Real Time Implementation of Adaptive Beam former for Phased Array Radar over DSP Kit

Muhammad Amir Shafiq, Saqib Ejaz & Nisar Ahmed
International Journal of Signal Processing (SPIJ), Volume (7) : Issue (1) : 2013 32
Real Time Implementation of Adaptive Beam former for Phased
Array Radar over DSP Kit
Muhammad Amir Shafiq amirshafiq@gmail.com
Faculty of Electronics Engineering
GIKI,
Topi, Pakistan
Saqib Ejaz saqibejaz@gmail.com
GIKI,
Topi, Pakistan
Nisar Ahmed nisarahmed@giki.edu.pk
GIKI,
Topi, Pakistan
Abstract
Mechanical positioners, rotating antennas and large size of early generation radars limited the
capability of the radar system to track laterally accelerating targets. Electronic Scanning Array
(ESA) such as used in Phased Array Radar (PAR) overcomed these limitations by providing
beam agility, good response time, variable scan rates and efficient use of energy. Early PAR
systems used analog phase shifting schemes that caused variations and component failures
resulting in overall degradation of radar performance. With the advent of new technology and high
performance embedded systems, digital beamforming has become powerful enough to perform
massive operations required for real time digital beamforming. MATLAB simulation of adaptive
beamformer is presented in this paper. Real time implementation of adaptive beamformer over
DSP kit (TMS320C6713) was also carried out and results were compared with MATLAB
simulations. GUI was also made in MATLAB for viewing results of real time implementation via
real time data exchange. Developed system can be used in digital beamforming PAR provided
array signals are routed to DSP kit through FPGA interfaced to high speed ADC’s.
Keywords: Digital Beamforming, Phased Array Radar, NLMS, Adaptive Beamformer, DSP kit
TMS320C6713.
1. INTRODUCTION
Phased Array Radar (PAR) has been used largely in communication systems and attracted the
interest of many researchers from last few decades. PAR technology is better than conventional
mechanically steered radars in terms of coverage, response time, scan rate, acquisition and
simultaneous tracking of multiple targets. Their capability to change radiation pattern
electronically, multi-beam capacity and high spatial resolution has made them attractive for
mobile communication applications as well [1].
PAR can generally be categorized as two types: Analog PAR and digital PAR. Adaptive
beamforming is integral part of digital PAR system. Digital beamforming offers advantages in
terms of power consumption, flexibility, and accuracy as compared to analog beamforming
techniques. In general, digital systems tend to consume less power in computation operations
and have programmable interface adding versatility to the system. Also digital systems provide

adaptive pattern nulling, super resolution, array element pattern correction, self calibration, and
radar power and time management [2].
Digital beamforming offers great deal of flexibility without degrading Signal to Noise Ratio (SNR).
Adaptive beamforming further improves the functionality of system by suppressing any
interference signals and offering high jamming resistance [3-4]. Adaptive beamforming can be
used to shape the radiation pattern for both reception and transmission of signal. Adaptive
beamforming optimize the beamformer response so that output contains minimal contribution due
to noise and interferers [5]. Design of transmitter and receiver of digital beam former is presented
in [6]. Simulations are carried out using 4-element linear array to demonstrate control of radiation
pattern of antenna. L-band signal processor using Field Programmable Gate Arrays (FPGA) for
digital beamforming antenna is presented in [7]. Hardware implementation in [7] offers selection
and synthesis of multiple beams formed using 16 element antenna.
In this paper, simulation and real time implementation of Adaptive Beamformer (ABF) is
presented. Implementation is based on the far-field plane wave incident on a 1x4 Uniform Linear
Array (ULA). Rest of paper is organized as follows. Problem statement is given in Section 2.
Theory of adaptive beamformer is presented in Section 3. Introduction to DSP kit and real time
implementation is given in Section 4. Experimental setup and results are also discussed in
Section 4. Finally conclusion is given in Section 5.
2. PROBLEM STATEMENT
Consider a point source P which is broadcasting in all directions shown in Fig. 1. Let  be an
angle that source subtends at ULA. ULA consists of four antennas spaced at 0.5 
θ
λ/2
x
z
y
P
FIGURE 1: 1x4 Uniform Linear Array.
The incident plane wave on ULA is given by (1) [8].
      
   nRF
nRFnn
ttx
ttxpte




cos
cos, (1)
where n = nRF , RF is a radio frequency and n is a time delay due to the nth antenna
element. This incident plane wave as received by the antennas of the ULA is modeled and
simulated in MATLAB. Our aim is to design a beamformer processor that will steer the main
beam of ULA to our desired angle. Developed adaptive beamformer processor will be used in
PAR. Simplified block diagram of PAR is shown in Fig. 2.

RF Front
End
ADC
Digital
Beamformer
Digital
Signal
processor
Display
FIGURE 2: Simplified Phased Array Radar Block Diagram.
3. ADAPTIVE BEAMFORMER
Beamforming is combination of radio signal from a set of non-directional antennas to simulate
one antenna with directional properties. Formed beam can be steered electronically using phase
shifters. Phase shifters can be adjusted to enhance or suppress reception of signal from certain
area. In this way, beamforming can also be used for direction finding applications as well [9].
In digital beamforming, phase shifting operation, amplitude scaling for each antenna element and
summation for receiving are done digitally [10]. General purpose Digital Signal Processor (DSP)
or dedicated beamforming chips are used for digital beamforming. Digital beamforming apply
different weights to adjust amplitude and phase of digital signal to form the beam in desired
direction. Amplitude and direction of beams are controlled by digital signal processor. Adaptive
beamforming not only steer main beam to desired direction but can also be used for steering nulls
in direction of jammer or interferer. Weights adapt to changes in location of interfering signal while
maintaining response towards the desired signal [11]. In adaptive beamformer, complex weights
are adapted continuously until desired results are obtained. In this way adaptive beamformer also
optimize the array pattern by controlling the weights of digital signal [3].
Multiple independent beams can be formed and steered in all direction using digital beamforming
processor [12]. Advantages of adaptive beamforming include improved dynamic range,
controlling of multiple beams, better and faster control of amplitude and phase [12]. Advanced
signal processing algorithms are used to locate and track the desired and interfering signals.
Adaptive system then dynamically minimizes interference signal and maximize desired signal
reception. PAR uses beam steering only, however beam steering and nulling both are used in
adaptive beamforming [13]. A generic adaptive beamformer is shown in Fig. 3.
The output y(k) at time instant k is given by a linear combination of the data at the M sensors
multiplied by weights vector. In vector form, it is written as (2)




1
0
*
)()()(
M
i
ii kxkwky (2)
where * represents complex conjugate. With adaptive beamformer system, we can point beam in
any direction and manipulate beam shape to optimize system performance by changing the
weights vector W. W is given by (3)

45.0°
w0 w1 w2 w3

Performance Index and
Error Calculation
Desired
Output
Array Output
Adaptive
Beamforming
Processor
FIGURE 3: A Generic Adaptive Beamformer System.
])(...)()([)(
*
1
*
1
*
0 kwkwkwkW M  (3)
Weight vector W is updated using adaptive algorithm. Weight vector W is chosen adaptively to
give desired peaks and null in radiation pattern of antenna array. Choice of adaptive algorithm
depends on speed, convergence and hardware complexity required. Most commonly used
adaptive algorithms are Least Mean Square (LMS), Normalized Least Mean Square (NLMS) and
Recursive Least Square (RLS) etc [14]. We have used NLMS for weight updation because it has
faster convergence as compared to LMS and low Mean Square Error (MSE). Also computational
complexity of NLMS is less than RLS.
3.1 NLMS
Adaptive complex weights )(*
kw i are calculated using adaptive NLMS algorithm [14]. Output of
adaptive beamformer is calculated based on moving average system i.e. output is approximated
by an adaptive Finite Impulse Response (FIR) filter. Mathematically,
   kkebeam as0 (4)
Where  kebeam is error mismatch between desired beam and digital beamformer output. Weight
update equation for NLMS is given by (5).
       
   11
1
1



kXkX
kX
kekWkW Tbeam

 (5)

Where  is learning rate,  kW is weight vector,  kX is input vector and  is very small real
number approximately equal to zero. These weights   )(*
kwkW i are obtained adaptively by
conforming the signal in accordance to the desired signal.
3.2 Simulation Results
Adaptive beamformer was simulated in MATLAB. Simulation results are shown in Fig. 4. If we
want to form a beam at 450
then we input this angle as our desired angle to adaptive beamformer
processor. NLMS adapts the weights and steer the beam to 450
. Radiation pattern at 450
desired
angle is shown in Fig. 4.
FIGURE 4: Radiation Pattern after Adaptive Beamforming.
NLMS adaptation error is shown in Fig. 5. It can be seen from the Fig. 5 that NLMS converges to
desired angle quickly and    kkebeam as0

FIGURE 5: NLMS Adaptation Error.
3.3 Direction of Arrival Estimation
For direction finding application, weights of adaptive beamformer are changed slowly and beam is
steered in space to find signal source. Angle at which quality of received signal maximizes
radiation pattern of antenna array is direction of arrival of signal [9-10].
4. REAL TIME IMPLEMENTATION
DSP processors are concerned primarily with real-time signal processing [15]. Real-time
processing requires the processing to keep pace with some external event, whereas non-real-
time processing has no such timing constraint. For non-real time system, data can be recorded
on PC and signal processing algorithms can be implemented offline. DSP based systems are less
affected by environmental conditions. DSP processors enjoy the advantages of microprocessors
[16]. They are easy to use, flexible, and economical. The DSP board which is used in this paper
is TMS320C6713 which is the floating-point DSP generation in the TMS320C6000 DSP platform.
The C6713 device is based on the high-performance, advanced Very Long Instruction Word
(VLIW) architecture developed by Texas Instruments (TI), making this DSP an excellent choice
for multichannel and multifunction applications [17]. Adaptive beamformer was successfully
implemented on TMS320C6713 DSP kit. Block diagram of experimental setup is shown in Fig. 6.

RTDX
Real Time Data
Exchange
DSP KIT TMS320C6713
NLMS
Implementaiton
Beamforming
Processor
Initialization
Signal
Generation
Beamforming
MATLAB
Beam
Radiation
Pattern
LMS Error
Plot
Desired
Angle
User Input
FIGURE 6: Block Diagram of Experimental Setup.
In current experimental setup synthetic data of ULA was generated over DSP kit and NLMS was
used to adaptive beamforming. Only input required from user is desired angle to steer the main
beam of ULA in that direction. In complete PAR system, desired angle will be controlled by signal
processor for beam steering. MATLAB was used for designing Graphical User Interface (GUI)
and plotting results of adaptive beamformer. Real Time Data Exchange (RTDX) was used for
duplex data communication between MATLAB and code composer studio. User inputs the
desired angle from GUI and after adaptive beamforming radiation pattern of ULA is plotted on
GUI. NLMS adaptation error is also shown in GUI. GUI is shown in Fig. 7.
FIGURE 7: Graphical User Interface (GUI).

We also profiled the execution of adaptive beamforming code over DSP kit. Adaptation time for
beamforming was found to be 3.31us. It can be seen from Fig. 4 and Fig. 7 that results of real
time implementation accede with MATLAB simulation results which verify the effectiveness of
DSP kit implementation. Hardware setup is shown in Fig. 8.
FIGURE 8: Hardware Setup.
Proposed hardware implementation block diagram for PAR is shown in Fig. 9. Developed system
can be used in digital beamforming PAR provided ULA signals are routed to DSP kit through
FPGA interfaced to high speed Analog to Digital Converter (ADC’s). ADC will sample the antenna
signals in reception mode and DAC’s will be used to send data to antenna elements in
transmission mode. ADC’s, Digital to Analog Converter (DAC’s) and other hardware will be
interfaced to FPGA because of its parallel processing feature. FPGA will also handle data transfer
and routing operation between DSP kit and ADC’s/DAC’s. DSP kit will perform beamforming and
signal processing operations, form desired beam, extract targets from raw data and forward
results to Plan Position Indicator (PPI) for display.
PROPOSED DIGITAL BEAMFORMING PAR
HARDWARE IMPLEMENTATION
DSP kit
Target
Extraction
PPI Display
Digital
Signal
Processing
Adaptive
Beamforming
FPGA
High Speed
ADC’s
High Speed
DAC’s
Interfacing
with other
hardware
Data
Handling &
Routing
FIGURE 9: Proposed Hardware Implementation Block Diagram for PAR.

5. CONCLUSION
Digital beamforming offers advantages in terms of power consumption, flexibility, and accuracy as
compared fixed beamforming techniques. In this paper, we have presented MATLAB simulation
of adaptive beamformer for digital phased array radar. Real time implementation of adaptive
beamformer over DSP kit was also carried out. DSP kit implementation results were also
compared with MATLAB simulation results. Experimental results verify effectiveness of
implementation. Complete solution for digital beamforming phased array radar is also proposed in
this paper. Developed system can be used for digital beamforming after interfacing high speed
ADC’s to DSP kit.
REFERENCES
1. R. Miura, T. Tanaka, I. Chiba, A. Horie and Y. Karasawa, “Beamforming Experiment with a
DBF Multibeam Antenna in a Mobile Satellite Environment,” IEEE Transactions on Antennas
and Propagation, Vol. 45, No. 4, pp. 707–714, 1997.
2. H. Steyskal, “Digital Beamforming – An Emerging Approach,” Military Communications
Conference, 1988. MILCOM 88, Conference record. Vol. 2, pp. 399-403, 1988.
3. M. Vavrda “Digital beamforming in wireless communications,” Institute of Radio Electronics,
Faculty of Electrical Engineering and Communication, Brno University of Technology,
Purkyňova, Czech Republic.
4. (2011, May.) [Online]. Available: Wikipedia, World Wide Web,
http://en.wikipedia.org/wiki/Active_Electronically_Scanned_Array
5. B. D. Veen and K. M Buckley, “Beamforming: A Versatile Approach to Spatial Filtering,” IEEE
ASSP Magazine, pp. 2-24, April 1988.
6. A. J. Rani and A.J. Lakshmi, “Phased Array Antenna and Beamforming Subsytems in Phased
Array Radar,” International Journal of Engineering Science and Technology, Vol. 2, No. 5, pp.
1253-1259, 2010.
7. T. Tanaka, I. Chiba, R. Mtura and Y. Karasawa, “Digital Signal Processor for Digital Multi
Beam Forming Antenna in Mobile Communication,” ATR Optical and Radio Communications
Research Laboratories, Japan, pp. 1507-1511, 1994.
8. D. M. Pozar, Microwave Engineering, Second Edition. John Wiley & Sons, New York, 1998.
9. S. Ejaz and M. A. Shafiq, “Comparison of Spectral and Subspace Algorithms for FM Source
Estimation,” Progress In Electromagnetics Research C, Vol. 14, pp. 11-21, 2010.
10. T. Haynes, Digital Beamforming Primer, November 2001.
11. Vecima, Beamforming in Software Defined Radio, Spectrum Signal Processing Webinar,
January 2011.
12. (2011, May.) [Online]. World Wide Web,
http://www.radartutorial.eu/06.antennas/an51.en.html
13. V. Vakilan, “Digital Beamforming Implementation of Switch-Beam Smart Antenna System by
using Integrated Digital Signal Processor and Field-Programmable Gate Array,” MS Thesis,
Faculty of Electrical Engineering, Universiti Teknologi Malaysia, May 2008.

14. A. D. Poularikas and Z. M. Ramadan, Adaptive Filtering Primer with Matlab, Taylor and
Francis Group, 2006
15. N. kehtarnavaz, Real-time digital signal processing based on the TMS320C6000, ELSEVIER,
2005.
16. R. Chassaing, Digital Signal processing and applications with the C6713 and C6416 DSK, A
John Wiley and Sons, Inc., Publications, 2005.
17. (2011, May.) [Online]. Available: Texas instruments DSP developer’s village. World Wide Web,
www.dspvillage.ti.com

Real Time Implementation of Adaptive Beam former for Phased Array Radar over DSP Kit

More Related Content

What's hot (20)

Similar to Real Time Implementation of Adaptive Beam former for Phased Array Radar over DSP Kit (20)

Recently uploaded (20)

Real Time Implementation of Adaptive Beam former for Phased Array Radar over DSP Kit