Adaptive echo cancellation method of affine projection maximum entropy sub-band

Abstract

The invention discloses an adaptive echo cancellation method of an affine projection maximum entropy sub-band. The adaptive echo cancellation method of the affine projection maximum entropy sub-band comprises the steps of A, far-end signal division, including, dividing a far-end signal vector U (n) into I far-end sub-band vectors Ui(n) through a first analysis filter, and dividing a near-end signal d(n) picked up by a near-end microphone into I near-end sub-band signals di(n) through a second analysis filter; B, signal extraction, including, carrying out I extraction for the far-end sub-band vectors Ui(n) through an extractor so as to obtain far-end input sub-band extraction vectors Ui(k), and carrying out I extraction for the near-end sub-band signals di(n) through the extractor so as to obtain near-end sub-band extraction signals di(k); C, affine projection; D, filter outputting, including, filtering a far-end input sub-band affine projection matrix X(k) through a filter so as to obtain an output vector Y(k); E, echo cancellation; F, tapping right vector updating, including, updating based on the maximum entropy V(k) of the current net signal vector so as to determine a right vector at the next extraction moment; and G, iteration, including, enabling k to be equal to k+1, and repeating the steps A, B, C, D, E and F until a call is ended. The echo cancellation effect is good, and the anti-jamming capability is strong.

Description

technical field [0001] The invention belongs to the technical field of adaptive echo cancellation for voice communication. Background technique [0002] In order to solve the problem of system echo interference during communication, various echo cancellation methods have been proposed. At present, echo cancellation methods mainly include frequency shift technology, sub-band center clipping technology, voice control switch technology, comb filter technology, microphone array technology, and adaptive echo cancellation technology. The first five methods have disadvantages such as high cost and difficult design. The adaptive echo cancellation method is not only low in cost, but also can obtain good performance. Therefore, the adaptive echo cancellation method is currently the mainstream echo cancellation technology. [0003] In the adaptive echo cancellation technology, the adaptive filter is a digital filter that can automatically adjust the performance according to the input...

AI Technical Summary

Problems solved by technology

However, when the algorithm converges, the algorithm weight vector is updated only through the sign of the error signal (positive or negative), regardless of the size of the error signal, so that although it has good anti-impulse interference ability, its convergence speed needs improvement

Images