We are trying to implement an adaptive filter with input weights. The adaptive parameters are obtained by simulating noise canceller on MATLAB. Simulink model of adaptive Noise canceller was developed and Processed by FPGA.

1. Done By
S . Malki Hussain S.Chand Basha
S . Md .Javeed B.Hussain Basha
S.Baba Fakruddin S.Minuddin
Submitted to
S . Fouziya Parveen

2.  Goal
 What is noise ?
 What is Noise Cancellation ?
 Simple Idea .
 Applications
 Adaptive Filter
 Adaptive Algorithm( LMS )
 Simulation
 Conclusion

3. Goal
The goal of the project is for
.

4. Equipment Lists
Design Tools
 MATLAB/Simulink
 Xilinx System Generator

5. Design Approach
Simulation
 MATLAB
. Least Mean Square (LMS)
 Xilinx
. Lease Mean Square (LMS)

6. What is noise?
 Noise consists of unwanted waveforms that can interfere
with communication.
 Sound noise: interferes with your normal
hearing
.Loud noises
.Subtle noise
.White noise (AWGN)

7. What is Noise Cancellation?
 Noise cancellation is a method to reduce or completely cancel out
undesirable sound.
 call Active Noise Cancellation .
 Noise cancellation tries to 'block' the sound at the source instead of
trying to prevent the sounds from entering our ear canals .
 These technologies are in their early stages.
 The hope is that one day that these technologies can be used to
minimize all sorts of unwanted sounds around us

8. Simple Idea
 Cancellation processes depend on simple principle
 adding two signals with the same
 amplitude and opposite phase the result will be zero
signals.
(H)

9. Simple wave cancellation

10. Applications
Headsets (headphone)
Honda cars.
Space satellite antennas.
 Use in apartment.
 Noise Muter

11. Adaptive Noise Cancelling
 Adaptive noise cancelling
- An approach to reduce noise based on reference noise
signals
- System output
- The LMS algorithm
K
k
 
u t s t n t w k n t k
( )  ( )  ( )  ( ) ( 
)
0 1 1 ( ) ( ) ( ) 1 w k u t n t  k

13. Adaptive filter
 nonlinear and time-variant .
 adjust themselves to an ever-changing environment .
 changes its parameters so its performance improves
through its surroundings.

14. Adaptive Filter
Output
signal
Input
signal
Adaptive
algorithm
Criterion of
performance
Filter
structure
 The coefficients of an adaptive filter change in time

15. Block diagram of adaptive system
No(n) S(n)+No(n)
?
Primary
signal
d(n)
N1(n)
Reference
signal
y(n)
output
e(n)
adaptive

16. Adaptive algorithm
An adaptive algorithm is used to estimate a time varying
signal.
By adjusting the filter coefficients so as to minimize the error.
There are many adaptive algorithms like Recursive Least
Square (RLS),Kalman filter,
but the most commonly used is the Least Mean Square (LMS)
algorithm.

17. LMS Adaptive Algorithm
 Introduced by Widrow & Hoff in 1959.
 Simple, no matrices calculation involved in the adaptation.
 In the family of stochastic gradient algorithms.
 Approximation of the steepest – descent method
 Based on the MMSE criterion.(Minimum Mean square Error)
 Adaptive process containing two input signals:
• 1.) Filtering process, producing output signal.
• 2.) Desired signal (Training sequence)

18. Stability of LMS
 The LMS algorithm is convergent in the mean square if and only if
the step-size parameter satisfy
 Here max is the largest eigenvalue of the correlation matrix of the
input data
 More practical test for stability is

19. LMS Algorithm Steps
Filter output
Estimated error

      
y n  u n 
k w n
k 
1
0
*
M
k
en dn yn

20. The LMS Equation
 The Least Mean Squares Algorithm (LMS) updates each coefficient
on a sample-by-sample basis based on the error e(n).
w (n 1) ( ) ( ) ( ) k w n e n x n k k    
 This equation minimises the power in the error e(n).
 The value of μ (mu) is critical.
 If μ is too small, the filter reacts slowly.
 If μ is too large, the filter resolution is poor.
 The selected value of μ is a compromise.

21. LMS algorithm
 Estimates the
solution to the
Widrow -Hoff
equations using gradient
descent method which
Finds minima by
estimating
the gradient.
X(n)
Transversal
Filter
C(n)
LMS
Y(n)
d(n)
e(n)
is the step size

22. Cont..
e(n)
Adaptive
filter
Unknown
system
X(n)
y(n)
d(n)
filtering operation with the
previous version of the coefficients.
Compare the computed output
with the expected output.
Update the coefficients using
the following computation.

23. Cont..
LMS algorithm
The most widely used real time adaptive filtering algorithm
Convergence speed of the LMS algorithm
 Controlled by the spread of eigenvalues of the autocorrelation
matrix of the input data
 Enhanced by reducing the eigenvalue spread

24. Advantages
 low computational complexity
 simple to implement
 allow real-time operation

26. Simulation
Xilinx System Generator Output

27. Conclusion
 Active noise cancellation is a method to cancel out
undesirable sound in real time
 The adaptive filter is used to estimate the error in
noisy wave
 Many algorithms are used in adaptive filter like LMS
RLS & MSE and the better is LMS .