| 摘要 |
PURPOSE:To improve call quality in a loud speaking call, by controlling the propagation characteristic of a pseudo echo path repeatedly so as to set n- sample values outputted from subtractors at minimum by using a bit of average correction information generated by averaging n-correction information. CONSTITUTION:For a constant time immediately after starting correction, the propagation characteristic of a sequential pseudo echo path 8a is controlled repeatedly at every phase so as to set the n-sample values outputted from the subtractors 10-1-10n at minimum by using the n-correction information generated from the n-sample values of delivery signals outputted from second thin-out circuits 15-1-15-n and the n-sample values outputted from the subtractors 10-1-10-n. Afterwards, the propagation characteristic of the pseudo echo path 8a is controlled repeatedly so as to set the n-sample values outputted from the subtractors 10-1-10-n at minimum by using the bit of average correction information generated by averaging the n-correction information. In such a way, it is possible to accelerate the convergence speed of the echo, and to increase an echo elimination quantity.
|
