Mamta M. Mahajanet al.

/ (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES

Vol No. 5, Issue No. 2, 172 - 176

Design of Least Mean Square Algorithm for

Adaptive Noise Canceller

Mamta M. Mahajan1 ( Research scholar1 ) S. S. Godbole2 (Assistant Professor2 )

Electronics & communication Engineering Department Electronics & ecommunication Engineering department
G. H. Raisoni College of Engineering Nagpur India G. H. Raisoni College of Engineering NagpurIndia
1 2
mmmmahajan@gmail.com godbole_123@yahoo.com

T
Abstract— This paper describes the concept of adaptive requires little or no a priori knowledge of signal or
noise cancelling, an alternative method of estimating noise characteristics. In this paper we have to used
signals corrupted by additive noise or interference. The adaptive filter for noise cancellation. Noise
method uses a “primary” input containing the cancelling is a variation of optimal filtering that is
corrupted Signal and a “reference” input containing
highly advantageous in many applications. It makes
noise correlated in some unknown way with the
use of an reference input derived from one or more
ES
primary noise. The reference input is adaptively filtered
and subtracted from the primary input to obtain the
signal estimate. A desired signal corrupted by additive
noise can often be recovered by an adaptive noise
canceller using the least mean squares (LMS)
algorithm. Computer simulations with uncorrelated
Gaussian noise and signals confirm the results of the
analysis and demonstrate the effectiveness of the least
sensors located S at points in the noise field where
the signal is weak or undetectable. This input is
filtered and subtracted from a primary input
containing both signal and noise. As a result the
primary noise is attenuated or eliminated by
cancellation.
mean squares (LMS) algorithms.This Adaptive Noise II. CONCEPT OF ADAPTIVE NOISE
Canceller is then useful for enhancing the S/N ratio of CANCELLING
data collected from sensors (or sensor arrays) working
in noisy environment, or dealing with potentially weak
signals.
A
Keywords—LMS Algorithm,; Adaptive Noise
Canceller;S/N ratio; error estimation; adaptive filter

I. INTRODUCTION
IJ

In numerous application areas, including

biomedical engineering, radar, sonar and digital
communications, the goal is to extract a useful signal
corrupted by interferences and noises.
Noise/interference removal is facilitated when
multiple sensors on different locations record the
biomedical phenomenon simultaneously.
The usual method of estimating a signal Fig. 1 Adaptive Noise Canceller
corrupted by additive noise’ is to pass it through a
filter that tends to suppress the noise while leaving
As the name implies, ANC is a technique used to
the signal relatively unchanged. Filters used for the
remove an unwanted noise from a received signal,
above purpose can be fixed or adaptive.
the operation is controlled in an adaptive way in
The design of fixed filters is based on prior
order to obtain an improved signal-to noise ratio
knowledge of both the signal and the noise. Adaptive
(SNR) . As shown in ― Fig.1‖, an ANC is typically a
filters, on the other hand, have the ability to adjust
dual-input, closed-loop adaptive feedback system.
their own parameters automatically, and their design
The two inputs are: the primary input signal d(n) (the

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 172

Mamta M. Mahajanet al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
desired signal corrupted by the noise) and the
reference signal x(n) (an interfering noise supposed to
be uncorrelated with the desired signal but correlated
with the noise affecting the desired signal in an
unknown way).
Fig. 1 shows the basic problem and the
adaptive noise Canceling solution to it. A signal s is
transmitted over a channel to a sensor that also
receives a noise no uncorrelated with the signal. The
combined signal and noise s +no form the primary
input to the canceller. A second sensor receives a
noise nl uncorrelated with the signal but correlated in
some unknown way with the noise no. This sensor
provides the reference input to the canceller. The
noise nl is filtered to produce an output y that is as
close a replica as possible of no. This output is
subtracted from the primary input s + no to produce
the system output z = s + no - y .
In the system shown in Fig. 1 the reference
input is processed by an adaptive filter. An adaptive
filter differs from a fixed filter in that it automatically
adjusts its own impulse response. Thus with the
ES
proper algorithm, the filter can operate under
changing conditions and can readjust itself
continuously to minimize the error signal. The error
signal used in an adaptive process depends on the
nature of the application.
In noise cancelling systems the practical objective
is to produce a system output z = s + no – y that is a
best fit in the least squares sense to the signal s.This
objective is accomplished by feeding the system
output back to the adaptive filter and adjusting the
filter through an LMS adaptive algorithm to
A
minimize total system output power.
In an adaptive noise cancelling system, in other
words, the system output serves as the error signal for
the adaptive process. It might seem that some prior
knowledge of the signal s or of the noises no and nl
would be necessary before the filter could be
designed, or before it could adapt, to produce the
IJ
noise cancelling signal y.
Assume that s, no, n 1, and y are statistically
stationary and have zero means. Assume that s is
uncorrelated with no and n l , and suppose that n1 is
correlated with no. The output z is
z=s+no-y. (1)
Squaring, one obtains
z2 = s2 +(no - y)2 + 2s(no - y). (2)
Taking expectations of both sides of (2), and
realizing that s is uncorrelated with no and with y,
yields
E[z2]=E[s2]+E[(n0-y)2]+2E[s(n0-y)]
=E[s2]+E[(n0-y)2] (3)
The signal power E [ s 2 ] will be unaffected as the
filter is adjusted to minimize E[z2 ]. Accordingly, the
minimum output power is
ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved.
Vol No. 5, Issue No. 2, 172 - 176
mine[z2]=E[s2]+mine[(no-y)2] (4)
When the fiiter is adjusted so that E[z2] is
minimized, E[(no -y)2 ] is, therefore, also minimized.
The filter output y is then a best least squares
estimate of the primary noise no. Moreover, when
E[(no - y)2] is minimized[(z -s)2] is also minimized,
since, from (I),
(z - s) = (no - y). (5)
Adjusting or adapting the filter to minimize the
total output power is thus tantamount to causing the
output z to be a best least squares estimate of the
signal s for the given structure and adjustability of the
adaptive filter and for the given reference input. The
output z will contain the signal s plus noise. From
(l),the output noise is given by (no -y). Since
minimizing E[z2] minimizes E[(no - y)2] minimizing
T
the total output power minimizes the output noise
power. Since the signal in the output remains
constant, minimizing the total output power
maximizes the output signal-to-noise ratio.
ANC technique has been successfully applied
to many applications, such as acoustic noise
reduction, adaptive speech enhancement and channel
equalization. In these cases and other related high
speed required applications, pure software
implementation would bring about long processing
time, thus cannot meet the requirement. An effective
way can be represented by a hardware
implementation on FPGA.
III. LMS ALGORITHM
The LMS algorithm is a widely used algorithm for
adaptive filtering. The algorithm is described by the
following equations:
M-1
y(n) = Σ wi(n) * x(n-i); (1)
i=0
e(n) = d(n) – y(n) (2)
wi(n+1) = wi(n) + 2ue(n)x(n-i); (3)
In these equations, the tap inputs x(n),x(n-
1),……,x(n-M+1) form the elements of the reference
signal x(n), where M-1 is the number of delay
elements. d(n) denotes the primary input signal, e(n)
denotes the error signal and constitutes the overall
system output. wi(n) denotes the tap weight at the nth
iteration. In equation (3), the tap weights update in
accordance with the estimation error. And the scaling
factor u is the step-size parameter. u controls the
stability and convergence speed of the LMS
algorithm. The LMS algorithm is convergent in the
mean square if and only if u satisfies the condition: 0
< u < 2 / tap-input power
where tap-input power =
Page 173
Mamta M. Mahajanet al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
The flowchart of the signals of the LMS algorithm
is shown in‖ Fig.2.‖
Fig. 1 Flowchart of the LMS algorithm
IV. LMS CORE IMPLEMENTATION
ES
A
IJ
Fig.3. Block Diagram of the LMS Core
1) Block Diagram
The block diagram of the LMS core is shown in
Fig.3.
The LMS core is divided into five blocks
1. The Control block arranges the timing of the
whole system. It produces four enable signals:
en_x,en_d,en_coee, en_err, which enable the Delay
Block, the Weight Update Block and the Error
Counting Block separately.
2. The Delay Block receives the reference signal
x_in and the primary input signal d_in under control
of the enable signal en_x and en_d. And it produces
the M tap delay signal x_out.
ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved.
Vol No. 5, Issue No. 2, 172 - 176
3. The Multiply Accumulator (MAC) Block
multiply the M_tap reference signal x_out with the
M_tap weight w separately, and add them together,
then we get yn.
4. The Error Counting Block subtract yn from dn
and get the error signal e_out, which is also the
output of the whole system. And it produces signal
xemu as a feedback by multiplying e_out, x_out and
the scaling factor u.
5. The Weight Update Block updates the weight
vector w(n) to w(n+1) that will be used in the next
iteration.
1] We programmed the delay block with VHDL
under the platform of Xilinx ISE 9.1i.In this block we
have to apply one reference input containing noise i.e
T
x_in and one primary input containing the corrupted
signal i.e. d_in and two enable signals en_x &en_d so
that we get the delay output x_out.
As shown in ― Fig 4‖ and ― Fig 5‖ when both the
enable signal get 1 then input equals to output i.e
x_in=x_out
When both the enable signal get 0 then we have to
produce one delay signal so that our reference input
and primary input are not mixed.
2] In control block we have to apply four inputs
clk,reset read & write which produces four enable
signals: en_x,en_d,en_coee, en_err, which enable the
Delay Block, the Weight Update Block and the Error
Counting Block separately. We programmed control
block in such a way that when clk=1 and read=1 then
all the four enable signal get 1 that means we read the
input and produce the output. As shown in ― Fig 6‖
and ― Fig 7‖ when clk=1 and write =1 then all the four
enable signal get 0 that means we are reading any
input just check the output.
3]in MAC block we have to apply two input one
reference input x_out and weight input w..MAC
block is basically multiple accumulator block.In this
block we have to multiply reference input x_out with
weight w separately, and add them together, then we
get yn. While programming MAC block as shown in
―Fig 8‖ and ― Fig 9‖ when clk=1 and reset=1 then
filter output=0 and when clk=0 and reset=0 then filter
output is equal to reference input x_out multiply by
weight w and add them with previous filter
output.Initialy we have to consider filter output as a
zero.
4]In weight update block we have to apply two input
signal.One enable signal en_coee and one feedback
signal xemu.In this block as shown in ― Fig 10‖ and
―Fig 11‖ when en_coee=1 then next weight is equal
to current weight plus feedback signal.
5]In error counting block we have to apply four input
yn,x_out,dn,en_err and produce two output e_out &
xemu.While programming error counting block as
Page 174
Mamta M. Mahajanet al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 172 - 176

shown in ― Fig12‖ we calculate error signal e_out by

subtracting yn i.e filter output from dn i.e primary
input signal which is the output of the whole system
and also calculate xemu as a feedback signal by
multiplying e_out,x_out and scaling factor u.

V Simulation result for LMS Algorithm

Fig 7.Simulation result for control block.

T
Fig 4.Simulation result for delay block.
ES Fig 8.Simulation result for MAC block
A
Fig 9.Simulation result for MAC block.
Fig 5 Simulation result for delay block.
IJ

Fig 10 .Simulation result for weight update block.

Fig 6.Simulation result for control block.

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 175

Mamta M. Mahajanet al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 172 - 176

pp. 1692-1716.
[2] D. Nicolae, R. Romulus, ― Noise canceling in
Audio signal with adaptive filter‖, University
of Oradea,Vol. 45, Number 6,2004, pp 599-
602
[3] A. Elhossini, S. Areibi and R. Dony,― An FPGA
Implementation of the LMS Adaptive Filter for
Audio Processing,‖ in Proc. IEEE International
Conference on Reconfigurable Computing and
Fig 11.Simulation result for weight update block. FPGAs,Sept.2006, pp. 1-8.
[4] S. Haykin, Adaptive Filter Theory, Prentice-
Hall, third edition, 2002.
[5] Boo-Shik Ryu, Jae-Kyun Lee, ― The
Performance of an adaptive noise canceller
with DSP processor‖,40th Southeastern
Symposium on System Theroy, New Orleans,
USA, March 16-18, 2008, pp 42-45.

T
[6] S. A. Hadei, ― A Family of Adaptive Filter
Algorithms in Noise Cancellation for Speech
Enhancement‖, International Journal of
Computer and Electrical Engineering, Vol. 2,
No. 2, April 2010, pp 307-315
[7] G. Long, F. Ling and J. G. Proakis, ― The LMS
algorithm with delayed coefficient adaptation,‖

VI. CONCLUSION
ES
Fig 12. Simulation result for error counting block.

Adaptive noise cancelling is a method of

adaptive filtering that can be applied whenever a
suitable 'reference input is available. The principal
IEEE Trans. on ASSP, vol. 37, Sept. 1989, pp.
1397-1405.
[8] L. K. Ting, R. F. Woods and C. F. N. Cowan,
― Virtex FPGA Implementation of a Pipelined
AdaptiveLMSPredictorforElectronicsupport
Measures Receivers,‖ IEEE Trans. VLSI
vol.
Syst.,

advantages of the method are its adaptive capability, 13, Jan. 2005, pp. 86-95.
its low output noise, and its low signal distortion. The [9] B. Widrow, ― Adaptive fdters 1: Fundamentals,‖
adaptive capability allows the processing of inputs Stanford Electronics Lab., Stanford Univ., Rep.
SUSEL-66-126, Dec. 1966
whose properties are unknown. Output noise and
[10]Kim, J.,Poularikas, A.D.:‖Performance of noise
signal distortion are generally lower than can be
A
canceller using adjusted step size LMS algorithm‖,
achieved with conventional optimal filter System Theory, 2002.Proceedings of theThirty-
configurations. In this paper, software Fourth Southeastern Symposium on, 18-19 March
implementation of LMS algorithm is presented to 2002.
implement the ANC system. [11] Xilinx Coproration ‖Spartan-IIE 1.8V FPGA
Family:
REFERENCES Functional Description‖,Nov. 2002.
IJ

[1] B. Widrow, J. R. Glover, J. M. McCool and et

al.,‖Adaptive Noise Cancelling: Principles and
Applications‖,Proc .IEEE, vol. 63, Dec. 1975,

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 176