Research On Fast Channel Estimation And Information Detection Method Based On Tensor

The performance of wireless communication system is greatly affected by wireless channel,and whether accurate channel and signal parameter information can be obtained is an important index to improve transmission reliability and effectiveness,so channel estimation is one of the important technologies in wireless communication system.Wireless channel often has a lot of randomness,and the signal in the transmission process generally contains different dimensions,such as space,time,frequency and so on,which increases the difficulty of designing the channel estimation method.Such signals with different dimensions can be characterized as tensors(also called multi-way arrays),and the tensor model can be used to signal detection and parameter estimation without channel state information(CSI)in blind signal processing technology,which can improve the spectrum efficiency.However,the traditional tensor fitting algorithm is alternating least squares(ALS)fitting algorithm.The algorithm has many invalid iterations and is prone to fall into local loops,which requires a large number of iterations to achieve convergence.In addition,unmanned aerial vehicles(UAVs)are widely used in the field of wireless communication,especially in harsh battlefield environments,UAVs have been acquired at high altitudes to take information and complete the task of information transmission of massive data such as image and video.It is very important to recover the detect signal accurately,but it is difficult for the UAV to detect information of two or more users.In order to solve the above problems,the main works in this paper can be divided into the following aspects.On the one hand,an improved alternating least squares(IALS)algorithm is proposed in this paper to reduce the number of iterations and the computational complexity of traditional iterative algorithms.The main idea of the improvement in the proposed algorithm is to use the parameter estimation value of the previous stageas the initial value of the first iteration.The step factors and the linear search direction are introduced to get the initial value of each iteration in the iterative process,which can improve the accuracy of the initial value of the iteration and accelerate the convergence.The simulation shows that,the proposed algorithm saves a lot of multiplication operations compared with the traditional ALS algorithm.On the other hand,in the face of the challenge of how the UAV can detect multi-user information in the wireless communication system,the generalized tensor contraction(GTC)operation is used in this paper to represent the slices multiplication,which can obtain a clear description of the data tensor using the generalized contraction of two tensors.The parallel factor decomposition(PARAFAC)model is used to get the model unfoldings,which makes the least squares(LS)algorithm more flexible and convenient,and can also be extended to higher-order model simply and clearly in multi-user scenarios.The simulation shows that this method is suitable for blind information detection without sacrificing algorithm performance in multiple-input multiple-output(MIMO)communication systems.