Generalized covariance multi-signal classification algorithm based on score function

Abstract

The invention provides a generalized covariance multi-signal classification algorithm based on a score function, and belongs to the technical field of communication. The method mainly comprises the following steps: 1, selecting proper parameters, and calculating and obtaining a score function; 2, acquiring a noise-mixed signal output by the antenna array, and constructing a vector of the signal; 3, calculating and obtaining a generalized covariance matrix based on a score function; 4, calculating and obtaining the estimation of the generalized covariance matrix based on the score function. 5,carrying out eigenvalue decomposition on the generalized covariance matrix based on the score function, carrying out calculation, and obtaining estimation of a noise subspace. 6, calculating and obtaining the estimation of the spatial spectrum according to the estimation of the noise subspace. and 7, estimating the direction of arrival of the radio signal through the spectral peak position. Experimental results show that the algorithm can accurately realize direction-of-arrival estimation of radio signals under the condition of impulse noise.

Description

technical field [0001] The invention belongs to the technical field of communication, relates to the estimation of the direction of arrival of radio signals, in particular to a multi-signal classification (SCORE-MUSIC) algorithm based on the generalized covariance of the score function. Background technique [0002] With the continuous development of communication technology, people are increasingly relying on radio communication in their daily study, work and life. At the same time, radio communication is also facing an increasingly complex electromagnetic environment. Such complex electromagnetic environment has increased The difficulty of radio signal analysis and processing, and impulsive noise is a typical representative of such a complex electromagnetic environment. Therefore, how to improve the ability of various signal processing methods to suppress impulsive noise has become an urgent problem to be solved. [0003] The multiple signal classification (MUSIC) algorit...

AI Technical Summary

Problems solved by technology

However, in some complex electromagnetic environments, the noise obeys the non-Gaussian distribution, its amplitude has a strong impulsiveness, and its probability distribution has a heavy tail, which can usually be represented by an Alpha stable distribution. At this time, the covariance is does not converge, so the performance of the traditional multi-signal classification algorithm will decline, or even completely fail.

Images