초록 ▼

The present invention relates to a method for cleaning signals for centralized antijamming. The invention makes it possible to provide as many cleaned antenna channels as sub-arrays with limited computational requirements. The method proceeds in two steps. Initially, an antijamming matrix is compute

The present invention relates to a method for cleaning signals for centralized antijamming. The invention makes it possible to provide as many cleaned antenna channels as sub-arrays with limited computational requirements. The method proceeds in two steps. Initially, an antijamming matrix is computed. This matrix depends on the noise covariance matrix, on a weighting vector representing the form of the desired antenna pattern on reception in an unjammed environment, and on constraints for preserving the shape of this antenna pattern. Subsequently, the signals arising from the antenna sub-arrays undergo a linear recombination effected by the antijamming matrix. The antijamming method is termed centralized since the data necessary for the antijamming processing are concentrated in the lone antijamming matrix. The invention applies notably to radar systems, notably to airborne radar systems.

대표청구항 ▼

1. A method for filtering noise using a covariance matrix Γ of the noise, said method comprising at least the following steps: receiving signals xi by a radar having N antenna sub-arrays, wherein each sub array receives a signal xi;computing coefficients Mij of a matrix M, i and j varying from 1 to

1. A method for filtering noise using a covariance matrix Γ of the noise, said method comprising at least the following steps: receiving signals xi by a radar having N antenna sub-arrays, wherein each sub array receives a signal xi;computing coefficients Mij of a matrix M, i and j varying from 1 to N, on the basis of the noise covariance matrix Γ, of a weighting vector wr representing a desired antenna pattern on reception in a clean environment, and of nc constraints for preserving the shape of this antenna pattern, said constraints being represented by a matrix C of dimension N×nc and by a vector f of dimension nc, the matrix M being equal to: wrwrH·wr·fH·(CH·Γ-1·C)-1·CH·Γ-1calculating using a computer processor the product of the matrix M times the vector xreceived formed of the signals received xi, the resulting vector being formed of the N filtered signals xi′. 2. The method for filtering noise as claimed in claim 1, wherein the product of the matrix M times the vector xreceived is parallelized as several scalar products calculated between a row of the matrix M and the vector Xreceived. 3. The method for filtering noise as claimed in claim 2, wherein the matrix M is recomputed periodically so as to following the jamming configuration. 4. The method for filtering noise as claimed in claim 1, wherein the matrix M is recomputed periodically so as to following the jamming configuration.