Research On Coal Mine Underground Channel Estimation Technology Based On Compressed Sensing

Efficient and reliable communication system is an important guarantee to achieve safe production in coal mines,and channel estimation technology is one of the key links,and its estimated performance will directly affect the communication quality of the entire system.However,the complex and changeable environment downhole greatly increases the impact on the wireless channel,resulting in the channel generally showing multipath fading characteristics.In order to effectively combat multipath interference and improve system performance,this paper is based on Orthogonal Frequency Division Multiplexing?Orthogonal Frequency Division Multiplexing,OFDM?technology to complete the signal transmission.At the same time,the rise and development of compressed sensing theory has opened up a whole new way of thinking in the field of channel estimation research.This article is built on the underground OFDM communication system,focusing on the application of compressed sensing reconstruction algorithm in channel estimation.The classical compressed sensing reconstruction algorithm has certain advantages in sparse channel estimation,but most of these algorithms use the inner product criterion measurement method in the atom screening stage,resulting in any two similar atoms in the sensor matrix when expanding the support set Will interfere with the matching of the signal residuals,resulting in the loss of part of the original signal.To this end,this paper proposes an improved strategy based on the generalized orthogonal matching tracking?GOMP?algorithm with faster convergence speed,that is,the Dice coefficient matching criterion is used when selecting the support set atoms,which is effectively avoided in this way.The problem that the optimal atom cannot be selected from the redundant dictionary due to the inner product criterion.Experiments show that under the condition that the signal-to-noise ratio is 40 and the step is set to3,the system bit error rate corresponding to the GOMP algorithm is0.8557?10^-2,and the D_GOMP algorithm reduces the bit error rate to0.2356?10^-2,which shows that the performance of the proposed algorithm in channel estimation has been greatly improved.However,one of the prior conditions of the above algorithm is that the sparsity is known,but the sparsity in the actual system is often difficult to obtain accurately,so this paper further studies the blind sparsity compressed sensing algorithm.Based on the Sparsity Adaptive Matching Pursuit?SAMP?algorithm,this paper introduces a new variable step size iteration idea,and selectively updates the step size according to the relationship between adjacent signal energies,effectively Improves the underestimation and overestimation problems of SAMP algorithm.At the same time,the algorithm combines the idea of atomic"weak"selection and the Dice coefficient matching criterion,thereby achieving further optimization in the atom selection stage.It is verified by experiments that the algorithm proposed in this paper has better estimation performance than SAMP algorithm in channel estimation with unknown sparsity.