Method and system for stereo echo cancellation for VoIP communication systems
An exemplary embodiment of the present invention is directed toward a method and system for cancelling line echo in the presence of a known secondary audio signal. Filter adaption is enabled in the presence of a known secondary audio source such as the sound of a computer game, a music signal or other secondary audio sources that would otherwise prevent echo cancellation due to an apparent double talk condition. It is emphasized that this abstract is provided to comply with the rules requiring an abstract which will allow a searcher or other reader to quickly ascertain the subject matter of the technical disclosure. It is submitted with the understanding that it will not be used to interpret or limit the scope or the meaning of the claims.
1. An echo canceller, comprising:
a combiner for combining a secondary audio signal and a far end primary telephony signal into a single combined reference signal; and
an adaptive filter coupled to the combiner for receiving the single combined reference signal as input, the adaptive filter having filter coefficients adapted to cancel a combination of an electrical echo and an acoustic echo in a near end signal by modeling in parallel an electrical echo path and an acoustic echo path, wherein the electrical echo is caused by imperfect impedance matching of a network interface in the electrical echo path.
2. The echo canceller of claim 1 , wherein the adaptive filter comprises a finite impulse response filter.
3. The echo canceller of claim 2 , wherein the finite impulse response filter comprises a linear transversal filter.
4. The echo canceller of claim 1 , wherein the adaptive filter uses a normalized least mean square algorithm.
5. The echo canceller of claim 1 , wherein the acoustic echo comprises at least a portion of the secondary audio signal that is broadcast by a speaker and received by a near end microphone coupled to the echo canceller.
6. The echo canceller of claim 5 , wherein a set top box generates, at least in part, the secondary audio signal.
7. The echo canceller of claim 1 , further comprising:
double talk logic to detect near end speech in the near end signal, the double talk logic controlling filter adaptation of the adaptive filter as a function of detection of near end speech in the near end signal.
8. The echo canceller of claim 7 , wherein the double talk logic uses a sample based double talk detection algorithm.
9. The echo canceller of claim 7 , wherein the double talk logic detects near end speech in the near end signal based on whether an absolute value of the near end signal is greater than a weighted sum of a voice signal of the far end primary telephony signal, the secondary audio signal and a background noise of the near end signal.
10. The echo canceller of claim 1 , wherein the secondary audio signal comprises a pulse metering tone.
11. The echo canceller of claim 1 , further comprising:
a decimator that downsamples the secondary audio signal to match a sample rate of the far end primary telephony signal.
12. The echo canceller of claim 1 , further comprising:
a difference operator to subtract an echo estimate of the single combined reference signal from the near end signal, wherein the adaptive filter generates the echo estimate of the single combined reference signal.
13. An echo canceller, comprising:
an adaptive filter having filter coefficients adapted to cancel an echo in a near end signal by modeling an echo path, wherein the echo comprises an electrical echo of a combination of a far end primary telephony signal and a secondary audio signal, and the electrical echo is caused by imperfect impedance matching of a network interface in the echo path.
14. The echo canceller of claim 13 , wherein the secondary audio signal comprises a pulse metering tone.
15. The echo canceller of claim 13 , further comprising:
double talk logic to detect near end speech in the near end signal, the double talk logic controlling filter adaptation of the adaptive filter as a function of detection of near end speech in the near end signal.
16. The echo canceller of claim 15 , wherein the double talk logic detects near end speech in the near end signal based on whether an absolute value of the near end signal is greater than a product of a gain of the network interface and a peak statistic of the combination of the far end primary telephony signal and the secondary audio signal summed with an estimated background noise of the near end signal.
17. The echo canceller of claim 13 , wherein the adaptive filter is further adapted to cancel a combination of the electrical echo and an acoustic echo in the near end signal by modeling in parallel the electrical echo path and an acoustic echo path.
18. The echo canceller of claim 17 , wherein the acoustic echo comprises at least a portion of the secondary audio signal that is broadcast by a speaker and received by a near end microphone coupled to the echo canceller.
19. The echo canceller of claim 18 , wherein a set top box generates, at least in part, the secondary audio signal.
20. The echo canceller of claim 13 , further comprising:
a decimator that downsamples the secondary audio signal to match a sample rate of the far end telephony signal.