Research On Channel Estimation Technologies Of OTFS System Under Time-Frequency Dual Selective Channel

After the development of mobile communications to 5G,high-speed moving objects,higher frequency carriers,and larger antenna arrays have all made some algorithms proposed to overcome dual-selection fading problem face great challenges.As a new modulation technology,Orthogonal Time Frequency Space(OTFS)is proposed in 2017 and able to work in various usage scenarios defined by 5G.The existing research on the receiver algorithm of this technology mainly focuses on channel estimation and equalization,and the equalization is the most widely studied,including linear equalization and nonlinear equalization.A number of studies have pointed out that accurate channel estimation is essential to improve equalization performance.However,there are relatively few studies on OTFS channel estimation,especially the pilot-aided channel estimation technology in the delay-doppler domain.At present,the technology of channel estimation by inserting pilots in the delay-doppler domain is mainly based on the amplitude of a single pilot impulse to obtain the tap coefficient of the transmission channel.In order to solve the problem that the performance of single pilot impulse deteriorates sharply under low SNR,the research on the channel estimation technology based on multiple pilot impulses,such as the multi-impulse threshold decision method based on prior information,has been carried out.However,this kind of method still has some room for improvement.Based on this,two improved channel estimation algorithms based on multiple pilot impulses are proposed in this paper:1.Considering that the existing multi-impulse channel estimation method does not make full use of the amplitude information of each pilot impulse,a multi-impulse incoherent channel estimation method is proposed in this paper.This method accumulates the amplitudes of multiple pilot impulses before making a threshold judgment,and uses the sum of the amplitudes as the channel fading coefficient.This not only improves the detection probability of channel taps,but also improves the estimation accuracy of the tap gain.Simulation results show that the proposed incoherent channel estimation method has significant improvements in both Normalized Mean Square Error(NMSE)performance andBit Error Rate(BER)performance compared with the existing method.When the SNR of pilot is 0dB-12 dB,the NMSE performance is improved by 3dB.When the SNR of pilot is0 dB and the SNR of data is 0dB-12 dB,the larger the SNR of data is,the more obvious the BER performance will be improved.2.In view of the performance improvements of the proposed multi-impulse incoherent channel estimation method,another new algorithm called multi-impulse coherent channel estimation method is proposed.This method calculates the amplitude after coherent accumulation of multiple pilot impulses,and then makes a threshold judgment and uses the amplitude as the channel fading coefficient.According to the principle of coherent operation,this method can obtain better anti-noise performance than incoherent method.Simulation results show that the proposed coherent channel estimation method significantly improves the NMSE performance of the estimated results and the BER performance of the system compared with the incoherent channel estimation method in the low SNR environment,while there is no significant difference between the two methods in the high SNR environment.When the SNR of pilot is-6dB-0dB,the smaller the value is,the more obvious the NMSE performance is improved.When the SNR of pilot is-5dB and the SNR of data is0 dB-12 dB,the BER performance is significantly improved.