|Download Help (Windows Only)|
The purpose of adaptive echo cancellation is to remove echoes from communication systems. Line echo, also known as network echo, and acoustic echo are two types of echoes. Line echo occurs over telephone lines due to circuit imperfections or impedance mismatches between telephone circuits. Acoustic echo occurs in applications such as audio conferencing and hands-free telephony. This topic describes acoustic echo cancellation only.
The following figure shows a diagram of acoustic echo cancellation.
When you place a speaker near a microphone in a full-duplex communication environment, acoustic echo occurs. As the previous figure shows, Microphone A acquires the voice from the near-end person and transfers the voice to Speaker B at the far end. Microphone B then picks the voice from Speaker B at the far end and transfers the voice back to Speaker A at the near end. When the near-end person talks to Microphone A, this person hears his or her own distorted and delayed voice in Speaker A. The distorted and delayed voice is acoustic echo. In the previous figure, h(n) is the room impulse response, which models the acoustic characters of the room, including reverberation, resonance, and reflections caused by the boundaries of the room and the objects in the room. The acoustic echo affects the communication negatively when the delay is large enough.
Adaptive echo cancellation uses adaptive filters at both ends to estimate the transfer function between each speaker and its corresponding microphone. As the previous figure shows, when Microphone A transfers the speech signal to Speaker B, the same signal also feeds into an adaptive filter at the far end. The adaptive filter then adjusts the filter coefficients iteratively to estimate the distorted and delayed speech. If the estimated signal is close to the signal from Microphone B, no acoustic echoes occur at Speaker A.
Refer to the Adaptive Echo Cancellation VI in the examples\Adaptive Filters\Applications\Adaptive Echo Cancellation directory for an example that uses the LabVIEW Adaptive Filter Toolkit to remove echoes from a signal.