Beam steering in smart antennas by using low complex

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 02 Issue: 10 | Oct-2013, Available @ http://www.ijret.org 545
BEAM STEERING IN SMART ANTENNAS BY USING LOW COMPLEX
ADAPTIVE ALGORITHMS
Amarnadh Poluri1
, Ashish kumar2
1
Student, 2
Asst Professor, Department of Electronics and Communications Engineering, Vignan's Institute Of Information
Technology, Duvvada, Visakhapatnam, Andhra Pradesh, India.
amarnadh401@gmail.com,.ashishkumar.ambastha@gmail.com
Abstract
Array antenna systems are often used to enhance the received signal to interference and noise ratio when the signal operates in
heavily jammed environment. Proper modeling of the received data at different antenna elements is important when evaluating the
performance of this system, especially when both the signal and interference have wide frequency bands.. The antenna output is the
linear combination of data from all the antenna elements. In conventional narrowband beam forming, time sequences at different
antenna elements are related by some fixed phase shift. The phase shift is determined by the wave forms direction of arrival (DOA).
In this paper, an efficient method for the pattern synthesis of the linear antenna arrays with the prescribed nulling and steering lobe is
presented. The proposed method is based on Least Mean Square (LMS) algorithm provide a comprehensive and detailed treatment of
the signal model used for beam forming, as well as, describing adaptive algorithms to adjust the weights of an array. In order to
improve the convergence rate of LMS algorithm in smart antenna system, in this paper we proposes a new normalized LMS (NLMS)
algorithm, This new algorithm can be treated as a block based simplification of NLMS algorithm which gives satisfactory
performance in certain applications in comparison with conventional NLMS recursion, i.e., BBNLMS algorithm. By taking advantage
of spatial filtering, the proposed scheme promises to reduce the bandwidth required for transmitting data by improving convergence
rate. The performance of the BBNLMS algorithm in the presence of Multi-path effects and multiple users is analyzed using MATLAB
simulations. The simulations when compared to that of the LMS algorithm, the results suggest that BBNLMS algorithm can improve
the convergence rate and lead to better system efficiency.
Keywords: BBNLMS, Convergence Rate, DOA, LMS, NLMS, Smart antenna.
----------------------------------------------------------------------***------------------------------------------------------------------------
1. INTRODUCTION
The performance of a communication system in an
interference environment can be severely degraded if the
received signal to interference ratio is unacceptably low.
Smart antenna is the critical technique of the third mobile
communication, while the core of smart antenna is adaptive
algorithm research. Smart antennas can be used to achieve
different benefits. Among those the most important is higher
network capacity [1], [2] by precise control of signal null
equality and mitigation of interference combine to frequency
reuse reduction distance improving capacity. The term
adaptive antenna is used for a phased array when the weight of
each element is applied in a dynamic fashion. The amount of
weighting on each channel is not fixed at the time of the array
design, but rather decided by the system at the time of
processing the signals to meet required objectives.
In other words, the array pattern adapts to the situation and the
adaptive process is under control of the system. For example,
consider the situation of a communication system operating in
the presence of a directional interference operating at the
carrier frequency used by the desired signal, and the
performance measure is to maximize the output signal to noise
ratio (SNR), in such systems the desirable is output SNR
should be maximized by cancelling the directional interference
using optimal antennas. The antenna pattern in this case has a
main beam pointed in the desired signal direction, and has a
null in the direction of the interference. Assume that the
interference is not stationary but moving slowly. If optimal
performance is to be maintained, the antenna pattern needs to
adjust so that the null position remains in the moving
interference direction.
In this paper we propose a simple and novel adaptive
algorithm for steering the antenna beam electronically.
Generally LMS algorithm is widely used in adaptive filter due
to its relatively low computational complexity, good stability
properties, and relatively good robustness against
implementation errors. However, the least mean square (LMS)
algorithm has poor convergence rate, which reduces the
system performance. In order increase the convergence rate,
LMS algorithm is modified by normalization, which is known
as normalized LMS (NLMS)[8]. In our work NLMS algorithm

__________________________________________________________________________________________
to increase the convergence rate and reducing computational
complexity Block processing of data in adaptive filters
significantly reduces computational burden and improves
convergence performance [9]. In literature several smart
antenna processing techniques were presented [10]-[17], to the
best of author’s knowledge reduction of computational
complexity in the array processing is not addressed. Finally in
our simulations we considered two received signals and three
directions of arrivals (DOAs). Simulation results confirm that
the proposed BBNLMS based beam steering is superior than
conventional LMS algorithm in terms of complexity and
convergence rate. Normalization of step size in LMS
algorithm improves the convergence rate and decreases excess
mean square error, by exploiting this we have implemented
beam forming using NLMS algorithms, i.e., NLMS algorithm
and BBNLMS algorithm.
2. ADAPTIVE BEAM FORMING
In certain applications the gain of a single antenna may not
sufficient, array antennas plays a vital role in such situations.
In array antennas the beam can be steered in two ways, ie.,
mechanical steering and electronic steering. Adaptive beam
forming can be performed in many ways using adaptive
algorithms. Several adaptive algorithms are presented in
literature. Most of the algorithms are concerned with the
maximization of the SNR. A functional diagram of an
adaptive array system is shown in fig 1
Fig-1: Functional block diagram of an adaptive array system
Adaptive array systems can locate and track signals (users and
interferers) and dynamically adjust the antenna pattern to
enhance reception while minimizing interference using signal
processing algorithms.
After the system down converts the received signals to
baseband and digitizes them, it locates the signal of interest
(SOI) using the DOA algorithm, and it continuously tracks the
SOI and signal not of interest (SNOI)s by dynamically
changing the complex weights (amplitudes and phases of the
antenna elements). Basically, the DOA computes the
direction-of-arrival of all the signals by computing the time
delays between the antenna elements, and afterward, the
adaptive algorithm, using a cost function, computes the
appropriate weights that result in an optimum radiation
pattern. Because adaptive arrays are generally more digital
processing intensive and require a complete RF portion of the
transceiver behind each antenna element, they tend to be more
expensive than switched-beam systems. Adaptive arrays
utilize sophisticated signal-processing algorithms to
continuously distinguish between desired signals, multipath,
and interfering signals, as well as calculate their DOA. This
approach updates its transmit strategy continuously based on
changes in both the desired and interfering signal locations.
In adaptive beam forming techniques, two main strategies are
distinguished. The first one is based on the assumption that
part of the desired signal is already known through the use of a
training sequence. This known signal is then compared with
what is received, and the weights are then adjusted to
minimize the Mean Square Error (MSE) between the known
and the received signals. In this way, the beam pattern can be
adjusted to null the interferers. This approach optimizes the
signal-to-interference ratio (SIR), and is applicable to non-
line-of-sight (NLOS) environments. Since the weights are
updated according to the incoming signals, not only the
interference is reduced but the multipath fading is also
mitigated. In the second one, the directions of arrivals from all
sources transmitting signals to the array antenna are first
identified. The complex weights are then adjusted to produce a
maximum toward the desired angle and null toward interfering
signals. This strategy may turn out to be deficient in practical
scenarios where there are too many DOAs due to multi paths,
and the algorithms are more likely to fail in properly detecting
them. This is more likely to occur in NLOS environments
where there are many local scatters close to the users and the
base station, thus resulting in a wider spread of the angle of
arrival. Another significant advantage of the adaptive antenna
systems is the ability to share spectrum. Because of the
accurate tracking and robust interference rejection capabilities,
multiple users can share the same conventional channel within
the same cell. System capacity increases through lower inter-
cell frequency reuse patterns as well as intra-cell frequency
reuse.
3. ADAPTIVE ALGORITHMS
An adaptive filter is a transversal filter trained by an adaptive
algorithm. The algorithm updates the weights at each iteration
by estimating the gradient of the quadratic MSE surface and
then moving the weights in the negative direction of the
gradient by a minute amount. The constant that determines
this amount is referred to as the step size (µ). When this step
size is small enough, the process leads these estimated weights
to the optimal weights. The convergence and transient
behavior of these weights along with their covariance
characterize the LMS algorithm and the way the step size and

__________________________________________________________________________________________
process of gradient estimation affect these parameters are of
great practical importance.
3.1 The Least Mean Square (LMS) Algorithm:
The LMS algorithm is probably the most widely used adaptive
filtering algorithm, being employed in several communication
systems. It has gained popularity due to its low computational
complexity and proven robustness. It incorporates new
observations and iteratively minimizes linearly the mean-
square error. This algorithm is based on the knowledge of
arriving signal. The knowledge of the received signal
eliminates the need for beam forming, but the reference can
also be a vector which is somewhat correlated with the
received signal. As shown in Figure 1, an adaptive beam
former consists of multiple antennas, complex weights, the
function of which is to amplify or attenuate and delay the
signals from each antenna element, and a summer to add all of
the processed signals, in order to tune out the signals not of
interest, while enhancing the signal of interest. Hence, beam
forming is some time referred to as spatial directions are
filtered out, while others are amplified. The response of the
LMS is as follows,
w’ (n+1) =w’ (n) +µu (n) e*(n)
Fig-2: Block diagram of adaptive transversal filter
3.2 Normalized Least Mean Square (NLMS)
Algorithm:
In structural terms, the normalized LMS filter is exactly the
same as the standard LMS filter. In the standard form of a
least mean square (LMS) filter studied in earlier the
adjustment applied to the tap-weight vector of the filter at
iteration n+1 consists of the product of three terms.
• The step-size parameter µ, which is under the
designer’s control.
• The tap-input vector u(n),which is supplied by a
source of information.
• The estimation error e(n) for real-valued data, or its
complex conjugate (n) for complex-valued data,
which is calculated at iteration n.
The adjustment is directly proportional to the tap-input vector
u(n). Therefore, when u(n) is large, the LMS filter suffers
from a gradient noise amplification problem. To overcome this
difficulty, we may use the normalized LMS filter. In
particular, the adjustment applied to the tap-weight vector at
iteration n+1 is “normalized” with respect to the squared
Euclidean norm of the tap-input vector u(n) at iteration n-
hence the term “normalized”. The normalized LMS filter is
manifestation of the principle of minimal disturbance, which
may be stated as follows. From one iteration to the next, the
weight vector of an adaptive filter should be changed in a
minimal manner, subject to a constraint imposed on the
updated filter’s output. The response of the NLMS algorithm
is as follows,
(n+1) = (n) + u(n)*e(n).
Here e(n) = d(n) - H(n) u (n),
3.3 Block based normalized LMS algorithm
(BBLMS):
In the conventional and normalized LMS algorithms described
earlier, the tap weights (free parameters) of a finite-duration
impulse response (FIR) filter are adapted in the time domain.
Recognizing that the Fourier transforms maps time-domain
signals into the frequency domain and that the inverse Fourier
transform provides the inverse mapping that takes us back into
the time domain, we see that it is equally feasible to perform
the adaption of filter parameters in the frequency domain. In
such a case, we speak of frequency –domain adaptive filtering
(FDAF),There are two main reasons for seeking adaption in
the frequency domain in one form or another, Self-
orthogonalizing adaptive filtering, mechanized in a different
way from that is used to improve the convergence
performance of the conventional LMS algorithm. And also a
more uniform convergence rate is attained by exploiting the
orthogonality properties of the discrete Fourier transform
(DFT) and related discrete transforms.

__________________________________________________________________________________________
Fig-3: Block Based Normalized LMS Algorithm
Fig-4: Block-based adaptive filter
In a block adaptive filter, the incoming data sequence u(n) is
sectioned into L-point blocks by means of a serial-to –parallel
converter, and the blocks of input data so produced are applied
to an FIR filter of length M, one block at a time. The tap
weights of the filter proceeds on a block-by-block basis rather
than on a sample-by-sample basis as in the conventional LMS
algorithm.
4. SIMULATION RESULTS
The performance of the LMS, NLMS, BB-NLMS algorithms
are examined in terms of convergence rate, number of active
taps used and beam pattern. In order to replicate realistic
mobile environments, for those simulations with more than
one multipath, each multipath experience a different gain,
which contains both amplitude and phase components. The
results will be accompanied by discussions of the observations
made and in addition comparisons will be made to that of
results achieved. The mean square error for various algorithms
is shown in Fig. From the figure is clear that the error
amplitude decreases as we are moving from LMS to
BBNLMS algorithm. By keen observation of Fig. we can say
that convergence starts for LMS approximately after 70th
sample, where as for NLMS algorithm it is approximately
after 30th sample, superiorly for BBNLMS convergence starts
approximately after 15th sample. More ever among all
algorithms BBNLMS has minimum error level. The main goal
of this project is to develop and study the effects of the
proposed algorithms for Smart Antenna systems. To critically
examine the performance of the proposed algorithm, the
following test cases were designed:
• One White signal with one DOA
• One White Signal with 3 DOA
• Two White Signals with 1 DOA each
• Two White Signals with Three DOAs each different
algorithm.
The selection of test cases enables us to investigate multipath
and multi-user effects in mobile communication system. The
first simulation investigated was the reception of one signal
with one path, which arrives at the base station at angle of
600. A gain with amplitude of 0.5 was introduced to the input
signal as it was propagated to the antenna. The simulation was
conducted for the following threshold control, α, values of 0.1,
0.5 and 1.0.This section we consider three cases. These are
Beam forming using LMS algorithm: one white signal with
one DOA The convergence characteristics of the three beam
forming algorithms are shown in figure 5.1. From the figure it
is clear that NLMS convergence faster than LMS. Whereas
BBNLMS convergence faster than NLMS. Normalized
algorithms convergences faster than conventional LMS
algorithm because of the normalization term in the
experiences large value in the denominator
Summary of the Block Based Normalized LMS
Algorithm
Parameters:
M = number of taps (i.e., filter length )
=adaption constant
0 < < 2 ,
Where
= error signal power
= input signal power
D(n) = mean-square deviation
Initialization. If prior knowledge about the tap-weight
vector (n) is available, use that knowledge to select an
appropriate value for (0). Otherwise, set (0)=0.
Data
Given:u(n) = M-by-1 tap input vector at time n
d(n) = desired response at time step n
To be computed: (n+1) = estimate of tap-weight vector
at time step n + 1
Computation: For n= 0,1,2…..,compute
e(n) = d(n) - H(n) u (n),
(n+1) = (n) +

__________________________________________________________________________________________
Fig-5: Typical Received signal errors of various signals with different DOA’s for LMS,NLMS, BBNLMS algorithms
LMS NLMS BBLMS
1
WHITE
SIGNAL 1
DOA
1
WHITE
SIGNAL 3
DOA
2
WHITE
SIGNAL 1
DOA
2
WHITE
SIGNAL 3
DOA

__________________________________________________________________________________________
Fig-6: Polar plot of the various signals with different DOA’s for LMS, NLMS, BBNLMS algorithms
LMS NLMS BBLMS
1
WHITE
SIGNAL
1
DOA
1
WHITE
SIGNAL
3
DOA
2
WHITE
SIGNAL
1
DOA
2
WHITE
SIGNAL
3
DOA

__________________________________________________________________________________________
CONCLUSIONS
In this paper several beam forming algorithms are presented
with some simulations and results. Using NMLS and
BBNLMS algorithms, compared to LMS algorithm, a 40%
and 55% increase in overall convergence rate is achieved in a
multi user multipath environment. With regards to beam
patterns, NLMS and BBNLMS algorithms are able to steer
beams in the direction of the desired signal and place nulls
elsewhere.
REFERENCES
[1] Carl B. Dietrich, Jr., Warren L. Stutzman, Byung-Ki Kim,
and Kai Dietze, “Smart Antennas in Wireless
Communications: Base- Station Diversity and Handset Beam
forming” lEEE Antennas and Propagation Magazine, Vol. 42,
No. 5, October 2000.
[2] Michael Chryssomallis, “Smart Antennas” lEEE Antennas
and Propagation Magazine, Vol. 42, No. 3, June 2000.
[3] G.S.N. Raju, Antennas and Wave propagation, Pearson
International Publications, Singapore,2010.
[4] R. E. Collin and F. J. Zucker, Antenna Theory, Mc-Graw
Hill, New York, 1969.
[5] R. E. Collin, Antennas and Radiowave Propagation, Mc-
Graw Hill, New York, 1985.
[6] Frank Gross, “Smart Antenna For Wireless
Communication” Mcgraw-hill, September, 2005
[7] Angeliki Alexiou and Martin Haardt, “Smart Antenna
Technologies for Future Wireless Systems: Trends and
Challenges”, IEEE Communication Magazine, Vol. 42, No.9,
pp. 90-97, September 2004.
[8] Bernard Widrow, Samuel D. Stearns, “Adaptive Signal
Processing”, Pearson Education Asia, Second Indian Reprint,
2002.
[9] Symon Haykin, “Adaptive filter theory”, Forth edition,
Pearson Education, Asia, 2002.
[10] S. D. Blostein and H. Leib, “Multiple antenna systems:
Role and impact in future wireless access,” IEEE Comm.
Mag., vol. 41, no. 7, pp. 94-101, July 2003.
[11] J. H. Winters, “Smart antennas for wireless systems,”
IEEE Personal Comm., vol. 5, no. 1, pp. 23-27, Feb. 1998.
[12] A. J. Paulraj, D. Gesbert, and C. Papadias, “Smart
antennas for mobile communications,” Encyclopedia for
Electrical Engineering, John Wiley Publishing Co., 2000, pp.
1-15.
[13]V. Kalinichev, “Analysis of beam-steering and directive
characteristics of adaptive antenna arrays for mobile
communications,” IEEE Antennas Propagat. Mag., vol. 43, no.
3, June 2001.
[14] E. Charpentier and J. Laurin, “An implementation of a
directionfinding antenna for mobile communications using a
neural network,” IEEE Trans. Antennas Propagat., vol. 47, no.
7, July 1999
[15] M. S. Choi, G. Grosskopf, D. Rodhe, B. Kuhlow, G.
Przyrembel, and H. Ehlers, “Experiments on DOA-estimation
and beam forming for 60 GHz smart antennas,” Proc.
Vehicular Technology Conference, 2003 (VTC 2003-Spring),
21-24 April 2003, Jeju, Korea, vol. 2, pp. 1041-1045.
BIOGRAPHIES
Amarnadh Poluri has obtained B.Tech degree
from Prasad Institute of Technology and
Sciences affiliated to JNTUK in the year 2011.
Now he is pursuing M.Tech Degree in the
Department of Electronics &
Communications, Vignan's institute of
Information and Technology, Visakhapatnam. He is interested
in the fields of wireless and mobile communication, signal
processing.
Ashish Kumar has obtained B.Tech degree
from Trident academy of technology affiliated
to BPUT in the year 2010.he obtained M.Tech
degree from Birla Institute of Technology
Mesra, Ranchi in the year 2013. Presently he
is working as an Asst Professor in the
department of Electronics & Communication Engineering,
Vignan’s Institute of Information Technology, and
Visakhapatnam. He is interested in the fields of wireless
communication, signal processing

Beam steering in smart antennas by using low complex

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Beam steering in smart antennas by using low complex (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Beam steering in smart antennas by using low complex