Fast Time-varying Channel Estimation Algorithm Based On ICI Cancellation In High Speed Railway Scene

The rapid development of the high-speed railway has been driven by the rapid advances of the economy and society.As China enters the 4G times,the application of broadband communication of Long Term Evolution for Railways(LTE-R)to the next generation high-speed railway communication system is the trend of development.However,the application of LTE-R faces many challenges,in which how to accurately obtain fast time-varying channel state information is a very important point.Therefore,taking the fast time-varying channel estimation method of high-speed railway communication system as the study object,this paper focuses on the characteristics of high-speed railway communication channel and the optimization method of its system performance.In order to provide strong support for the subsequent research of channel characteristics,several common channel measurement methods are analyzed firstly.Under the high-speed railway communication environment,no additional measurement information is allowed,so the measurement method based on the existing communication network is more applicable.On this basis,the small-scale fading characteristics of the high-speed railway communication system are studied,namely time-selective fading and frequency-selective fading.There are many parameters to be estimated in the communication channel of high-speed railway,and the Inter-carrier Interference(ICI)will cause serious damage to the Orthogonal Frequency Division Multiplexing(OFDM)system,and also affect the accuracy of channel estimation.Therefore,according to the characteristics of high speed railway communication channel,the corresponding channel estimation method is improved.Considering the interaction between ICI and channel estimation,based on the traditional Discrete Karhunen-Loeve-BEM(DKL-BEM)algorithm,a twice channel estimation algorithm based on ICI cancellation is proposed,which is called the twice channel estimation algorithm.Both the twice channel estimation algorithm and the traditional DKL-BEM algorithm use DKL-BEM to model the channel,and reduce the estimated parameters of the channel.In addition,in order to avoid error propagation and difficult convergence from the iteration,using simplified parallel interference cancellation algorithm to eliminate ICI after the first channel estimation,and then the received signal after the elimination of ICI is used for the second channel estimation to reduce the effect of ICI on channel estimation.The simulation results show that the proposed twice channel estimation algorithm can significantly improve the channel estimation accuracy under the condition of small SNR,but the performance under the condition of large SNR is inferior to the traditional DKL-BEM algorithm.Analysis shows that when the SNR exceeds a certain threshold,the twice channel estimation algorithm performance is inferior to the traditional DKL-BEM algorithm,which is reflected in the performance curve,that is,there is a cross point of the performance curves of the two algorithms;further study found that the scatter cross point varies with speed has a certain distribution.Based on the comprehensive consideration of the accuracy and algorithm complexity of the fitting curve,the quadratic polynomial curve fitting is performed on the movement trajectory of the cross point,and the trajectory of the cross point changing with the speed is given.The theoretical value calculated from the fitting curve is compared with the actual value obtained in the simulation experiment.The maximum relative error is 3.57%,which verifies the accuracy and effectiveness of the fitting curve.Based on the above,a dual-mode switching channel estimation algorithm based on SNR decision is proposed.The cross point is used as the boundary,and the twice channel estimation algorithm is adopted under the condition of small SNR;the traditional DKL-BEM algorithm is adopted under the condition of large SNR.The algorithm can adaptively implement the switching of the two algorithms,so it can be referred to as an adaptive channel estimation algorithm.The simulation results show that the adaptive channel estimation algorithm not only retains the advantages of the twice channel estimation algorithm under the condition of small SNR,but also abandons its disadvantages under the condition of large SNR.The adaptability and effectiveness of the channel estimation algorithm are improved.Twice channel estimation algorithm and adaptive estimation algorithm all based on simplified PIC algorithm to cancel ICI,while avoiding the error propagation and difficult convergence resulting from the iteration,but it also affects the effect of ICI cancellation.In other words,residual errors still exist after ICI cancellation,which affects the accuracy of channel estimation.In order to improve the accuracy of channel estimation as a whole and reduce the influence of the residual error,a twice fast time-varying channel estimation algorithm using weighted coefficients to cancel ICI is proposed.According to the current moving speed and the SNR,the algorithm gives the appropriate weighted coefficient value on the basis of the minimum mean square error criterion,and improves the simplified PIC algorithm.Simulation results show that compared with the traditional DKL-BEM algorithm,the proposed algorithm can significantly improve the channel estimation accuracy under different speeds and different SNR conditions.