Research On Compressed Sensing And Simulation Of Signal Acquisition In Ground Penetrating Radar

Due to the difference in the dielectric properties between the different underground materials, the ground penetrating radar can respond to the underground by exciting and receiving electromagnetic waves. At the signal acquisition side, in order to follow the Nyquist sampling law, and thus the amount of collected data is too large , Making the radar equipment in the development of a lot of difficulties. As a new signal acquisition and imaging technique,compressed sensing can restore the original signal by collecting data that is much lower than that of Nyquist in the condition that the signal satisfies its sparseness or sparsely in the transform domain. The storage pressure of the equipment is of practical significance.This paper studies the sampling law and the mechanism of digital signal acquisition of traditional instrument and equipment. The sparse characteristics of ground penetrating radar signals are discussed. The characteristics of the underdetermined linear system and the sparsity of the solution of the underdetermined linear system are improved by p norm.Compressed sensing is actually in the signal acquisition at the same time to achieve the data compression, is a dimension reduction process, rather than in the source code sense of compression. It is mainly composed of the sparse representation of the signal, the collection and recovery of the signal. This paper makes a comparative study on the compensation algorithm of compression perception (orthogonal matching pursuit and base pursuit) in recovering the signal, the correct probability of support set, and the computational complexity.In this paper,a simulation model with underground target is established,which is discretized to the underground space. At the same time,the radar signal is associated with the underground target body by creating a dictionary,and the signal is collected by the low-dimensional Gaussian random matrix. Simulation experiments show that compression sensing can greatly reduce the amount of data collected and achieve the reconstructed imaging of the underlying target.