Micro-programmed Adaptive Filtering Applications - Digital Design of Signal Processing Systems: A Practical Approach - page 463

for i=MAX_SPREAD:Length_Training
xn=Training_Data(i:-1:i-MAX_SPREAD+1);
y(i)=xn*hn;
d(i)=Training(i);
e(i)=d(i)-y(i);
hn = hn + mue*e(i)*xn';

end

11.5 Echo Canceller
An echo canceller is an application where an echo of a signal that mixes with another signal is
removed from that signal. There are two types of canceller, acoustic and line echo.

11.5.1 Acoustic Echo Canceller
In a speaker phone or hands-free mobile operation, the far-end speaker voice is played on a speaker.
Multiple and delayed echoes of this voice are reflected from the physical surroundings and are
picked up by the microphone along with the voice of the near-end speaker. Usually these echo paths
are long enough to cause an unpleasant perception to the far-end speaker. An acoustic echo canceller
is used to cancel the effects. A model of the echo cancellation problem is shown in Figure 11.6.
Considering all digital signals, let u[n] be the far-end voice signal. The echo signal is r[n] and it
gets added in to the near-end voice signal x[n]. An echo canceller is an adaptive filter h[n] that ideally
takes u[n] and generates a replica of the echo signal, r½n . This signal is cancelled from the received
signal to get an echo-free signal,

x½n , where:

x½n¼x½nþr½nu½n * h ½n
x½n¼x½nþr½nr½n:

u(n)

Adaptive
Algorithms

h(n)

^

r[n]

r(t)

-

x(t)

+

e(n)

near-end

x(n) + r(n)

Figure 11.6 Model of an acoustic echo canceller