Research On Large-scale MIMO-FBMC Information Detection Algorithm Based On Low Rank Tensor

In 5G(The Fifth Generation),the system performance of massive MIMO-FBMC(Massive Multiple-input Multiple-output-Filter Bank Multi-carrier)system is improved by the information detection.For example,we inspire to improve the system transmission rate,the system capacity,the number of system users,reduce the system error rate and so on.In the study for massive MIMO-FBMC system at domestic and foreign,researchers hope to be able to further improve the efficiency of the system diversity.And the low-rank tensor model reduces the dimension of high-dimensional signals into low-dimensional signals,it makes the signal have better sparseness and better diversity.In recent years for the research of massive MIMO-FBMC systems,the existing literature has applied the low-rank tensor model to massive MIMO-FBMC systems.In the context of multiuser and dynamic user scenarios,this paper uses the information detection and channel estimation based on the massive MIMO-FBMC system in the low rank tensor model to perform the following research:Firstly,this paper proposes a low-order tensor system model of massive MIMO-FBMC for low-order tensor models to solve the problems of massive MIMO-FBMC systems.The tensor model is decomposed,the high-dimensional received signals are decomposed.For low-order tensor models,ALS(Alternate Least Squares)and LM(Levenberg-Marquardt)are used to process the received signals.In this model,the tensor coding algorithm is further studied to map the user transmission symbols under different time-frequency space resources to realize antenna diversity of massive MIMO-FBMC.Through experimental simulation,the results verify that the low rank tensor model can effectively improve the antenna diversity and spectrum transmission efficiency of the massive MIMO-FBMC system.Then,in this paper,a high-density MIMO-FBMC system based on low rank tensor model is proposed for high-density user scenarios.For low-order tensor system models in high-density user scenarios,as the number of system users increases exponentially,the multi-user low-order tensor system model for massive MIMO-FBMC needs to be further studied,using ALS and LM respectively.The quantity algorithm processes the received signal of the base station antenna and effectively separates the low rank tensor information of different users.Throughexperimental simulation,this algorithm can effectively separate low-rank tensor information of different users in high-density user scenarios,and has significantly improved system transmission rate and system capacity.Finally,this paper proposes a massive MIMO-FBMC system information detection algorithm for dynamic user scenarios in massive MIMO-FBMC system.Using the low rank tensor system model,we can establish the dynamic users’ model of massive MIMO-FBMC system.And the feature space of the dynamic user,the three-dimensional channel space are studied.ALS and LM are used to process the receiving signal of base station.The receiving signal of the antenna effectively separates low-rank tensor information of multiple users in different motion states.Through experimental simulation,this algorithm can effectively separate the dynamic user information and reduce the system error rate.