Channel Estimation Of MIMO-OFDM Communication System Based On Compressed Sensing

In order to meet people’s increasing demand for wireless network communication capacity and transmission speed,the number of antennas has been expanded to dozens of times that of the traditional mode,and more antennas have brought more complex channel models.For multiple-input multiple-output orthogonal frequency-division multiplexing systems,accurate channel information is an important guarantee for high-quality communication.However,the traditional pilot-based channel estimation method has too high resource consumption in the frequency domain,and the channel cannot be effectively restored under limited pilot carrier conditions.Through the sparse characteristics of channels,it is found that the application of compressed sensing technology to channel estimation can not only reduce the number of pilots,improve spectral efficiency,but also improve the reconstruction accuracy of channels.At present,the use of compressed sensing technology for channel estimation has become a research hotspot.In this paper,the following research work is mainly done on the MIMO-OFDM channel estimation problem based on compressed sensing:Firstly,the channel estimation problem of MIMO-OFDM is analyzed,and a channel estimation model based on compressed sensing method is established to facilitate subsequent work.Then,aiming at the problems of non-optimal pilot optimization criterion and local solution of search algorithm in the existing pilot design method,the observation matrix characteristics required for pilot design are theoretically derived,and then a high-power sum and high-power variance criterion is proposed to replace the existing pilot optimization criterion.Complete the performance analysis and proof of the high-power sum and high-power variance criteria.On this basis,a joint pilot design algorithm based on tree structure is proposed.The simulation results show that the pilot designed by this method has the best channel estimation performance under multiple conditions.Finally,the existing greedy algorithms are analyzed,and in view of the problems of insufficient calculation accuracy and excessive number of iterations of the current algorithms,a piecewise adaptive backtracking orthogonal least squares algorithm is proposed by adopting the idea of segmented estimation and adaptive step size.Wireless communication simulation is carried out under multiple comparison conditions,and experimental simulation shows that the algorithm has less time consumption,higher recovery accuracy and robustness than the traditional algorithm.