Synchronization of echo cancellers in a voice processing system
View Patent ↗Methods and apparatus for processing a received voice signal from a variable-delay network by a communications system that has at least one echo canceller. An echo canceller determines delay characteristics that are associated with the received voice signal and correspondingly modifies echo cancellation processing. In a variation of the embodiment of the invention, a first echo canceller sends silence descriptor information to a second echo canceller. The second echo canceller reconstructs a voice signal based upon the silence descriptor information and excitation vector information.
1. In a communication system comprising at least two echo cancellers, a method comprising:
(a) determining, by a first echo canceller of the at least two echo cancellers, silence descriptor information related to a transmitted voice signal sent from the first echo canceller to a second echo canceller of the at least two echo cancellers, the first echo canceller and the second echo canceller being at different ends of the commiunication network;
(b) synchronizing a local random codebook excitation selection at the second echo canceller, the codebook being used to determine an excitation vector of a speech coder;
(c) transmitting, by the first echo canceller to the second echo canceller, the silence descriptor information and excitation vector information; and
(d) reconstructing, by the second echo canceller, the transmitted voice signal based upon the silence descriptor information and the excitation vector information.
2. The method of claim 1 , wherein the silence descriptor information comprises spectral information regarding the transmitted signal.
3. The method of claim 2 , wherein the excitation vector information identifies a particular excitation vector that, when filtered according to the spectral information, provides an estimate of the transmitted signal.
4. The method of claim 1 , further comprising:
(e) receiving, by the first echo canceller from the second echo canceller, a received voice signal based upon the silence descriptor information and the excitation vector information; and
(f) modifying, by the first echo canceller, echo cancellation processing on the transmitted voice signal based on the received voice signal.
5. The method of claim 1 , further comprising:
(a)(1) including the silence descriptor information in a packet that is sent from the first echo canceller to the second echo canceller;
(b)(1) including synchronization information in the packet; and
(b)(2) synchronizing the local random codebook excitation from the synchronization information.
6. The method of claim 1 , (b) comprising:
(b)(i) synchronizing random number generation between the different ends of the communication network.
7. The method of claim 6 , (b) further comprising:
(b)(ii) utilizing unused bits in a packet containing the silence descriptor information.
8. The method of claim 6 , further comprising:
(e) in response to (b)(i), accessing an echo signal when discontinuous transmission (DTX) is active and comfort noise generation is occurring at a far end of the communication network.
9. The method of claim 8 , further comprising:
(f) updating coefficients when the DTX is active and the comfort noise generation is occurring.
10. A communications system comprising:
a first echo canceller that determines silence descriptor information related to a transmitted voice signal sent from the first echo canceller to a second echo canceller of at least two echo cancellers and that sends synchronization information to the second echo canceller;
a second echo canceller that receives the silence descriptor information, the synchronization information, and excitation information from the first echo canceller, that reconstructs the transmitted voice signal based upon the silence descriptor information and the excitation vector information, and that uses the synchronization information to synchronize a local random codebook excitation selection, the codebook being used to determine an excitation vector of a speech coder.
11. The communications system of claim 10 , wherein the first echo canceller receives a received voice signal from the second echo canceller based upon the silence descriptor information and the excitation vector information and modifies echo cancellation processing on the transmitted voice signal based on the received voice signal.
12. The communications system of claim 10 , wherein the first echo canceller and the second echo canceller synchronize random number generation to synchronize the local random codebook excitation selection.
13. A communication system comprising:
means for determining, by a first echo canceller, silence descriptor information related to a transmitted voice signal sent from the first echo canceller to a second echo canceller;
means for synchronizing a local random codebook excitation selection at the second echo canceller, the codebook being used to determine an excitation vector of a speech coder;
means for transmitting, by the first echo canceller to the second echo canceller, the silence descriptor information and excitation vector information; and
means for reconstructing, by the second echo canceller, the transmitted voice signal based upon the silence descriptor information and the excitation vector information.
14. A computer-readable medium having computer-executable instructions for performing the steps comprising:
(a) determining, by a first echo canceller of the at least two echo cancellers, silence descriptor information related to a transmitted voice signal sent from the first echo canceller to a second echo canceller of the at least two echo cancellers, the first echo canceller and the second echo canceller being at different ends of the communication network;
(b) synchronizing a local random codebook excitation selection at the second echo canceller, the codebook being used to determine an excitation vector of a speech
(c) transmitting, by the first echo canceller to the second echo canceller, the silence descriptor information and excitation vector information; and
(d) reconstructing, by the second echo canceller, the transmitted voice signal based upon the silence descriptor information and the excitation vector information.
15. The computer-readable medium of claim 14 having computer-executable instructions for performing the further steps comprising:
(e) receiving, by the first echo canceller from the second echo canceller, a received voice signal based upon the silence descriptor information and the excitation vector information; and
(f) modifying, by the first echo canceller, echo cancellation processing on the transmitted voice signal based on the received voice signal.
16. The computer-readable medium of claim 14 having computer-executable instructions for performing the further steps comprising:
(b)(i) synchronizing random number generation between the different ends of the communication network.
17. The computer-readable medium of claim 16 having computer-executable instructions for performing the further steps comprising:
(b)(ii) utilizing unused bits in a packet containing the silence descriptor information.
18. The computer-readable medium of claim 16 having computer-executable instructions for performing the further steps comprising:
(e) in response to (b)(i), accessing an echo signal when discontinuous transmission (DTX) is active and comfort noise generation is occurring at a far end of the communication network.
19. The computer-readable medium of claim 18 having computer-executable instructions for performing the further steps comprising:
f) updating coefficients when the DTX is active and the comfort noise generation is occurring.