Rapid MUSIC spectrum decomposition method, device and equipment for large-scale antenna

Abstract

The invention discloses a rapid MUSIC spectrum decomposition method, device and equipment for a large-scale antenna, and the method comprises the steps: receiving a signal X, and estimating a high-dimensional autocorrelation matrix R according to the signal X; performing skeleton extraction on the high-dimensional autocorrelation matrix R to obtain a low-dimensional representation matrix C; calculating to obtain a low-rank matrix Y and obtaining a low-dimensional approximate decomposition CY of the high-dimensional autocorrelation matrix R; obtaining an SVD approximate decomposition (shown inthe description) of the high-dimensional autocorrelation matrix R is obtained by conducting SVD decomposition on the low-dimensional approximate decomposition CY; constructing a signal space K by means of the SVD approximate decomposition (shown in the description) of the high-dimensional autocorrelation matrix R, estimating a spatial spectrum P (theta) by means of the signal space K, and conducting target signal detection and estimation according to the spatial spectrum P (theta). According to the invention, while estimating the MUSIC spatial spectrum accurately, the calculation complexity ofSVD is reduced from cubic growth to square and even linear growth rate, high-precision and low-complexity MUSIC spatial spectrum estimation is realized, matrix norm minimization is used as an optimization criterion, and the approximate error precision of a high-dimensional autocorrelation matrix is ensured.

Description

technical field [0001] The invention relates to the field of information, in particular to a fast MUSIC spectrum decomposition method, device and equipment for large-scale antennas. Background technique [0002] While large-scale antenna arrays increase the estimated signal-to-noise ratio and spatial resolution, the required computational complexity and processing delay will also become unbearable. In the implementation of millimeter-wave radar target detection and phased array radar beamforming, etc. It becomes difficult to apply in typical application scenarios. The core idea of ​​the MUSIC method is to divide the signal or noise subspace based on eigenvalue decomposition or singular value decomposition, and estimate the spatial position of the target through the signal or noise subspace. When the number of antennas is large, the computational complexity is extremely high; even if the processing process uses High-performance CPUs also require a long processing delay, whic...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the object of the present invention is to propose a fast MUSIC spectrum decomposition method and device for large-scale antennas, to solve the defects of high processing delay, low precision and high computational complexity in the prior art

Images