Research On PAPR Suppression And Channel Estimation In FBMC System

AS an air interface solution in 4G,Orthogonal Frequency Division Multiplexing(OFDM)technology can effectively resist multipath fading with low implementation complexity.However,OFDM has such defects as high requirements on carrier orthogonality,which makes it sensitive to frequency deviation,low spectral efficiency caused by CP introduction,and high out-of-band power radiation caused by unformed subcarriers.These problems make OFDM unable to meet the requirements of next generation communication.Filter Bank Multi Carrier(FBMC)is a multi-carrier transmission scheme with high spectral efficiency and relatively low complexity.FBMC can effectively resist ISI and ICI by shaping each subcarrier with a prototype filter with good time-frequency focusing characteristics at both transmitter and receiver.There is no requirement of CP in FBMC,so its spectral efficiency is higher.In addition,its out-of-band attenuation is rapid so that no additional out-of-band reduction is required.However,FBMC also has problems such as high peak-to-average Power Ratio(PAPR),overlap of adjacent symbols and inherent imaginary interference in the system caused by the orthogonality between sub-carriers,which can only be met in the real domain due to the prototype filter's properties.Therefore,PAPR reduction and channel estimation of FBMC systems are more complex than OFDM systems.The prototype filter in FBMC lasts for several symbol times,leading to overlapping signals of adjacent symbols in the time domain.This character makes PAPR reduction techniques in traditional OFDM unable to be directly applied to the FBMC system.This paper mainly focuses on the PAPR reduction methods of FBMC based on the Partial transmission Sequence(PTS).Firstly,Complementary Cumulative Density Function(CCDF)is defined to measure the probability of PAPR exceeding the given energy threshold.Then it is concluded that the core idea of PTS method is to find the best rotating phase vector to minimize the peak value of FBMC signal in time domain.For this reason,several PTS methods combining with multiple data blocks are studied to suppressed the PAPR of FBMC system.In the end,an improved PTS method is proposed to approach the theoretical optimal method with lower complexity.On the other hand,the classical channel estimation method based on pilots is easy to implement.It can estimate the channel response value by sending the known symbol at transmitter and recovering it at the receiver.This paper mainly studies the channel estimation method based on pilots.The pilot methods can be divided into discrete pilot and block pilot according to its own structural characteristics.In terms of Interference Approximation Method(IAM),this paper proposes a pilot structure based on EIAM-C.Under the premise of good estimated performance,proposed method greatly reduces the peak value of FBMC signals and thus improved channel estimation performance.In the aspect of channel estimation of mimo-fbmc system based on space multiplexing,this paper extends the method in a single antenna by taking the number of 2x2 transceiver antennas as an example.In order to solve the problem of overlapped pilot interference calculation region in Coded Auxiliary pilot(CAP)and other pilot coding regions,an iterative CAP method is proposed to effectively eliminate imaginary part interference of pilot.