基于张量分解的隧道信道估计算法OA

A fast channel estimation algorithm based on tensor decomposition is proposed to realize timely acquisition of channel state information in wireless communication systems assisted by reconfigurable intelligent surfaces(RISs)in tunnel environments.Firstly,a cascade channel matrix is constructed by a small number of passive reflector units,and a third-order tensor model is built.Then,the third-order tensor is compressed by high-order singular value decomposition(HOSVD),and the cascade matrix estimation problem is transformed into a tensor decomposition problem.Finally,the enhanced plane search algorithm is used to dynamically update the channel matrix information,and the optimal relaxation factor is calculated at the same time to accelerate convergence and improve estimation accuracy.The simulations verified the channel estimation characteristics.The introduction of intelligent surface improves the channel estimation performance of tunnel wireless communication.In comparison with the traditional algorithms,by changing the number of antennas and the number of reflector units,the proposed algorithm has lower normalized mean square error(NMSE),can converge to the order of 10-6,and has a faster running speed.In addition,the variable factors in different tunnel environments were simulated,which reflected the applicability of the algorithm in complex environments.