Complex RCS data interpolation method based on compressed sensing

Abstract

The invention relates to a complex RCS data interpolation method based on compressed sensing, and belongs to the technical field of comprehensive experimental testing. A specific implementation mannerof the method comprises: constructing a complex domain reconstruction model according to the RCS measurement data; solving the complex domain reconstruction model, acquiring a sparse distribution ofthe RCS data; and reconstructing the target frequency domain data based on the sparse distribution to implement complex RCS data interpolation. Compared with the conventional method for simply diagonalizing the complex matrix, the invention has rapid processing speed and high precision without reducing or increasing the original constraint range, and is capable of effectively shortening the measurement time to improve the RCS measurement efficiency.

Description

technical field [0001] The invention relates to the technical field of comprehensive testing and testing, in particular to a complex RCS data interpolation method based on compressed sensing. Background technique [0002] The traditional interpolation algorithms mainly include spline interpolation and Fourier transform method. Spline interpolation is a real number method and a pure interpolation algorithm. For darkroom RCS measurement data, it lacks the support of physical concepts and physical meanings. The Fourier transform method has two defects. One is that it is limited by the Nyquist sampling theory, so that the target expansion is not blurred in the range and azimuth directions, and the amount of sampled data cannot be reduced. The second is that the Fourier transform method is an overall algorithm. Even if the target signal is only in a small range of the unambiguous window, the overall transformation must be carried out, and the algorithm efficiency is low. However...

AI Technical Summary

Problems solved by technology

A common practice for complex field data is to convert the complex matrix to diagonal real numbers. The problem is that the original circular constraint space for the complex field has become a square constraint space, which is equivalent to increasing the scope of the constraint space, which will inevitably increase the noise of the algorithm.

A common practice for two-dimensional matrix data is to use the matrix vectorization method of stacking rows and columns. The disadvantage is that the dimension of the compressed sensing measurement matrix is ​​greatly increased by a geometric order, which greatly increases the computational memory and time consumed by the algorithm.

Images