Appeal No. 2002-1197
Application 09/131,167
should be kept within a predetermined range that does not cause
distortion. However, Koski teaches not updating the filter
coefficients based on the condition of double talk detection, not
only updating the coefficients based upon the condition that the
output signal of the microphone is within a predetermined range
of possible output signal values, as claimed. Not updating the
filter coefficients based on double talk detection will
indirectly keep the output signal of the microphone within a
predetermined range of output signal values, but this is not the
same as updating based on the output signal value. We agree with
appellant's argument "that when double-talk does not exist, the
double-talk detector of Koski will not prevent the parameters of
the adaptive filter from being updated, regardless of the
prevailing value of the output signal of the microphone because
Koski discloses freezing the parameters of the adaptive filter
only when double-talk exists" (RBr3). Thus, the condition under
which Koski freezes the coefficients of the adaptive filter are
not based on whether the output signal is within a predetermined
range. The examiner has not presented any reasoning that since
distortion is caused when the magnitude of the superposed near-
end and echo signals is too great, due to double talk, it would
have been obvious to one skilled in the art to freeze the
coefficients based on the magnitude of the output signal rather
than based on the presence of double talk.
- 11 -
Page: Previous 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Next
Last modified: November 3, 2007