Nonlinear echo suppression
View Patent ↗Presented is a method and associated system for suppression of linear and nonlinear echo. The method includes dividing an input signal into several frequency bands in each of a several of time frames. The input signal may include an echo signal. The method further includes multiplying the input signal in each of the several frequency bands by a corresponding echo suppression signal. Calculating the corresponding echo suppression signal may include estimating a power of the echo signal in a particular frequency band as a sum of several component echo powers, each of the several component echo powers due to an excitation from a far-end signal in a corresponding one of the several frequency bands. Calculating the corresponding echo suppression signal may further include subtracting the power of the echo signal in the particular frequency band from a power of the input signal in the particular frequency band.
1. A system for echo suppression, said system comprising:
a microphone configured to receive an input signal; and
a processor configured to:
divide the input signal into a plurality of frequency bands in each of a plurality of time frames, said input signal comprising an echo signal; and
multiply said input signal in each of said plurality of frequency bands by a corresponding echo suppression signal, wherein said processor is configured to calculate said corresponding echo suppression signal by:
estimating a power of said echo signal in a particular frequency band as a sum of a plurality of component echo powers, each of said plurality of component echo powers due to an excitation from a far-end signal in a corresponding one of said plurality of frequency bands;
subtracting said power of said echo signal in said particular frequency band from a power of said input signal in said particular frequency band;
wherein said processor is configured to estimate each of said plurality of component echo powers by:
applying a delay to said far-end signal in said corresponding one of said plurality of frequency bands; and
applying a weight to said far-end signal in said corresponding one of said plurality of frequency bands.
2. The system of claim 1 , wherein said processor is further configured to calculate said corresponding echo suppression signal by dividing by said power of said input signal in said particular frequency band.
3. The system of claim 1 , wherein said echo signal is a residual echo signal from a linear echo canceller.
4. The system of claim 1 , wherein said plurality of component echo powers are uncorrelated with one another.
5. The system of claim 1 , wherein said delay is chosen to maximize a cross-correlation between said input signal and said far-end signal in said corresponding one of said plurality of frequency bands.
6. The system of claim 1 , wherein said weight is a normalized cross-correlation between said input signal in said particular frequency band and said far-end signal in said particular frequency band.
7. The system of claim 1 , wherein said plurality of component echo powers comprise only component echo powers due to an excitation from said far-end signal in said corresponding one of said plurality of frequency bands, a frequency band having a frequency of one half of said corresponding one of said plurality of frequency bands, and a frequency band having a frequency of one third of said corresponding one of said plurality of frequency bands.
8. A system for echo suppression, said system comprising:
a microphone configured to receive an input signal; and
a processor comprising a linear echo canceller, said processor configured to:
divide said input signal into a plurality of frequency bands in each of a plurality of time frames; and
multiply said input signal in each of said plurality of frequency bands by a corresponding echo suppression signal, wherein said processor is configured to calculate said corresponding echo suppression signal by:
estimating a power of said echo signal in a particular frequency band as a sum of a plurality of component echo powers, each of said plurality of component echo powers due to an excitation from a far-end signal in a corresponding one of said plurality of frequency bands; and
subtracting said power of said echo signal in said particular frequency band from a power of said input signal in said particular frequency bands;
wherein said echo signal is a residual echo signal from said linear echo canceller, and
wherein said processor is configured to estimate each of said plurality of component echo powers by:
applying a weight to a synthesized echo in said corresponding one of said plurality of frequency bands, said synthesized echo obtained from said linear echo canceller.
9. The system of claim 8 , wherein said weight is a normalized cross-correlation between said input signal in said particular frequency band and said synthesized echo in said corresponding one of said plurality of frequency bands.