Research On FBMC-OQAM Technology In Doubly Selective Fading Channel

Compared with the traditional Orthogonal Frequency Division Multiplexing(OFDM)system,the Filter Bank Multi Carrier with Off-set Quadrature Amplitude Modulation(FBMC-OQAM)system are getting more and more attention due to its higher spectrum utilization rate and effective resistance to frequency offset.In the doubly selective fading channel,the inherent imaginary interference of FBMC-OQAM system is crosstalked into complex interference,resulting in Inter Symbol Interference(ISI)and Inter Carrier Interference(ICI),and the performance of FBMC-OQAM system is sharply deteriorated.In this paper,the signal detection method of FBMC-OQAM system is studied,and the technique of improving the performance of FBMC-OQAM system is explored by signal processing means in the doubly selective fading channel.The main work and contributions are as follows:1.An interference cancellation method based on oversampled FBMC-OQAM system is proposed in doubly selective fading channel.This scheme can change the distribution of interference existing in the received signal by oversampling in the time domain,and then a reasonable signal to interference noise ratio(SINR)sampling point can be obtained,further more,the detection algorithm is used to minimize or eliminate interference,and ultimately the ICI and ISI are suppressed.In addition,considering that the ICI of the FBMC-OQAM system mainly comes from adjacent subcarriers,a low complexity receiver for the oversampled signal is proposed,which can effectively reduce the complexity of the detection operation.The simulation results show that the proposed scheme can vastly eliminate the ISI and ICI in doubly selective fading channel.2.A conjugate transmission scheme based on oversampled FBMC-OQAM system is proposed.In this system,the bits are modulated into a pair of conjugate complex symbols by conjugate modulation scheme at the transmitter and these conjugate complex symbols is oversampled at the receiver.Then,similarly,the detection algorithm is used to minimize or eliminate interference,and the detected conjugate complex symbols are combined in a linear manner.What's more,the ISI and ICI caused by the doubly selective fading channel and the inherent imaginary interference of the FBMC-OQAM system are reduced.Moreover,a sub-block MMSE receiver for oversampled conjugate signal is proposed,and the receiver can reduce the computational complexity of signal detection.Simulation results show that,when BER is-410,the performance of the proposed scheme is improved by about 9d B and 11 d B respectively compared with the oversampled FBMC-OQAM scheme at relative movement speeds of 300 km/h and 600 km/h.3.A signal detection scheme based on deep learning classification technique for the FBMC-OQAM conjugate system is proposed.First of all,the constellation point redundant real data set was produced by the author of this paper,then the neural network is trained in the way of constellation point classification by using the data set,and save the optimal model.Further more,the received conjugate complex symbols of the FBMC-OQAM system is detected by the optimal neural network model,and the original digital symbol of the transmitter is directly restored.In the proposed detection scheme,the neural network model can learn all the operating principles of OQAM post-processing and conjugate signal combining,and then realize the interference cancellation and obtain linear combining diversity gain.The test results show that the proposed scheme can improve the comprehensive detection performance of the system.