and scheme of GFDM system.The specific research contents are as follows:In this thesis,the PAPR performance of GFDM system is analyzed deeply.By summarizing the research work of existing papers,the PAPR performance optimization of GFDM system is obtained,and the requirement of PAPR suppression algorithm and scheme for GFDM signal is given,where the complexity is not too high,without sacrificing BER performance or a big impact on the GFDM signal spectrum.Aiming at the high complexity of the existing PAPR suppression algorithm in GFDM,considering that different filters have their own phase-frequency characteristics,a new algorithm based on randomly assigning filters to reduce PAPR in GFDM(RAF-GFDM)is proposed.According to the feature of GFDM system that the filter structure could be selected flexibly,different filters are assigned to subcarriers randomly to adjust the phase congruency due to signals superimposition.Additionally,this method can suppress PAPR effectively without causing signal distortion.Theoretical analysis and simulation results demonstrate that the proposed algorithm does not bring additional calculation costs while reducing PAPR and almost has no effect on the bit error rate(BER),and it may provide a higher application value.Aiming at the problem of sacrificial bit error rate(BER)performance for PAPR suppression schemes of existing GFDM systems,three PAPR suppression schemes based on precoding matrices are presented in this thesis.Firstly,the relationship between the PAPR performance of GFDM system and the autocorrelation function between data signals transmitted in the same time slot is deduced theoretically.The distribution of PAPR can be improved by linear matrix transformation,The phase scrambling codes are introduced by linearly transforming the input complex data signal with three different precoding matrices,DFT-Precoding,WHT-Precoding and DHT-Precoding,and the instantaneous power of the GFDM signal is reduced by changing the correlation between the data symbols transmitted on different subcarriers but in the same slot to achieve the purpose of reducing the system PAPR.These schemes are linear processing,which suppress PAPR while ensuring that BER performance is not affected and there is also no significant effect on the spectrum of the GFDM signal.Theoretical analysis and simulation results also validate the performance and effectiveness of this scheme.