Echo Cancellation System in VOIP Using MATLAB
Uploaded by
ijsretEcho Cancellation System in VOIP Using MATLAB
Uploaded by
ijsretInternational Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
42
Volume 6, Issue 1, January 2017
Echo Cancellation System in VOIP using MATLAB
SAURABH K. DAHIVADKAR1, MARUTI LIMKAR2
1
(Department of Electronics & Telecommunication, MUMBAI University, INDIA
Email: [email protected])
2
(Department of Electronics & Telecommunication, MUMBAI University, INDIA
Email: [email protected])
ABSTRACT instantaneously while audio input output delay variation
can calculate pure delay between room echo and buffer.
Acoustic echo cancellation (AEC) method reduces
undesired acoustic echo arriving at microphone and also 2.2 Conventional AEC:
emphasizes the target talkers voice. It cancels the non
The conventional AEC method is shown in figure 1.
linear acoustic echo which appears due to distortion and
ultimate pure delay caused due to both room echo and
audio input and output buffers. This AEC method
reduces more than 40 dB of the undesired echo.
Keywords Acoustic Echo Cancellation, Conventional
AEC, Adaptive Filter, Voice Over Internet Protocol.
I. INTRODUCTION
Now a days voice over internet protocol (VOIP) phone
applications which run over smart phones and tablets
have very popular AEC helps to prevent detrimental
acoustic echo and howling which is generated due to Fig.1 Conventional AEC Method.
coupling between loudspeakers and microphones. There
are several AEC techniques of which consist of an The echo reduction gain can be calculated using the
adaptive filter (ADF). It identifies the acoustic echo path following Wilmer filtering method,
and cancels out the acoustic echo but still some residual
echo remains back due to which echo reduction is used
which reduces the residual echo in the error signal.
II. PROPOSED SYSTEM
Where,
2.1 ADF and ER:
is the estimate of the residual echo level and
The ADF and ER when combined together improve the is given by,
performance. But there are three limitations of AEC
loudspeaker distortion, microphone sensitivity variation
and audio input output delay variation. Hence a new
AEC method is proposed which uses three new Where,
techniques that is loudspeaker distortion cancels out both
linear as well as non linear echoes, microphone
sensitivity variation tracks the residual echo level
www.ijsret.org
International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
43
Volume 6, Issue 1, January 2017
Where, E[.] is the reasmble average.
is the estimated power frequency response
Proposed AEC: of residual echo path. Delay estimation technique
calculates the echo path transfer segment as:
The proposed AEC Method, reduces the issues of
distortion variations in level and delay with ADF, ER
and DE techniques. ADF eliminates linear and non
linear echoes, ER takes the residual echo levels and DE
calculates pure delay caused from echo and buffer
produces audio input output.
Non linear ADF compensates non linear echo caused by
the saturation effects because of small loudspeaker and
poor amplifier. The output signal X(n) is given by,
Where,
is frequency domain microphone signal, H is
complex conjugate transposition, L and M are the
number of multi delay filter.
Where, a(n) denotes non negative hard clipping
III. RESULTS AND DISCUSSIONS
threshold.
This u(n) is transformed into frequency domain signal The proposed AEC method consist of sampling
u() and frequency domain estimate of echo signal can frequency switch (SFS), Sampling frequency converter
be calculated as: (SFC), Analysis filter (AF), Loss control (LC), Synthesis
filter (SF), Sound device control, Delay estimator,
Buffer and acoustic echo controller. The signal enters
decode followed by AEC software and the sampling
Instantaneous ER eliminates the residual echo level by frequency is selected by SFS. Then it is decoded into
separating level and spectral structure from the residual two or three sub bands whose ranges are 0-4, 4-8, 8-16
echo. The residual echo level changes when the KHz. The delay between the acoustic echo and sound
microphone sensitivity varies even if the echo path device is estimated by delay estimator and then the
changes. As a result, the power spectrum of residual signal undergoes gain control by the LC. Synthesis filter
echo can be calculated by, (SF) resynthesizes the sub band signals and selecting the
sampling frequency the output is obtained at the
encoder.
Where,
www.ijsret.org
International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 0882
44
Volume 6, Issue 1, January 2017
[5] J. Casar-Corredera and J. Alcazar-Fernandez, An
acoustic echo canceller for teleconference systems,
Proc. IEEE International Conference on Acoustics,
Speech, and Signal Processing, Tokyo, Japan, vol. 11,
pp. 13171320, Apr. 1986.
Fig.2 Block diagram of AEC software implemented
in VOIP application.
IV. CONCLUSION
The proposed AEC method helps to reduce the undesired
acoustic echo in smart phones and tablet model which is
implemented in voice over internet protocol (VOIP)
hand free phone application which is better than
conventional AEC method. The proposed AEC method
estimates instantaneous residual echo variation non
linear acoustic echo and pure delay which results in
better performance than conventional AEC method.
REFERENCES
[1] Gerald Enzner, Peter Vary, A softpartitioned
frequency-domain adaptive filter for acoustic echo
cancellation,Proc. IEEE Conference on Acoustics,
Speech and Signal Processing, (ICASSP 2003),Vol. 5,
2003, 393-396.
[2] Mr. K. G. Gunale, Ms. S. N. Motade, Dr. S. L.
Nalbalwar, Frequency domain adaptive filter using FFT
algorithm for acoustic echo cancellation, Third
International Conference on Emerging Trends in
Engineering and Technology, IEEE, 2010, 582-587.
[3] L. Caviglione, A simple neural framework for
bandwidth reservation of VoIP communications in cost-
effective devices, IEEE Trans. Consumer Electron., vol.
56, no. 3, pp.12521257, Aug. 2010.
[4] H.-H. Choi, J.-R. Lee and D.-H. Cho, On the use of
a power-saving mode for mobile VoIP devices and its
performance evaluation, IEEE Trans. Consumer
Electron., vol. 55, no. 3, pp.15371545, Aug. 2009.
www.ijsret.org
