Content uploaded by Jozef Sofka
Author content
All content in this area was uploaded by Jozef Sofka on May 08, 2015
Content may be subject to copyright.
Performance of a laser communication system with acousto-optic
tracking: An experimental study
V. Nikulin, R. Khandekar, and J. Sofka
Dept. of Electrical and Computer Engineering
SUNY-Binghamton, Binghamton, NY 13902-6000
ABSTRACT
Laser communication systems hold great promise for broadband applications. This technology uses much higher-than-
RF region of the spectrum and allows concentration of the signal within a very small spatial angle, thus offering
unsurpassed throughput, information security, reduced weight and size of the components and power savings.
Unfortunately, these intrinsic advantages do not come without a price: small beam divergence requires precise
positioning, which becomes very critical at high bit rates. Complex motion patterns of the communicating platforms,
resident vibrations, and atmospheric effects are known to cause significant signal losses through the mechanisms of the
pointing errors, beam wander and other higher-order effects. Mitigation of those effects is achieved through the multiple
means of fast tracking and wavefront control. In this paper we focus on the application of a beam steering technology
and its effect on the communication performance of the system. We present the results of an experimental study of a
laser communication link subjected to pointing distortions. These distortions are generated by a special disturbance
element in the optical setup, which recreates specific operation environments with particular spectral characteristics. The
acousto-optic technology is used to build an agile tracking system to assure the maximum signal reception in spite of the
harsh operational conditions. The received communication signal is recorded and statistically analyzed to calculate the
bit-error-rates. This paper presents the synthesis of a tracking system and the experimental results characterizing the
communication performance under uncompensated pointing disturbance and with tracking.
Keywords: tracking, Bragg cell, quadrant detector, bit-error rate
1. INTRODUCTION
While RF and fiber optics still dominate global communications, free-space laser-based systems are viewed as the
technology that in the nearest future could handle most of the information transit throughout the world [1], including the
last-mile links. Laser communication employs highly directional laser beams, thus affording intrinsically high bandwidth
with small antennas and at low power, low probability of interception and detection and high resistance to jamming [2]-
[4]. However, these advantages do not come without a price: due to low divergence, laser beams must be very accurately
positioned on the target or the receiving station. In many aerospace applications when the transmitting optical platform is
placed on board of an airplane the ability to track the target is affected by the complex maneuvers performed by the
airplane, often at supersonic speed, the resident vibration of the airframe and atmospheric effects.
A laser-positioning system must satisfy the accuracy and the steering range requirements prompted by a particular
application. It is generally preferred to have wider tracking bandwidth in order to compensate most of the pointing
disturbance frequencies and to assure accurate and stable positioning of the laser beam on the receiver aperture.
Unfortunately, devices that offer such characteristics typically have small range of operation and may not be used alone.
Consequently, a laser positioning system is likely to feature a combination of a high-precision high-bandwidth steerer
performing agile tracking and compensation of resident vibrations, and a gimbal system responsible for the initial
pointing and retargeting of the laser beam.
While modern gimbals offer the required field of regard and have reasonably fast response, the search for the most
effective and reliable way to perform agile steering is an on-going effort. Non-mechanical devices such as acousto-optic
Bragg cells are virtually inertia-free and; therefore, have a great potential for this task [5], [6]. They outperform most of
the mechanical devices and offer a viable solution to the tracking problem.
Free-Space Laser Communication Technologies XVIII, edited by G. Stephen Mecherle
Proc. of SPIE Vol. 6105, 61050C, (2006) · 0277-786X/06/$15 · doi: 10.1117/12.640112
Proc. of SPIE Vol. 6105 61050C-1
Vertical cell Quadrant
Detector
Lu Soirce OPt. Bnn c.n /
B—— p©iiti©ii
Irdbiti©ii
&veto nt
RB signal
Drive Electronics
(with intece crd direct-digital
synthesis card and RFsection)
In this paper we investigate performance of a laser communication system with fast acousto-optic tracking and use the
results of an experimental study to demonstrate the effects of agile optical tracking on the bit-error-rate (BER) in the
presence of the pointing disturbance.
2. TRACKING SYSTEM AND ITS MAIN COMPONENTS
A generic configuration of an acousto-optic tracking system is shown in Fig. 1. It includes two Bragg cells required to
perform 2-dimensional beam steering and a quadrant detector, which provides beam position feedback to the controller
that regulates frequencies of the RF signals.
Figure 1. Acousto-optic tracking system
An acousto-optic cell, when used as a beam deflector, utilizes the effect of Bragg diffraction of the laser beam incident
upon a volume grating created by an ultrasonic signal [7], [8]. Experimental studies show that an acousto-optic device
has dynamics of a first-order system [9], [10] which could be approximated by the following transfer function.
G(s) =
θ
D/(f-fc) = [ )/(nv
λ
] * [ wb / (s+ wb) ] (1)
where
θ
D – beam deflection angle,
f – control frequency of the acoustic signal,
fc – central frequency of the acoustic signal,
λ
- optical wavelength,
n – refractive index,
v - acoustic velocity in the interaction medium,
wb – open-loop bandwidth of the Bragg cell.
The Bragg cells used in this study are made from tellurium dioxide (TeO2 or paratellurite) and are driven by ultrasonic
signals with a center frequency at 24MHz and the acoustic bandwidth of 12 MHz. The electronic subsystem uses direct-
digital-synthesis (DDS) and consists of a control interface card, two DDS cards, and an RF section, all self-contained in
one box. The electronic subsystem is interfaced to a PC, which generates control and frequency information via a digital
input-output card. FIFOs in the Control Interface Card buffer and then latch two channels of frequency information to
the DDS cards upon command from the PC.
Proc. of SPIE Vol. 6105 61050C-2
The position-sensing detector is an avalanche photodiode with a quadrant structure connected to four trans-impedance
amplifiers (TIAs), whose outputs, denoted as A, B, C, and D, are used to generate the azimuth and elevation feedback
signals. Just like in any photodiode circuit these signals will be contaminated with noises, generated primarily through
the shot noise mechanism (signal shot noise, background noise, and dark current noise) and thermal (Johnson) noise
mechanism. Let the standard deviations of the quadrant noises be na, nb, nc, and nd, then for a reasonably large signal-to-
noise ratio (SNR) the signals representing the azimuth and elevation displacements could be expressed as follows [11]
Σ
−−+
+=
Σ
−−+
+=
dcba
cbda
nnnn
yfEl
nnnn
xfAz
)(
)( (2)
where
DCBA +++=
Σ
(3)
and
Σ
−−+
=
Σ
−−+
=
DCBA
yf
CBDA
xf
)(
)( (4)
The presence of noise components shown in (2) will lead to additional angular errors, which result from the position
detection process. The extent of this phenomenon is expressed in terms of the noise equivalent angle (NEA), also known
as the tracking jitter [11], which represents additional variance of the pointing errors, generated by the tracking system in
response to the measurement noise, and could be found as follows
S
N
R
SF
TJ 2
21
=
σ
(5)
where SF – angular slope factor of the position sensing circuitry, and the SNR in the tracking system is defined as
2222
2
)(
dcba nnnn DCBA
SNR +++
+++
= (6)
It should be noted that (6) could also represent the SNR of the communication system if the quadrant APD is also used to
detect the information signal. The easiest way to implement this is by adding up the four signals generated by the
quadrants. Then using (6) for the analysis of the communication performance requires calculation of the noises using the
bandwidth of the communication system rather than the tracking bandwidth.
3. CONTROL SYNTHESIS
The dynamics of a Bragg cell is characterized by a first-order transfer function of Eq. (1), which is also supported by the
results of our step response experiments [9], [10]. Considering that the access time of these devices could easily be on
the order of tens of microseconds or less, their steering bandwidth is typically very large (usually on the order of tens of
kHz). Therefore, a simple gain controller in the feedback appears to be sufficient to reject most of the distortions, and an
equation for the control effort applied to a Bragg cell could be written as follows
f=fc+H*vaz,el , (7)
where vaz,el – azimuth or elevation feedback signal from the quadrant detector.
This approach; however, does not work in practice. Our quadrant photodiode will be a source of several types of noise,
including signal shot noise, background noise, and dark current noise; while thermal noise will be generated in the
electronic circuitry. System performance will be affected by all noise frequencies within the passband of the tracking
Proc. of SPIE Vol. 6105 61050C-3
Center
Frequenc
Signal
Compensating
system, which will pose a significant problem. Indeed, a device as agile as a Bragg cell would respond to almost any
signal coming from a quadrant detector, regardless of whether the signal represents an actual displacement of the laser
beam or just the additive noise. Therefore, a constant gain controller in the feedback needs to be complemented by
intelligent filtering of the position measurement signal. A block diagram of the proposed control system, per channel,
either azimuth or elevation, is presented in Fig. 2.
Figure 2. Control system configuration
A disturbance signal with a specific spectrum, e.g. representing aircraft jitter, continuously affects the pointing direction
of our transmitter. If this disturbance is not completely compensated by a fast acousto-optic steering device (FAOS), the
resultant pointing angle error causes response in the quadrant detector. Since a signal from the detector is contaminated
with noise, it is first filtered, and then used by a constant-gain controller to adjust the frequency of FAOS around
fc=24MHz. The purpose of a Kalman filter is to estimate the state of a system from measurements, which contain random
errors due to the noises in the tracking system.
Since the controlled plant (acousto-optic device) in this case is a first-order system, then implementation of a first-order
Kalman filter would probably be sufficient for measurement noise rejection. Generally, a first-order system could be
expressed in the discrete-time domain as follows:
1−
⋅+= nnn yaxy , (8)
where x – filter input;
y – filter output;
a – parameter of the model.
Then an equation for a first-order Kalman filter is
()
11
1|
1| −−
−
−⋅+⋅−⋅
+
=nnn
nn
nn
nyayax
Ms
M
y, (9)
where s is the noise variance and
()
1|
1|
1|
1
2
11|
1
−
−
−
−−−
+
−
⋅=
⋅−+⋅=
nn
nn
nnn
nnnnn
Ms
M
MM
yaxaMM
(10)
Proc. of SPIE Vol. 6105 61050C-4
C
C C
The above control and adaptive filtering algorithms are iterative and could be programmed in software. The frequency at
which the control outputs to the AOS system are updated could be very high (tens of kHz or more) and is simply a
function of the hardware characteristics and the latency in software-hardware interaction.
4. LABORATORY SYSTEM CONFIGURATION
An experimental laser communication system with fast tracking was assembled on an optical table as presented by the
schematic in Fig. 3. The parts on the right-hand-side are installed in the transmitter, while L, M3, M4, and QD could
represent receiver components.
Figure 3. Acousto-optic tracking testbed
A pigtailed laser diode with λ=633nm is used to send a beam through an optical path, which includes two orthogonal-
mounted acousto-optic deflectors for horizontal and vertical steering. The beam is expanded to approximately 12mm in
diameter to fill the apertures of the AODs. Since the Bragg cells are designed to work with circularly polarized incident
beams, two sets of half-wave plates and quarter-wave plates are used for this purpose. Lens L, representing the receiver
aperture is used to focus the beam on the quadrant detector. A number of folding mirrors are also used in the setup.
In our initial approach the laser diode was modulated directly with a square wave signal to generate a sequence of bits
representing communication data. An evaluation electronic board was used to switch the source on and off at a rate of
15kHz, thus sending 30kbps. However, the hardware constraints did not allow sufficiently fast switching, thus limiting
the modulation depth, communication bandwidth and introducing additional noise into the signal. All these factors made
it very difficult to build an effective tracking system and demonstrate its advantages in the presence of pointing
disturbances. Hence, external modulation was considered as an alternative. The acousto-optic cells used in our testbed
have high dynamic characteristics and could simply “steer the beam away from the receiver” every time ‘0s’ need to be
transmitted and keep it on the receiver aperture while sending ‘1s’. This feature was implemented by forming an
additional RF signal varying between 0 and 3MHz at a rate of 20kHz, and adding it to the input of one of the Bragg cells,
to modulate communication signal at 40kbps.
Optical tracking experiments also require some disturbance to be applied to alter the pointing direction from the
transmitter to the receiver. Acousto-optic cells could carry this function in addition to tracking, as was demonstrated in
[12]. These devices offer linear characteristics such that proportional increments/decrements of the acoustic frequency
result in linear increase/decrease of the pointing angles, which could represent the effects of the platform vibration or
optical turbulence. Hence, the function of the Bragg cells in our optical setup is tri-fold: external modulation of the laser
signal, introduction of the pointing angle disturbance and agile tracking. Details of this approach as well as the
electronics setup are presented in Fig. 4.
S
Source
M1-M4 Folding Mirrors
HWP Half-Wave Plate
QWP Quarter-Wave Plate
EX Beam Expander (8X)
AOD Acousto-Optic Deflector
L Biconvex Lens
QD Quadrant Detector
Proc. of SPIE Vol. 6105 61050C-5
Bragg Cell Electronics
(with intectacecardDDS
card and SF section) SF Control
Br AmplifieCell
(tracking device, mockdator,
and disturbance source)
Dedicated Processor
(with LabVIEW
Real-Time module)
Ethnet
PC-Based
center
position
info
Analysis System
I
to drive
electronics
Position lnfcmiation TI Amplifiers
QuadDtctor —
Figure 4. Electronic system setup
As the above schematic demonstrates, the compensating signal from the feedback gain controller and the noise are
combined to give a cumulative effect of the pointing errors and the tracking action. In addition, a signal from the on-off-
keying (OOK) modulator is added to facilitate binary data transmission. The noise is initially generated with an infinite
spectrum, but then it is changed by a “spectrum shaper.” The latter represents a single filter or a combination of filters
capable of producing a signal whose spectrum could be similar to a practical communication scenario, e.g. vibrations of
an aircraft, spacecraft, or the effects of turbulence.
The control system is built using National Instruments data acquisition (DAQ) boards and LabVIEW Real-Time
software [13]. A Pentium-III class desktop computer was converted into a real-time dedicated hardware target that runs a
single-kernel real-time operating system. Experimental algorithms were first developed on a Windows host computer
and then downloaded onto the real-time hardware target via an Ethernet link. Upon completion of each experimental run
another computer equipped with a faster acquisition board was used for temporal and spectral data analysis.
5. EXPERIMENTAL RESULTS
The designed tracking system has been tested under pointing disturbance conditions, which could represent mechanical
vibrations of the communication platform or the effects of optical turbulence, resulting in the change of the beam arrival
angle. The baseline noise characteristics were obtained to identify the fundamental limits of the hardware used in our
system. First, spectral response of the receiver was obtained in the absence of the incident laser beam, as shown in Fig. 6
(a). Then the beam was centered on the quadrant detector by moving the latter manually with precision mount knobs.
This resulted in zero mean signals representing the azimuth and elevation displacements, but due to the noises generated
in the system, mostly through the shot noise mechanism and the thermal noise mechanism, we always had nonzero
signals from the position sensing circuitry. A spectral characteristic of the radial displacement signal, representing our
noise floor, is featured in Fig. 5.
Proc. of SPIE Vol. 6105 61050C-6
-66
-70 --
-76 --
101 io io4
Frequency [Hz]
102 I0 ID
Frequency [Hz]
(a) (b)
Figure 5. Spectral characteristic of the electronic noise
(a) no signal; (b) with signal
An important aspect of a tracking experiment is the choice of a disturbance signal, whose spectral characteristics are
supposed to represent a specific communication environment. These could be, for example, recorded vibration spectra
from satellites, aircraft, ground vehicles, etc. All these characteristics have very particular shapes, possibly with resonant
peaks, representing operation of specific subsystems onboard the communication platform as well as its motion patterns.
In principle any of these spectral characteristics representing motion, vibrations or turbulence effects could be
implementing in our experimental setup by tuning the “spectrum shaper” module shown in Fig. 4.
However, for the purpose of this demonstration we chose a more generalized spectrum. A disturbance signal was formed
by filtering random noise with a third-order low-pass filter with a bandwidth of 2kHz. This results in almost flat
spectrum extending to 2kHz, which exceeds the effects of most of the realistic environments where precise pointing of a
laser beam is adversely affected by vibrations and atmospheric effects. This signal was generated twice and applied to
both horizontal and vertical channels of our system without activating tracking, while azimuth and elevation responses
were recorded. The two outputs were combined into a radial signal, whose power spectral density is shown in a solid line
(‘uncompensated’) in Fig. 6. It should be noted that this spectral response includes both, the actual beam motion across
the quadrant detector, and the electronic noise, with a spectrum shown in Fig. 5(b). Sampling of the two outputs was
performed at a rate of 20,000 samples/second; therefore, the highest frequency that could be detected by a frequency-
domain analysis procedure is 10kHz. It can also be seen from Fig. 6 that at frequencies exceeding 8kHz only the effects
of electronic noise are observed.
The same disturbance signal was used to test the performance of the tracking system. The control loop, implementing the
procedure given by Eq. (7)-(10), was designed to operate at a rate of 20kHz, same as the sampling rate. The spectral
response of the tracking system is shown in a dotted line in Fig. 6. As could be seen from the results, 7 to 9 dB of noise
rejection is assured at frequencies up to a few hundred Hz. The bandwidth of the tracking system, defined at the
frequency where 3dB reduction of the disturbance is assured, is approximately 2kHz. This is sufficient for rejecting most
of the vibration effects and could, to a significant degree, compensate beam wander or wavefront tilt due to the
atmospheric distortions.
Proc. of SPIE Vol. 6105 61050C-7
11111111]
HHW :1::::
-66 --H--
Uncompensated
Compensated I
102 I
-IU
Frequency [Hz]
4xlt H1tttFht1H
3.6
2.6
04 08 08 I12
4000
3000
3000
2600
2000
1600
1000
600 1LI—
Ft ii
8000
0000
4000
3000
2000
1000
02 04 00 08 I12
Figure 6. Spectral response of the acousto-optic tracking system
It could also be observed that fast steering action of the acouso-optic cells introduces additional noise at higher
frequencies. However, this happens in the stopband of the tracking system, where the magnitude of the spectrum is
already low. Additional experimental results are featured in Fig. 7.
(a) (b) (c)
Figure 7. Histograms of the radial displacement signal
(a) no pointing disturbance; (b) disturbance uncompensated; (c) disturbance compensated
Fig. 7(a) presents a histogram of the radial displacement signal calculated from the readings of the quadrant detector
while only electronic noise is present. Fig. 7(b) and (c) show the effects of the pointing disturbance without
compensation and with tracking, respectively.
Assuming that the total signal from all quadrants is used to detect the information sent via the optical link, as was
suggested earlier, we could evaluate communication performance of our system by collecting sufficient bits data. Since
the laser source is modulated to send ones and zeros, the corresponding received signals need to be arranged into two
sets for statistical analysis. The BER of a direct detection optical link is [11]
(
)
25.0 QerfcBER =, (11)
01
01
σσ
+
−
=II
Q, (12)
where erfc – complementary error function;
I1, I0 – signal currents (voltages) for logic ‘1’ and logic ‘0’, respectively;
σ
1,
σ
0 – noise currents (voltages) for logic ‘1’ and logic ‘0’, respectively.
Proc. of SPIE Vol. 6105 61050C-8
ID
B B
-8 -8 -4
Communication SinaI LV]
Bi18BillN818
N811
2:
3
E——Nfint
NfItU:
2.6
1.6
6.6
26 -26 AL-16 -ID
Communication SinaI LV]
14 ____________
BilUBill
12 NIlNfilB
ID
021201 B-I 0-B 0B
Clmmulicalill BilaI LV]
Since (11) implies Gaussian noise statistics, curve fitting of the collected data distribution was performed to find signals
and noises to be used in (12). The results are presented in the figure below.
(a) (b) (c)
Figure 8. Histograms and normal approximations of the communication signals
(a) no pointing disturbance; (b) disturbance uncompensated; (c) disturbance compensated
The approximation curves in Fig. 8 were obtained using standard functions in the MATLAB statistical toolbox. The
resulting mean values and variances corresponding to ones and zeros detected by our system are presented below.
Table. Communication performance analysis ‘0’ ‘1’
µ σ2 µ σ2 BER @
40kbps Max. bit
rate, kbps
No pointing disturbance 0.1712 0.0017 10.947 0.0697 3.57*10-273 1385
Disturbance uncompensated 0.1832 0.1015 11.625 1.1982 2.84*10-16 72.9
Disturbance compensated 0.1594 0.0028 11.373 0.2957 4.94*10-79 392
The last two columns of the above table show the results of the bit error rate analysis. Since our system was operated at
40 kbps, the corresponding BER values were found to be very small due to short link range and overall favorable
conditions. Therefore, we extrapolated the performance characteristics assuming that OOK modulation is implemented at
a much faster rate. In this case signal levels in (12) would remain the same, while variances of the noises would grow
proportionally to the increase of the communication bandwidth. Assuming that BER≤10-9 is required, we found the
maximum bit rates of our system, as presented in the last column of the table. Furthermore, we calculated
communication performance of the system subjected to the pointing disturbance, but not equipped with tracking when
data is sent at 392kbps. In this case the BER increases from 10-9 to 0.0049.
CONCLUSIONS
This paper presents results of an experimental study of a laser communication system equipped with non-mechanical
pointing and tracking. The latter was implemented using acousto-optic Bragg cells to assure agile beam steering.
Experimental results were obtained using a laboratory prototype, and the tracking bandwidth of 2kHz was demonstrated
with a rate of the control loop equal to 20kHz. The system is capable of compensating the effects of most practical
vibration environments and could also provide partial mitigation of the optical turbulence. It was demonstrated that
implementation of agile tracking results in the increase of the maximum bit rate by a factor of 5.37 or improves the BER
by several orders of magnitude. It is most likely that increasing the rate of the control loop would bring additional
benefit, thus extending the tracking bandwidth, decreasing the additional noise introduced by the Bragg cells at higher
frequencies, and reducing the noise in the received communication signal.
Proc. of SPIE Vol. 6105 61050C-9
REFERENCES
1. “Global Satellite Communications Technology and Systems,” WTEC Report, 1998.
2. D.K. Paul, F. Faris, R. Garlow, T. Inukai, B. Pontano, R. Razdan, A. Ganz, and L. Caudill, “Optical Intersatellite
Links: Application to Commercial Satellite Communications,” Proc. 14th AIAA Int. Communication Satellite
Systems, Washington, DC, Mar. 1992.
3. B.I Edelson and G. Hyde, “Laser Satellite Communications, Program Technology and Applications,” IEEE-USA
Aerospace Policy Committee Report, Apr. 1996.
4. D.K. Paul, “Optical Crosslinks for Advanced Satcom Networks,” Proc. Asia Pacific Microwave Conf., New Delhi,
India, Dec 1996.
5. P.J. Oleski, K.W. Dorrian, and T.E. Busch, “Non-mechanical Laser Beam Steering/Spoiling Methods for Inter
Satellite Cross Links,” Proc. SPIE, Vol. 2123, 1994.
6. H. Presley, “High efficiency large time-bandwidth acousto-optic laser deflector for free space optical
communications,” Harris Corp. Technical Report, 1995.
7. A. Yariv, Optical Electronics, CBS College Publishing 1985.
8. J. Xu, R. Stroud, “Acousto-Optic Devices. Principles, Design and Applications,” John Wiley & Sons,1992.
9. V. Skormin, V. Nikulin, and M. Bouzoubaa, “Analysis and Characterization of an Acousto-Optic Beam Steering
System,” Technical Report for NASA Jet Propulsion Laboratory, Pasadena, Ca, Sept. 2000.
10. V. Nikulin, M. Bouzoubaa, V. Skormin, and T. Busch, “Modeling of an Acousto-Optic Laser Beam Steering System
Intended for Satellite Communication,” Optical Engineering Journal, Vol. 40, No. 10, October 2001, pp.2208-2214.
11. S.G. Lambert, W.L. Casey, “Laser Communications in Space,” Artech House, 1995.
12. V.V. Nikulin, R. Khandekar, J. Sofka, and G. Tartakovsky, “Acousto-optic pointing and tracking systems for free-
space laser communications,” Proc. SPIE, Free-Space Laser Communications V, San Diego, CA, July-Aug. 2005.
13. www.ni.com
Proc. of SPIE Vol. 6105 61050C-10
