Dynamic Sparse Channel Joint Estimation Of Multi-dimensional Based On Compressed Sensing

In wireless communication system,wireless channel characteristics must be fully understood in order to improve the reliability of wireless communications.Due to the influence of obstacles,the signals at different phases of the receiving end are superposed which causes multipath fading.This fading phenomenon seriously deteriorates the quality of received signal and affects communication reliability.Over the past decade,Orthogonal Frequency Division Multiple(OFDM)technology has become more and more widely used in wireless broadband communication systems.Under high-speed mobile environment,the time-domain fading caused by Doppler effect will destroy the orthogonality between sub-carriers and cause inter-carrier interference.For the time-frequency dual-channel selection,a large number of pilot sub-carriers are required due to the estimation of large channel coefficients.Making use of the sparsity of wireless channel,it will reduce the number of pilot subcarriers by using compressed sensing(CS)technology.However,under the environment of high-speed railway,the driving environment changes rapidly which results in rapid change of the channel.That is,some non-zero channel taps may become zero at the next slot and some new non-zero channel taps may appear at the same slot,so the channel sparsity changes dynamically.As a result,the traditional channel estimation methods based on CS or distributed CS(DCS)are not satisfactory for the channel reconstruction.Aiming at the channel estimation method of dynamic sparse channel,this thesis combines the joint sparsity and dynamic characteristics of channel coefficients,and achieves some results and conclusions mainly from the following two aspects:In this paper,aiming at the problem that the performance of channel estimation algorithm affected by the dynamic change over time of sparse channel in multi-symbol scene,this paper proposes a channel estimation algorithm based on CS and combining with the joint sparsity and time correlation of channels.Firstly,the system model under multi-symbol scenario is established by using complex exponent base extension model(CE-BEM)and DCS technology.Then,the corresponding channel is time-correlated which is based to determine the non-zero channel taps newly generated or disappeared at the current time so as to obtain the actual channel sparsity;Finally,the joint sparse channel coefficient vector between different BEM orders are restored jointly.In this paper,for the problem of dynamic sparse channel estimation in multi-symbol multi-antenna scenario,an estimation algorithm suitable for large-scale channel coefficients is designed.Firstly,the block sparsity of channel coefficients in MIMO system is analyzed,and then the channel coefficient vector is rearranged according to the common joint sparsity characteristics of channel coefficients of dynamic sparse channels in different transmit-receive antenna pairs and different BEM orders.Since the channel corresponding to different OFDM symbol periods has a dynamic characteristic,the time correlation of the channel is used to determine the dynamic change of the non-zero channel tap at the current moment,so as to determine the sparseness of the channel.Finally,the channel coefficients are used to jointly recover the channel information in different transmit-receive antenna pairs and the block sparsity between different order CE-BEM coefficients.Finally,this article summarizes the above research work and puts forward to some future research directions.