MVDR an Optimum Beamformer for a SmartAntenna System in CDMA Environment
M Yasin
1
, Pervez Akhtar
2
, M Junaid Khan
3
Department of Electronics and Power Engineering
1, 2, 3
Pakistan Navy Engineering College, NUST, Karachi, PAKISTAN
Abstract:
Efficient utilization of limited radio frequencyspectrum is only possible to use smart/adaptive antennaarray system. Minimum Variance Distortionless Response(MVDR) algorithm is an option for smart antenna toexploit spatial distribution of the users and the access delaydistribution of signal paths to enhance mobile systemscapabilities for quality voice and data communication. Thispaper analyzes the performance of MVDR (blindalgorithm) and Kernel Affine Projection Algorithm(KAPA) (nonblind algorithm) for CDMA application. Forthe first time, KAPA is implemented in [1] in the context of noise cancellation but we are using it for adaptivebeamforming which is novel in this application. Smartantenna incorporates these algorithms in coded form whichcalculates optimum weight vector which minimizes the totalreceived power except the power coming from desireddirection. Simulation results verify that MVDR a blindalgorithm has high resolution not only for beam formationbut also better for null generation as compared to nonblindalgorithm KAPA. Therefore, MVDR is found more efficientBeamformer.
Keywords: Adaptive Filtering, Minimum Variance Distortionless Response (MVDR) Algorithm and Kernel Affine Projection Algorithm (KAPA).
I. INTRODUCTIONSince Radio Frequency (RF) spectrum is limited and itsefficient use is only possible by employingsmart/adaptive antenna array system to exploit spatialdistribution of the users and the access delay distributionof signal paths to enhance mobile systems capabilitiesfor data and voice communication. The name smartrefers to the signal processing capability that forms vitalpart of the smart/adaptive antenna system which controlsthe antenna pattern by updating a set of antenna weights.Smart antenna, supported by signal processingcapability, points narrow beam towards desired users butat the same time introduces null towards interferers, thusimproving the performance of mobile communicationsystems in terms of channel capacity, extending rangecoverage, tailoring beam shape and steering
multiplebeams to track many mobiles electronically. Consider asmart antenna system with elements equally spacedand user’s signal arrives from an angle
Ne
()
d
0
Φ
asshown in Fig 1 [2].
Fig.1. Smart/adaptive antenna array system
Adaptive beamforming scheme that is MVDR (blindalgorithm) and KAPA (nonblind algorithm) is used tocontrol weights adaptively to optimize signal to noiseratio (SNR) of the desired signal in look direction
0
Φ
.The array factor for
(
elements equally spacedlinear array is given by
)
Ne
()
d
21((cos))0
().
d N jnnn
AFAe
π
−Φ+αλ=
Φ =
∑
(1)
where
α
is the inter element phase shift and is describedas:
00
2cos
d
π λ
−α = Φ
(2)
and
0
Φ
is the desired direction of the beam.In reality antennas are not smart; it is the digitalsignal processing, along with the antenna, which makes
.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 4, July 201099http://sites.google.com/site/ijcsis/ISSN 1947-5500
the system smart. When smart antenna is deployed inmobile communication in Code Division MultipleAccess (CDMA) environment in which different codesare assigned to different users, it radiates beam towardsdesired users only. Each beam becomes a channel, thusavoiding interference in a cell. Because of these, eachcoded channel reduces co-channel interference, due tothe processing gain of the system. The processing gain(PG) of the CDMA system is described as:
10log(/)
b
PGBR
=
(3)
where
B
is the CDMA channel bandwidth and
b
R
isthe information rate in bits per second
.
If a single antenna is used for CDMA system, then thissystem supports a maximum of 31 users. When an arrayof five elements is employed instead of single antenna,then capacity of CDMA system can be increased morethan four times. It can be further enhanced if array of more elements are used [4] [5] [7] [8] [9].The rest of the paper is organized as follows: Section 2introduces MVDR algorithm with simulation results.KAPA with simulation results are presented in section 3.Finally the concluding remarks of this work are providedin section 4.II. MVDR ALGORITHM
A. Theory
MVDR is a direction of arrival (DOA) estimationmethod in which the direction of a target signal isparameterized by the variable
0
Φ
and all other sourcesare considered as interferences. In beamformingliterature, this estimation method is called MVDR inwhich the weights of the smart antenna array are chosenso as to pass the desired directional signal without anydistortion (preserving unity gain) whereas to suppress theinterferers maximally. MVDR is a blind algorithm whichdoesn’t require a training signal to update its complexweights vector but utilizes some of the known propertiesof the desired signal. Assuming that is thesteering vector and is independent of the data obtainedfrom sensors. The data obtained from sensors isgiven by
0
(
s
Φ
)
nn
011
(){,,........}
n
unuuu
−
=
(4)
MVDR beamformer output in the look directionwith input signal is described as:
()
yn
()
un
()(1)()
T
ynwnun
= −
(5)
The autocorrelation matrix
R
of the sensor data vectoris given by
{()()}
T
REunun
=
(6)
where
E
is the expectation operator. The output powerfor each looking direction is defined by
2
{}{()()}
TTT
w
(7)
PEywEununwwR
= = =
In adaptive beamforming algorithm, the weight vectorsare correlated with the incoming data so as to optimizethe weight vectors for high resolution DOA detection ina noisy environment. MVDR is graded an adaptivebeamformer, therefore, some constraints are imposed as, ensures that desired signals are passed with unitygain from looking direction whereas the output powercontributed by interfering signals from all otherdirections are minimized using a minimization criterionas described in .
(8)(9)
T
wsg
=
(8)
where
g
denotes the gain of MVDR which is equal tounity.
(
T w
)
MinPwRw
=
constrained to
(9
1
T
ws
=
)
Solving by Lagrange multiplier method, we obtainthe weight vector as:
(9)
11
T
RswsRs
−−
=
(10)
When we put the value of
(1
into, the outputpower
0)(9)
0
(
P
)
Φ
for a single looking direction is obtainedas:
01
1()
T
PsRs
−
Φ =
(11)
MVDR algorithm computes the optimum weight vectorbased on the sampled data that ultimately forms abeampattern and places null towards interferers [3] [6].
B. Simulation Results
Computer simulation is carried out, to illustrate that howvarious parameters such as number of elements
()
Ne
.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 4, July 2010100http://sites.google.com/site/ijcsis/ISSN 1947-5500
and element spacing, affect the beam formation. Thesimulations are designed to analyze the properties of MVDR and KAPA algorithms. The desired signal isphase modulated, used for simulation purpose. It is givenby
()
d
sin(2)
()
j
Ste
∗π∗ƒ∗
=
t
(12)
where
f
is the frequency in Hertz.
1) Effect of number of elements on array factor
Uniform linear array is taken with different number of elements for simulation purpose. The spacing betweenarray elements is taken as
( )
.
/8
λ
-100 -80 -60 -40 -20 0 20 40 60 80 100-60-50-40-30-20-100Angle of Arrival(degree)
N o r m a l i z e d A r r a y F a c t o r ( d B )
Ne=20Ne=15Ne=10
Fig.2. Normalized array factor plot for MVDR algorithm with AOA fordesired user is 0 degree and
- 30 degrees for interferer
with constantspace of
(
between elements
)
/8
λ
Angle of Arrival (AOA) for desired user is set at 0degree and for interferer at -30 degrees as shown in Fig.2 which provides deep null at -30 degrees but at the sametime forms narrow beam in accordance to number of elements.
-100-80-60-40-20020406080100-60-50-40-30-20-100Angle of Arrival(degree)
N o r m a l i z e d A r r a y F a c t o r ( d B )
Ne=20Ne=15Ne=10
Fig.3. Normalized array factor plot for MVDR algorithm with AOA fordesired user is 20 degrees and
- 20 degrees for interferer
withconstant space of
( )
/8
λ
between elements
Similarly in Fig.3, we achieved a deep nullapproximately at -20 degrees and the desired user isarriving at 20 degrees. Therefore, it is proved that for afixed spacing and a frequency, a longer array
(2
Ne
0)
=
results a narrower beam width but thishappens at the cost of large number of sidelobes.
-100-80-60-40-20020406080100-60-50-40-30-20-100Angle of Arrival(degree)
N o r m a l i z e d A r r a y F a c t o r ( d B )
Ne=20Ne=15Ne=10
Fig.4. Normalized array factor plot for MVDR algorithm with AOA fordesired user is - 10 degrees and
40 degrees for interferer
withconstant space of
( )
/8
λ
between elements
In Fig. 4, AOA for desired user is obtained at -10 degreesand deep null is shown at – 40 degrees for
/4
d
λ
=
.Again it is proved that for a fixed spacing and afrequency, a longer array results a narrowerbeam width but this happens at the cost of large numberof sidelobes.
(2
Ne
=
0)
.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 4, July 2010101http://sites.google.com/site/ijcsis/ISSN 1947-5500
Search History:
Result 00 of 00
00 results for result for
p.
MVDR an Optimum Beamformer for a Smart Antenna System in CDMA Environment
Efficient utilization of limited radio frequency spectrum is only possible to use smart/adaptive antenna array system. Minimum Variance Distortionless Response (MVDR) algorithm is an option for smart antenna to exploit spatial distribution of the users and…
(More)
209
Reads
3
Readcasts
0
Embed Views
More From This User
Notes
Load more
