| The low-orbit broadband Internet satellite communication integrated with 5G faces a series of challenges brought by high-frequency band,large bandwidth and high-order modulation of signals.Satellite communication channel modeling and transmission performance optimization have become current research hotspots.The main research contents and innovations of this paper are as follows:1.This paper studies the nonlinear modeling technology of on-board filters in low-orbit broadband Internet satellite channels.Aiming at the problems of many fitting parameters and complex structure in the existing spaceborne filter model modeling,an improved spaceborne filter characteristic model based on multi-dimensional kernel function was proposed.Through the linear and nonlinear combination of the kernel function,the model can effectively characterize the fluctuations in the passband,transition band suppression and nonlinear group delay characteristics of the spaceborne filter,so as to realize the accurate characterization of the nonlinear model of the spaceborne filter..2.This paper studies the satellite channel equalization technology with strong nonlinear onboard filter distortion in low-orbit broadband satellite communication.Aiming at the problem that the adaptive equalization algorithm in low-orbit broadband satellite communication does not deal with the group delay distortion when compensating for the nonlinear distortion of the channel,the system bit error rate increases and the convergence speed is slow.Adaptive step size equalization algorithm for phase weights.The algorithm obtains the eigenvalues from the nonlinear distortion model of the on-board filter,and initializes the weights of the equalizer as a priori information,so that the weights of the equalizer can be close to the optimal solution,reduce the solution time,and improve the convergence speed of the algorithm;at the same time,the algorithm constructs An adaptive variable step size factor that can compensate for phase distortion is proposed.In the algorithm iteration process,the step size factor is set to be jointly controlled by the signal error and the time domain phase weight factor,which can improve the nonlinear group delay characteristics of the spaceborne filter.The resulting phase distortion is compensated,which reduces the demodulation threshold and bit error rate of the entire system.Experimental results show that the complexity of the algorithm is low,the overall processing time of the link is reduced,and compared with the traditional adaptive LMS,adaptive MMSE and other equalization algorithms,the bit error rate is reduced to 10-6,and the signalto-noise ratio performance is improved by 0.4 to 2.2dB.3.At the end of this paper,the hardware implementation of the model is also studied.According to the multi-dimensional kernel function spaceborne filter model proposed in this paper,the hardware design is completed for the spaceborne filter function module of the low-orbit broadband Internet satellite channel simulator based on the NI Lab View platform.and implementation,and the filter function is verified through the built verification platform. |
