Research On Channel Estimation And Channel Tracking In Spatial Modulation

People's demand for high speed data transmission and high frequency spectrum efficiency is the key factor driving the development of wireless communication technology in the future.In the fourth generation mobile communication technology,Multiple Input Multiple Output(MIMO)technology plays a very key role.However,Inner Channel Interference(ICI)inevitably exists in the application of MIMO technology.In recent years,Spatial Modulation(SM)is a technology proposed for wireless transmission using multi-antenna architecture,which reduces the complexity of the receiving end and completely avoids the problem of ICI in MIMO system.As a high-potential multi-input multi-output technology,it has become a research hotspot of wireless communication technology in the future.In the wireless communication system,the accuracy of receiving end's channel estimation affects the reliability of subsequent signal detection and demodulation.After the introduction of spatial modulation,there are few researches on channel estimation in this scheme.In addition,there is an increasing demand for wireless communication scenarios.In the case of high-speed terminal movement,channel changes will be very drastic,so it is urgent to study the channel tracking scheme under the time-varying channel model.The main research content of this paper is to study a channel estimation method based on neural network and a channel tracking method based on kalman filter under the framework of spatial modulation system.The main contents and achievements of this paper include:(1)research on channel estimation in spatial modulation system: several common channel estimation methods in spatial modulation system are analyzed.This paper introduces the basic principle of neural network,designs the neural network structure and the channel estimation scheme suitable for spatial modulation system.This scheme transfers the complex computation of the receiver to the cloud and reduces the complexity of channel estimation in the actual communication process.A two-layer feedforward neural network structure was designed,and the scaled conjugate gradient(SCG)algorithm was used to train the neural network.The simulation and verification of the proposed scheme were carried out,and the results showed that the bit error rate performance of the scheme was better than that of LS algorithm and MMSE algorithm.In addition,the effects of the training sequence length and the number of hidden layer neurons in neural network on the performance of the channel estimation method are studied and verified by simulation.The results show that the longer the training sequence is,the better the performance of the algorithm.The number of hidden layer neurons is related to the complexity of the system and should be at least larger than the number of antennas at the receiving end.(2)Research on Channel Tracking in Spatial Modulation System.It mainly includes data-aided channel tracking method and Kalman filter algorithm applied to channel tracking of space modulation system to improve the drawbacks of data-aided algorithm.The necessity of channel tracking in time-varying fading channels is analyzed,and the data-aided tracking methods commonly used in channel tracking in spatial modulation systems are introduced.Aiming at the problems in data-aided channel tracking,Kalman filter is used to improve channel tracking.The scheme of Kalman filter channel tracking method is designed for spatial modulation system.The steps of the algorithm are introduced in detail and the simulation results show that the bit error rate performance of the algorithm is better than that of DACT method under low SNR.The effects of pilot ratio and channel variation parameters on channel tracking are analyzed according to the simulation results.The higher pilot ratio,the lower channel variation rate is,channel tracking effect is better.