Research On ICI Mitigation For Differential Underwater Acoustic OFDM Communications

The construction of the space-air-ground-sea integrated network motivates the need for the high-data-rate underwater acoustic(UWA)communication.Subject to the performance bottleneck,i.e.,the communication distance multiplied by the data rate is no larger than 100km·bps,the existing UWA communication technology cannot keep up with its application in engineering.Driven by this fact,the development of theories and methods for high-data-rate UWA communication is required.Differential Orthogonal Frequency Division Multiplexing(OFDM)technology is likely to effectively avoid strong dependence on channel information,thus significantly reducing the time-frequency resources occupied by channel estimation,and providing a potential solution to the above bottleneck.However,due to the inherently large Doppler spread(equivalent to the carrier spacing)of the UWA channel,the detection performance of differential UWA-OFDM systems is subject to severe inter-carrier interference(ICI).The existing ICI mitigation methods,based on partial or fractional fast Fourier transform,have limited weight compensation accuracy and poor adaptability to dynamic Doppler in the UWA channel.To this end,we devote to research on ICI mitigation methods with high compensation accuracy and strong adaptability for differential UWA-OFDM communications to deal with limited ICI mitigation capabilities in the prior art and difficulty in improving detection accuracy.Our primary research contents and results are summarized as follows:1.In view of the limited weight compensation accuracy of the classical partial fast Fourier transform(P-FFT),we propose a time-domain ICI mitigation method based on weight prediction-correction.Specifically,inspired by the influence of the weight compensation accuracy on the P-FFT interference suppression performance,we establish a weight optimization model to minimize mean-square error(MSE).In this manner,a weight estimation-correction algorithm is designed to obtain the optimal compensation coefficient across adjacent carrier interference.Theoretical analysis shows that the proposed algorithm,adopting the eigendecomposition in the prediction stage and the adaptive iteration in the correction stage,combines the stability of non-adaptive methods with the simplicity of adaptive explicit iteration to improve the accuracy of weight compensation after filtering and segmenting the disturbed signal.Simulation results show that,compared with the existing P-FFT,MSE can be reduced by 62.85%and 91.54%under high Doppler factor(1.5×10^-4)and large carrier number(2048),respectively.2.To cope with the poor adaptability of the classical fractional fast Fourier transform(F-FFT)method to Doppler shift,we propose a frequency-domain ICI mitigation method adapted to dynamic Doppler spread.Specifically,based on the analysis of the existing frequency domain compensation structure,we propose the concept of fiducial frequency offset.In view of this,we further verify the influence of this parameter on the MSE performance of the F-FFT,and obtain the analytical relationship between the two.With this in mind,an adaptive fiducial frequency offset estimation algorithm is designed,which transforms the frequency shift compensation interval from a fixed spacing to a fractional multiple of the fiducial frequency offset adapted to the dynamic Doppler spread.Furthermore,considering weight compensation accuracy,we propose a frequency-domain ICI mitigation method that combines adaptive fiducial frequency offset estimation and weight estimation-correction.Simulation results show that,compared with the existing F-FFT,the proposed method can obtain higher accuracy of ICI mitigation(MSE can be reduced by 51.81%-99.46%within the range of 2⁹-2¹¹ carriers).Through the above research,we deeply explore the ICI mitigation technology for differential UWA-OFDM communications from the perspective of time domain and frequency domain,respectively.The proposed methods exhibit significant interference mitigation ability under large carrier number and high Doppler factor,thus expanding application prospects of differential UWA-OFDM communications.