Adaptive filtering algorithms for acoustic noise

cancellation – LMS and RLS

S. Abinaya S. Sanjana
IIITDM Kancheepuram IIITDM Kancheepuram
esd15i003@iiitdm.ac.in esd15i020@iiitdm.ac.in

Abstract— For the cancellation of noise signal adaptive

filtering technique is used. In many applications of noise III. PROPOSED ALGORITHM
cancellation, the changes in signal is quite fast. This A. Least Mean square algorithm
requires adaptive algorithm, which converge rapidly.
From this point of view LMS and RLS algorithms suit There are many techniques that can be used in noise
these situations best. LMS algorithm is one of the most cancellation. Two such algorithms are Least Mean
used algorithms in many signal processing applications. Square Algorithm (LMS) and Filtered-x Least Mean
Unfortunately, LMS algorithm has high computation Square Algorithm (FXLMS). Any LMS adaptive process
complexity and stability problems. Here, we have runs through 2 sub processes – Filtering and Adaption.
presented RLS, which is a deterministic algorithm that Filtering involves computing the output of a linear
works well for high frequency noise signals. traversal filter in response to the input signal where all
the samples are independent of each other. Also involves
Keywords—Active Noise Cancellation (ANC); Tap-
generating the estimation error comparing the output
length; LMS; Adaptive filters; RLS;
with the desired response. Desired signal is generated in
I. INTRODUCTION the due course from past input and output signals which
are not corrupted by signal impurities such as active
Noise cancellation in acoustic signals is done by noise, echo and reverberation. In most of the cases, the
using optimum filtering techniques, i.e., adaptive filters. desired signal is an exact replica of input signal and it is
The coefficients of adaptive filter are varied with time found with the process of auto regressive analysis
using different algorithms. FIR adaptive filters are 3(ARA). Adaptation involves automatic adjustment of
widely used and are described in terms of tapped-delay- the parameters of the filter (weights of coefficients) in
line and lattice models. In FIR adaptive filters, the tap- accordance with the estimation error. The combination
length is the total number of filter coefficients or weights of filtering and adaptation forms a closed loop feedback
of the filter. Tap-length plays a major role in achieving system where the feedback signal is non-additive in
the steady state of the system. This is because, the MSE nature.
rises the number of weights are very less, but having a The LMS algorithm is classified as adaptive filtering in
large number of coefficients might introduce more which is a self-designing and time-varying system that
complexity. continuously adjust its tap weight in the algorithm [1].
LMS algorithm has been used to supress the effect of
II. LITERATURE REVIEW acoustic noise in speech signals. is the weight
vector of the filter at the time instant n. e(n) is the
difference between the desired and output signal.
Generally, in adaptive noise cancellers, the input is
filtered using a finite impulse response filter. The filter
coefficients would then be adjusted as required using
algorithms such as steepest descent or gradient search in
order to minimize the error function.

After solving, we get where

R is the input autocorrelation matrix.
 B. Recursive least square algorithm
RLS algorithm recursively finds the coefficients to
minimize the following weighted linear least squares
cost function relating to the input signal.

Fig. 1. LMS Algorithm

is known as the forgetting factor. It is chosen
On the other hand, the FXLMS algorithm will need a depending on the desired speed and frequency selectivity
filtered reference signal as input as requires a filtered in estimation. A small value of the forgetting factor may
version of the reference signal as input in which the filter result in the estimation to happen fast but it will be less
is having the same impulse response as the cancellation selective. That is, the convergence is achieved soon but
path. FXLMS has been used on cancelling the periodic steady state value might not be as required. Hence it is
noise generated by laptop fan after identification [4]. necessary to find a proper trade-off value of
There was another development in 2013. To improve
convergence rate and noise reduction ratio in ANC, the The value of wopt(n) that minimize J(w,n) is taken as the
variable tap length and step size FXLMS were filter weights.
introduced [5]. Yang et al. (2014) reduced 8dB noise
level in the cabin of high-speed elevator by
implementing modified FXLMS algorithm [6].
However, the FXLMS algorithm has higher complexity
compared to LMS algorithm due to the consideration of Where,
the secondary path parameter.

Algorithm :
With the given input vector X(n) and the desired output
signal d(n),
1) Update the input to the filter.

2) Compute the filter output using the previous set of

filter coefficients w(n-1).

3) Compute the error e(n)

4) Compute the Kalman Gain vector k(n)

Note: To achieve fast estimation when a change occurs

in the input signal, the gain matrix of the RLS algorithm
k(n) must be increased for a short time.
5) Update P according to the following equation
 6) Update the filter coefficients for the next iteration
Convergence of RLS is much faster than LMS though
computational complexity of RLS substantially
increases.
IV. SIMULATION

A. Least Mean square algorithm

Audio file of Waveform Audio File Format (.wav) was

used as input for the simulation model which was done
in MATLAB software environment. Input is a speech
signal of sampling frequency 8 kHz. An acoustic noise
with a frequency of 50 Hz is added to the input signal.
The algorithm is initiated by setting column weight
vector w(n) and filtered output e(k) as a zero column
vector. The initial value of μ is taken as 0.005 randomly
with the following LMS algorithm condition [2]: The simulation using the LMS adaptive filter algorithm
attenuates the noise gradually and finally cancels the
noise in the corrupted speech signal. The images (***)
Taking total number of iterations to infinity, it is not show the clean speech signal after the filtering done. It is
practically realisable. Hence a more stringent condition only slightly different from the original noise-free
for step size variation is used for LMS algorithm to speech signal. The amplitude and the noise of the signal
restrict the values of μ beyond a particular limit which are reduced after filtering.
causes severe degradation in steady state performance.
Input is multiplied with the weight vector to generate a
weighted signal y(n). This is subtracted from the desired B. Recursive Least Square Algorithm
signal to obtain error signal e(n). After each iteration, the Audio file of Waveform Audio File Format (.wav) was
weigh vector updates according to equation (*). The used as input for the simulation model which was done
filtering process goes on until it reaches the length of in MATLAB software environment. Input is a speech
primary input signal. The output of this process and the signal of sampling frequency 8 kHz. An acoustic noise
original error-free speech signal are compared in both with a frequency of 500 Hz was added to the signal. The
time and frequency domain. The implementation of the algorithm is initiated by setting the initial input
LMS adaptive filter is tested this way. covariance estimate as 0.1 and P matrix as an identity
matrix of order M. In our case, M was taken as 32. The
performance of RLS algorithm is tested by comparing
the plots of the convergence of the adaptive filter
response to the response of the FIR filter using
Timescope.
 V. CONCLUSION
The LMS and RLS adaptive filtering techniques are
implemented.
The simulation model establishes the fact that LMS
adaptive algorithm is capable of updating the weights
and essentially cancelling the noise. Lower frequency
noise (upto 5kHz) can be removed using this algorithm.
From the implementation of this project and the
synthesis report it can also be concluded that the
computation time and the area required for LMS
algorithm is less. On the basis of simulation, a
comparison is presented which shows that the RLS
algorithm has better performance in terms of noise
cancellation at higher frequencies.

REFERENCES

1. Hamidia M and Amrouche A 2016 Improved

variable step-size NLMS adaptive filtering algorithm
for acoustic echo cancellation Digital Signal
Processing 49 pp 44-55
2. Apolin JA, Diniz PSR, Laakso TI and Campos MLR
1998 Step-size optimization of the BNDRLMS
algorithm 9 th European Signal Processing
Conference
3. E. Horita, K. Sumiya, H. Urakami and S. Mitsuishi
A leaky RLS algorithm: it’s optimality and
implementation IEEE Transactions on Signal
Processing ( Volume: 52 , Issue: 10 , Oct. 2004 )
4. Cordourier-Maruri HA and Orduna-Bustamante F
2009 Active Control of Periodic Fan Noise in Laptos:
Spectral Width Requiremnts in Delayed Buffer
Implementation Journal of Applied Research and
Technology 7(2) pp 124-35
5. Chang DC and Chu FT 2013 Active noise cancellation
with a new variable tap length and step size FXLMS
algorithm IEEE International Conference on
Multimedia and Expo
6. Yang IH, Jeong JE, Jeong UC, Kim JS and Oh JE 2014
Improvement of noise reduction performance for a
high-speed elevator using modified active noise
control Applied Acoustics 79 pp 58-68