Research On Demodulation Algorithm Of OTFS Signal In Fast Time-Varying Channels

OTFS(Orthogonal Time Frequency Space)is a new type of modem technology with wide application prospects in high-speed rail,satellite transmission,aviation field,etc.There are a large number of atmospheric attenuation materials in the high-speed mobile transmission environment,which requires real-time communication system to have the advantages of anti-multipath,anti-interference,large amount of transmitted information and high reliability.The channel estimation algorithm based on pilot assistance can ensure the performance of communication system while saving transmission resources,but the traditional channel estimation algorithm is difficult to track the rapid change of channel,which seriously affects the estimation accuracy of channel matrix.The inter-subcarrier interference due to transmission characteristics cannot be completely eliminated in the signal detection at the receiver side,which eventually leads to the degradation of demodulated signal accuracy.The paper investigates the channel estimation and equalization techniques that can be applied to fast varying channels,including fast varying channel characteristics,OTFS modem techniques,channel matrix theory,channel estimation algorithms,basis expansion model(BEM)and iterative interference cancellation algorithms.The main innovations of the paper are as follows:(1)In terms of fast varying channel characteristics,OTFS modem techniques and channel matrix theories,the characteristics of fast varying channels are introduced,and the impact of these characteristics on the received signal is studied,and the relevant characteristic parameters of typical fast varying channels are analyzed.The OTFS signal generation method is improved by OTES modem technology through theoretical derivation of OTFS signal generation methods and analysis of the conversion relationship between OTFS signals and OFDM(Orthogonal Frequency Division Multiplexing)signals.The channel matrix theory is derived for the frequency-domain channel matrix and the timedelayed Doppler-domain matrix,and the energy distribution of the two channel matrices in two-dimensional space is simulated experimentally,in addition the characteristics of the two channel matrices are analyzed.(2)In terms of channel estimation algorithm and BEM,an improved BEM-based channel estimation algorithm is proposed based on fast varying channel characteristics.The conventional channel estimation algorithm has a computationally intensive matrix inversion operation and relies on the interpolation algorithm,which leads to low accuracy of the channel matrix estimation and makes it difficult to eliminate the effect of inter-subcarrier interference.The improved algorithm can generate the basis function matrix directly in the process of estimating the time-domain channel impulse response,and can improve the accuracy of the basis coefficient estimation by LMMSE algorithm,which effectively reduces the computational effort.(3)In terms of channel estimation algorithm and iterative interference cancellation algorithm,an improved channel estimation algorithm by means of piecewise scattering guide frequency is proposed,which can effectively save transmission resources and use pregenerated channel parameters for estimation to reduce the computational effort and improve the computational efficiency.To further improve the channel matrix accuracy,a lowcomplexity iterative interference cancellation algorithm is proposed to eliminate intersubcarrier interference,which has a simple structure and requires less iteration.The above improvement schemes are simulated and verified in OTFS and OFDM systems,and it is proved that the bit error rate(BER)is lower comparing with the traditional schemes,about 3db.