한국통신학회 2015년 하계종합학술발표회

NLMS Adaptive Beamforming Algorithm for smart antenna

Prakash Thapa, Jinho Yang, Wootaek Lim, Kyoungwoo Park and Seong Ro Lee
Mokpo National University, South Korea
e-mail : [email protected], [email protected]

Abstract
The smart antennas are widely used for wireless
communication because it has an ability to
increase the coverage capacity of a
communication system. The main purpose of the
smart antenna system is the selection of smart
algorithms for adaptive array. By using beam
forming algorithms the weight of the antenna
array can be adjusted to form certain amount of
adaptive beam to track corresponding users
Fig. 1: Smart Antenna Array System
automatically and to minimize interference
arising from others by introducing nulls in their
2. NLMS Algorithm
direction. On this paper, one of the beam
The NLMS Algorithm which is also known as
forming algorithms called Normalized Least
projection algorithm is very useful method for
Mean Square Algorithm (NLMS) is presented.
adapting the coefficients of a finite impulse
The efficiency of NLMS algorithm is compared
response filter for number of signal processing
on the basis of normalized array factor and
and control application.
mean square error (MSE) for mobile
One of the drawbacks of Least Mean Square
communication.
Algorithm is more sensitive to the step size of its
input signal. This drawback of LMS algorithm
Keywords: Smart Antenna, NLMS Algorithm
can be reduced by using NLMS Algorithm. So
NLMS algorithm is used for stable calculation
1. INTRODUCTION:
and faster convergence [4][5][6]]7].
Smart antennas are also known as adaptive array
According to the LMS algorithm, the weight of
antennas. Smart antennas are used to increase
the output signal is given by
the efficiency of digital wireless system [1]
which combines multiple antenna elements with w(n+1 = w(n) + (1)
a signal processing capability to optimize its The NLMS algorithm can be formulated by
radiation and reception pattern automatically in natural modification of the LMS algorithm
response to the signal environment [2].Smart which is given by the following equation
antennas enhance the capacity of wireless w(n+1) = w(n) +
networks through a combination of diversity
gain, array gain and interference suppression by or, w(n+1) = w(n) + (2)
reducing multipath and co-channel interference.
It increased the capacity of data rates for a given To prevent weight vectors from diverging and
number of user or more users for a given data makes the algorithm more stable and faster
rate per user. These antennas are used at the base converging a positive constant ‘ɛ’ may be added
stations which radiates narrow beam to serve to the denominator of the input vector u(n). So
different users [3]. the weighted vector can be computed by

- 275 -
 한국통신학회 2015년 하계종합학술발표회

w(n+1) = w(n) + (3) S(t) = ejsin(ɷt) (11)

or, Where, f is the frequency in Hertz (Hz).

w(n + 1) = w(n) + μ(n)e*(n) (5)
where
1) Effect of Number of Elements on Array
µ(n) = (6) Factor
Where, the NLMS algorithm reduces the step
size ‘ɛ’ to make large changes in the update
weight vectors. This prevents the update weight
vectors from diverging and makes the algorithm
more stable and faster converging than when a
fixed step size is used.

3. Implementation of the NLMS algorithm

The practical implementation on NLMS
algorithm is similar to the LMS algorithm which Fig.2: Array factor plot for NLMS algorithm
can be used as an extension of the standard LMS with AOA for different array element
algorithm.
The NLMS algorithm requires these steps in the Figure 2 shows the array factor plot of NLMS
following order for each iteration. algorithm. In NLMS algorithm space between
1. The output of the adaptive filter is the array element play an important role in beam
calculated by forming technique and it is taken as 0.5 lamda.
y(n) = wH(n)u(n) (7) Where number of array element 5, 9 and 12,
2. An error signal is calculated by desired user is arriving at 0 degree, interference
e(n) = d(n) − y(n) (8) angle of -30 degree by keeping spacing distance
3. The step size value for the input vector is “d” constant.
calculated by The simulation result shows that when number
µ(n) = (9) of array element increasing from 5 to 12 the
beam width become narrow and the number of
3. The updated weights for the next side lobes goes on increasing. But the level of
iteration calculated side lobes is low as compared to those generated
w(n + 1) = w(n) + μ(n)e*(n) (10) by small number of elements.
Each iteration of the NLMS algorithm requires
3N+1 multiplication, which is only N more than 2).Effect of Spacing between Elements on
the standard LMS algorithm. This is an Array Factor
acceptable increase considering the gains in
stability and echo attenuation achieved.

4. Simulation Results
The simulation of NLMS algorithm is carried
out by using MATLAB to illustrate how various
parameters such as number of antenna elements
(N), inter element spacing (d) and variation in
SNR parameter affect the beam formation and
convergence of algorithm. The desired signal is
phase modulated with SNR 10dB, used for Fig3: Normalized array factor plot for NLMS
simulation purpose. It is given by algorithm with AOA for different distance

- 276 -
 한국통신학회 2015년 하계종합학술발표회

MSE is decreases and conversed it after 20

Figure 3 shows the performance of NLMS iteration.
algorithm when the distance between the array
elements is 0.125, 0.25 and 0.5 and the desired 5. CONCLUSION
user is arriving angle at 00 interference is at an
angle of -300 and number of array elements are In this paper NLMS adaptive beam forming
14. algorithms is discussed. This algorithm is used
The simulation result shows that array element in smart antenna array system in coded form, to
produces narrow beam when inter element enhance mobile communication performance.
spacing is increased. When the separation The sensitivity of LMS algorithm to the step size
distance between array elements is equal to can be reduced by using NLMS algorithm.
length granting lobs are created. It is also determined from the simulation results
that the performance of NLMS algorithm is
better to minimize MSE for different number of
elements and for different spacing maintained
between elements using performance function of
the algorithm that minimized the average power
in the error signal. Therefore, NLMS algorithm
must be the better option to avoid interference
and to increase the capacity of mobile
communication system.

ACKNOWLEDGMENT

Fig.4:LMS output of desired signal with This work was supported by Priority
Phase Research Centers Program through the
National Research Foundation of
The figure 4 shows the graph of signal versus Korea(NRF) funded by the Ministry of
number of iteration. In which array acquire and Education, Science and Technology(2009-
track after 20 iterations. 0093828) and by the MSIP(Ministry of
Science, ICT and Future Planning), Korea,
under the the C-ITRC(Convergence
Information Technology Research Center)
(IITP-2015-H8601-15-1006) supervised by
the IITP(Institute for Information &
communications Technology Promotion)

REFERENCES
[1] RAO, AMARA PRAKASA, and NVSN
SARMA. "Adaptive Beam forming Algorithms for
Smart Antenna Systems." methods 4, no. 8
(2014): 10-11.
Fig.5:Plot of MSE versus LMS output [2] Joshi, Revati, and Ashwinikumar Dhande.
"ADAPTIVE BEAMFORMING USING LMS
Figure 5 shows the relationship between mean ALGORITHM."
square error and number of iteration. In which [3] Kumar, N. A., Rao, S. R., & Rao, V. S. Analysis
of Performance Improvement in Adaptive Beam

- 277 -
 한국통신학회 2015년 하계종합학술발표회

Forming Using RLMS Algorithm in Smart Antenna

systems (SAS).
[4] Kumar, N. Anil, SV Rama Rao, and V. Srinivasa
Rao. "Analysis of Performance Improvement in
Adaptive Beam Forming Using RLMS Algorithm in
Smart Antenna systems (SAS)."
[5] Haykin, Simon S. Adaptive filter theory. Pearson
Education India, 2007.]
[6]Kumar, KR Shankar, and T. Gunasekaran.
"Performance analysis of adaptive beam forming
algorithms for micro strip smart
antennas." International Journal of Computing
Science and Communication Technologies 2.1 (2009).
[7] Yasin, M., and Pervez Akhtar. "Performance
Analysis of LMS and NLMS Algorithms for a Smart
Antenna System." International Journal of Computer
Applications 4.9 (2010): 25-32.

- 278 -