Vous êtes sur la page 1sur 12



Microwave Photonics Programs at DARPA

Richard W. Ridgway, Senior Member, IEEE, Carl L. Dohrman, Senior Member, IEEE,
and Joshua A. Conway, Senior Member, IEEE
(Invited Paper)

AbstractOver the past ten years, DARPA has made significant

investments toward advancing the field of microwave photonics.
This paper reviews DARPA-funded progress in this subject over
the past decade. DARPA-funded research has advanced the stateof-the-art for microwave-photonic components, including low noise
laser diodes, electrooptic modulators and high power photodiodes,
as well as microwave photonic link configurations, including photonic downconversion, reconfigurable optical filters and optical
phase-locked loops. These investments have yielded dramatic improvements in spurious-free dynamic range (SFDR). Measured
performance includes SFDRs exceeding 115 dB Hz2/3 at 16 GHz
using broadband externally modulated links; exceeding 120 dB
Hz2/3 at 10 GHz using sub-octave electrooptic modulators; near
135 dB Hz2/3 at 100 MHz using optical phased-locked loops
as linear phase demodulators; and exceeding 125 dB Hz2/3 at
5 GHz using optical filtering, downconversion and predistortion
Index TermsElectrooptic modulator, high power photodiode,
microwave photonics, noise figure, optical fiber links, optical
phased-locked loops, photonic downconversion, spurious free dynamic range (SFDR).


ULTIFUNCTIONAL receivers, capable of performing

communications, radar, and electronic warfare functions, are of considerable interest to the military. The next generation of multifunctional receiver technology will require substantial gains in several key performance parameters, including
increased frequency of operation (f ! 18 GHz), increased instantaneous bandwidth (B ! 1 GHz), increased spurious-free
dynamic range (SFDR > 120 dB Hz2/3 ) and enhanced receiver
sensitivity (S " 90 dBm). Furthermore, there is a continued
desire to reduce the system size, weight, power, and cost. The
application of microwave photonic components and links in
such receivers could yield significant improvements to these
performance parameters. Microwave photonic components operate at very high frequencies with very wide bandwidths [1][4]
and can efficiently transfer signals from the RF to the optical
Manuscript received January 15, 2014; revised March 27, 2014; accepted May
11, 2014. Date of publication June 17, 2014; date of current version September
1, 2014. The views opinions, and/or findings contained in this article are those
of the authors and should not be interpreted as representing the official views
or policies, either expressed or implied, of the Defense Advanced Research
Projects Agency or the Department of Defense.
R. W. Ridgway and J. A. Conway are with the U.S. Defense Advanced
Research Projects Agency (DARPA), Arlington, VA 22203 USA (e-mail:
richard.ridgway@darpa.mil; joshua.conway@darpa.mil).
C. L. Dohrman is with the Booz Allen Hamilton, Inc., Arlington, VA 22203
USA (e-mail: dohrman_carl@bah.com).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/JLT.2014.2326395

domains and back. Furthermore, microwave photonic technologies are compatible with frequency channelization to simultaneously allow wide instantaneous bandwidths and good receiver
sensitivity. Finally, microwave photonic links use optical fiber
to transport the signals around the platform, which reduces the
RF loss and significantly reduces the cable size and weight. As
an example, the weight of a representative low-loss militarygrade RF cable (113 kg/km [5]) is more than three times higher
than the weight of a comparable military-grade fiber-optic cable (31 kg/km [6]), while the loss of the RF cable (0.72 dB/m
at 18 GHz) is orders of magnitude more than fiber-optic cable (0.2 dB/km). While there are numerous components and
architectures that have been used to create microwave photonic
links, the three dominant components that enable the effective
use include:
1) Low-noise, high-power laser diodes.
2) Low-loss electrooptic modulators with low drive voltage,
3) High-power, highly-linear photodiodes.
Over the last ten years, the Defense Advanced Research
Projects Agency (DARPA) has invested in numerous programs
aimed specifically at the improvement and maturation of these
three components. The Ultra-Wideband Multifunction Photonic
Transmit/Receive Module (ULTRA-T/R) and Photonic Simultaneous Transmit and Receive (P-STAR) programs aimed to
improve the electrooptic modulator as a means of achieving
good RF isolation between transmit and receive signals sharing
a common aperture at X-band. The Transmit and Receive Optimized Photonics (TROPHY) program focused on improving
lithium niobate modulators and high power photodiodes to increase the RF transmit power of microwave photonic links. The
Linear Photonic RF Front-End Technology (PHOR-FRONT)
program investigated the use of optical phase-locked loops as a
means of improving the linearity of microwave photonic links
and explored photonic downconversion. The Photonic Analog
Signal Processing Engines with Reconfigurability (PhASER)
program considered RF signal processing in the optical domain,
using reconfigurable optical filters and delay lines to improve the
SFDR of an RF photonic link. The Network Enabled by WDMHighly Integrated Photonics (NEW-HIP) program investigated
low-noise laser diodes and high-power photodetectors to create
a wavelength-division-multiplexed network capable of transmitting analog and digital signals on a common single-mode
fiber-optic network. While these programs measured various
performance metrics, a common metric was SFDR. This paper will summarize some of the results from these and other
DARPA-funded programs and describe how these performance

0733-8724 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.



Fig. 2.

Block diagram of an external modulation microwave photonics link.

Fig. 1. SFDR and receiver sensitivity as a function of bandwidth for a typical

RF receiver, assuming an SNR requirement of 5 dB, noise figure of 5 dB, and
SFDR of 110 dB in a 1 Hz band.

metrics could translate to system attributes of interest to military

The three performance metrics of interest for a microwave
receiver include the instantaneous bandwidth, the receiver
sensitivity, and the dynamic range. The receiver sensitivity (S),
which is sometimes referred to as the minimum detectable signal, is determined by 1) the thermal noise power associated with
the bandwidth of the receiver, 2) the noise figure of the receiver,
and 3) the signal-to-noise ratio needed to detect or demodulate
the received signal:

Fig. 3. Plot of measured laser relative intensity noise (RIN) versus optical
output power. Filled markers denote shot-noise-limited measurements and the
open markers denote estimates of the laser RIN, with the shot-noise removed.
The red X depicts the desired performance of a low-RIN laser.

S = 10 log (kT B) + N F + SNRreq



where k is Boltzmanns constant, T is the apparent receiver temperature in Kelvins, and B is the bandwidth in Hertz. As the
bandwidth increases, the thermal noise will increase, causing
degradation in the receiver sensitivity. The red line in Fig. 1
shows an example of this relationship for a receiver with 5 dB
noise figure and 5 dB signal-to-noise ratio requirement. The dynamic range of a receiver is also impacted by the instantaneous
bandwidth. The SFDR of a receiver is often normalized to a
1 Hz bandwidth. For an instantaneous bandwidth, B, the SFDR
for a receiver can be calculated as:
SFDR (B) = SFDR (1Hz) 10log (B) .


The black dashed line in Fig. 1 shows an example relationship

between bandwidth and SFDR for a receiver with normalized
SFDR = 110 dB Hz2/3 .
Clearly, the bandwidth and the noise figure influence the
receivers sensitivity as well as the SFDR. The dashed line at
110 dB Hz2/3 is representative of the SFDR of a state-of-the-art
electronic receiver. The solid line denotes the sensitivity as
a function of bandwidth for a typical X-Band electronic
receiver chain with a noise figure of 5 dB. To achieve the

desired dynamic range of greater than 60 dB and sensitivity of

90 dBm requires a receiver bandwidth of about 10100 MHz.
A block diagram of an external modulation microwave photonic link is shown in Fig. 2. A low-noise, continuous-wave
laser diode provides light to an electrooptic modulator, such
as a lithium niobate modulator configured as a MachZehnder
interferometer (MZI). A microwave signal drives the electrooptic modulator, which modulates the light that is directed over
a single-mode fiber to a high-power photodiode. This section
summarizes the recent progress made on these three key microwave photonic components.
A. Low-Noise, High-Power Laser Diodes
Minimizing the relative intensity noise (RIN) of the laser
diode while attaining high output power is an essential step
in achieving the desired link noise figure and SFDR. DARPA
has made a number of investments in achieving low-RIN, highpower lasers at = 1550 nm for microwave photonics, with
several noteworthy successes, including demonstration of semiconductor lasers operating at or below the shot-noise floor [7] of
state-of-the-art measurement systems, as summarized in Fig. 3.
The DARPA RF Lightwave Integrated Circuits (RFLICS) program, researchers demonstrated a tunable sampled-grating distributed Bragg grating laser with shot-noise-limited RIN below



160 dB/Hz with 6 mW output power and 50 nm of wavelength

tunability [8]. On another DARPA-funded program, InP-based
slab-coupled optical waveguide external cavity lasers achieved
shot-noise-limited system RIN (# 163 dB/Hz) from 0.2 to
10 GHz with output powers of 370 mW at 1550 nm [9]. Additionally, under the DARPA NEW-HIP program, low-noise, highpower InP-based DFB lasers were demonstrated. Unpackaged
versions of these devices demonstrated output powers of over
200 mW with shot-noise-limited system RIN measurements of
158 dB/Hz, with estimated laser RIN of less than 165 dB/Hz
[10]. Packaged versions of these devices also demonstrated
shot-noise-limited RIN measurements while also showing laser
operation over a temperature range from 10 to 85 C. These
packaged lasers demonstrated 50 mW output power with shotnoise-limited RIN of 155 dB/Hz from 0.1 to 20 GHz, with
estimated laser RIN of 166 dB/Hz [11]. Fig. 3 shows the
summary of these RIN measurements.
B. The Electrooptic Modulator
The MachZehnder modulator (MZM) has become the
predominant electrooptic modulator for high-performance
microwave-photonic applications. One of the attractive features
of the MZM is its well-defined transfer function, which allows
for accurate link performance analysis as well as the potential
for linearization techniques using pre- or post-distortion. The
optical transfer function of a MachZehnder electrooptic modulator can be written as [12]:
(VB + vm sin(m t))
Tff P1
1 + cos
Pout (t) =
where Tff is the fiber-to-fiber transmission of the modulator
when biased for maximum transmission, P1 is the optical input
power in milliwatts, VB is the dc bias voltage, V is the voltage
needed to induce a phase shift, m is the input modulation
voltage, and m is the modulation frequency in radians per
second. The performance of a MZI is often characterized by its
slope efficiency [12], given by
sm zi (, f ) =

Tf f P1 Rs
2V (f )


where Rs is the source impedance and is angle relative to the

quadrature bias point (where = 0). Although there has been
a great deal of interest and exploration of the role of the bias
condition of the modulator [13], the most common bias point is
at quadrature, where the second-order distortion is minimized.
When the modulator is biased at quadrature, the third-order
distortion limits the SFDR, as will be discussed later.
Critical metrics for the electrooptic modulator include Tf f ,
V , maximum optical power handling, and the 3 dB cutoff
frequency. DARPA has explored these parameters in a number of programs, including ULTRA-T/R [13], P-STAR, and
TROPHY. Each of these programs aimed to increase the operational frequency of the modulators by reducing the RF-optical
velocity mismatch and by reducing the attenuation coefficient of
the traveling-wave electrodes [14]. Achieving RF-optical velocity matching requires increasing the velocity of the RF signal by

Fig. 4.

Measured V as a function of frequency [13].

reducing the effective dielectric constant of the RF signal. Methods of reducing the effective dielectric constant have included
reducing the thickness of the lithium niobate [15], increasing the
thickness of the microwave electrodes [14], and etching away
unneeded portions of the lithium niobate [16].
An innovative lithium niobate modulator developed under
ULTRA-T/R focused on achieving velocity matching between
the traveling wave RF signal and optical signal and maximizing the length over which this velocity matching occurred. The
novel design resulted in an interaction length between the RF
and optical signals of nearly 14 cm [17]. The dual-drive z-cut
design yielded one of the lowest measured V for a broadband
lithium niobate modulator: 1.4 V at 12 GHz. Fig. 4 summarizes
the measured V as a function of frequency [13]. The fiber-tofiber optical insertion loss (10 log(Tff )) was measured to be
8 dB [13]. While this low V modulator has a great deal of utility for microwave-photonic links, the focus of the ULTRA T/R
program was to allow simultaneous transmit and receive of microwave signals from a common antenna. Experiments carried
out under the ULTRA T/R program achieved 40 dB of isolation
between transmitted and received signals above 10 GHz [18].
The TROPHY program focused on improving the linearity
of microwave-photonic links as well as moving to higher operational frequencies. TROPHY included significant electrooptic
modulator development, and the focus was again on achieving velocity matching between the RF and optical signals. The
TROPHY program explored both thick electrodes and etching
of unneeded lithium niobate as a means of increasing the velocity of the microwave signal. As part of the TROPHY program, a
lithium niobate modulator was packaged with 1 mm coaxial connectors and demonstrated to have operational performance out
to 110 GHz. Fig. 5 shows the measured RF-optical frequency
response of the modulator. The fiber-to-fiber optical insertion
loss was measured to be 3 dB at a wavelength of 1550 nm [19].
On a DARPA-funded Small Business Innovation Research
program, velocity matching between the RF and optical signals
was achieved by transferring a very thin layer (110 m) of
lithium niobate onto a quartz substrate using crystal ion slicing [20]. The dc V of the device was shown to be 2.5 V at
1550 nm wavelength with a V L product of 4.75 Vcm [21].



Fig. 5. Measured RF-optical frequency response of a velocity-matched

lithium niobate modulator [19].

While these devices have not been packaged and fully characterized at high frequencies, preliminary results show preservation
of the electrooptic coefficient and low optical loss in the optical
waveguides formed in the thin films of lithium niobate. Furthermore, electromagnetic models show that the low dielectric
constant of the quartz substrate enables simultaneous velocity
matching and low RF loss out to millimeter-wave frequencies
In addition to lithium niobate modulators, DARPA has also
made significant investments in electrooptic modulators based
on alternative materials. While lithium niobate has a relatively
high electrooptic coefficient (r33 ) of 30.8 pm/V, novel materials engineering approaches have been investigated to produce
electrooptic polymers with potentially much higher r33 . The
DARPA Supermolecular Photonics Engineering (MORPH) program developed electrooptic polymers and investigated their
use in electrooptic modulators. Under this program, electrooptic polymers based on poly(methyl methacrylate)-anthrylmethyl
methacylate (PMMA-AMA) were synthesized and shown to
have r33 of over 300 pm/V. Additionally, MZMs were fabricated using these polymers, and dc V as low as 0.75 V was
demonstrated, while insertion losses of 17.1 dB were measured
[22]. Subsequent work with electrooptic polymer-based MZMs
has resulted in commercial products; as an example, an electrooptic polymer-based MZM with dc V of 2.5 V, insertion
loss of 6.5 dB, and speeds of up to 45 Gbit/s for differential
phase shift keying is commercially available [23].
Another materials system currently under development is the
GaAs/AlGaAs system. GaAs-based MZMs have been demonstrated in the past with V as low as 0.3 V in a pushpull configuration [24], but their overall performance is limited by the r33
of GaAs. Additionally, the small mode sizes of these devices result in increased coupling losses, which limits their utility in microwave photonic systems. Recently, DARPA has funded work
to increase the electrooptic efficiency of GaAs-based modulators through the use of InAlGaAs/InAlAs multi-quantum well
structures in the waveguide core [25]. These structures were
used to demonstrate MZMs with V of 2 V for single-arm-drive
and device length of 1.8 mm, resulting in a V L product of
0.36 Vcm. The device has potential for redesign in a pushpull
configuration to reduce V by a factor of two. The capacitance
of the devices was measured as 2 pF/cm, which is expected to
enable high-speed operation.

Fig. 6. Plot of V versus 3 dB bandwidth for a number of MZMs in the recent

literature. Data markers are color-coded to indicate DARPA funded work, and
shape-coded to indicate materials system. Data points which did not include
3 dB frequency measurements are shown at 1 GHz. Based upon data compiled
by [37].

A summary of the state-of-the-art in MZMs is shown in

Fig. 6, which plots V versus bandwidth. It should be noted
that insertion loss and power handling also play an important role in modulator performance but are not captured in this
C. Photodetectors
When operating in the linear region of the photodiode response, the output current of the photodiode is given by:
idet = rd Pout


where rd is the detector responsivity in A/W and Pout is the light

from the electrooptic modulator. Neglecting the frequency response of the photodiode and electrooptic modulator, the output
RF signal is given by:
PRFout = i2det RL


where RL is the load resistor. To address the high-power,

high-speed, high-linearity requirements of microwave photonic
receivers, the unitraveling carrier (UTC) photodiode has established itself as the leading photodiode architecture for highperformance receivers. Under the DARPA TROPHY program,
significant advances in high-power, high-linearity UTC photodiodes were demonstrated [38]. Specific developments include
the demonstration of UTC photodiodes with RF output power
up to 750 mW at 15 GHz with an output third-order intercept point (OIP3) of >55 dBm. The maximum power level
was achieved using InP/InGaAs-based UTC photodiode structures bonded to a high thermal conductivity substrate. Using a
novel charge-compensated modified unitraveling-carrier (CCMUTC) photodiode design, high RF output powers were attained in the V-band, including devices with 3 dB bandwidths
of 50 and 65 GHz demonstrated with saturation currents of 95



Fig. 8.
Fig. 7. Plot of saturation current versus frequency for high-power, high-speed
photodiodes. Based upon data compiled by J. Campbell and A. Beling [48].

and 55 mA, respectively [39]. A plot of the state of the art in

saturation current versus 3 dB bandwidth is shown in Fig. 7.
Another area of DARPA-funded photodiode research is
the topic of balanced photodetectors. Balanced photodetectors
have potential to reduce the impact of RIN on the NF in a microwave photonic link, by enabling a differential measurement
of the RF signal relative to the original optical carrier. Balanced
photodetectors also enable increased optical power handling.
While the prospect of noise cancellation is appealing, the implementation of microwave photonic links using balanced photodetectors is limited by the ability to achieve high-precision
optical power splitting and equal optical delay in each path.
Under the DARPA TROPHY program, an InP/InGaAs modified UTC photodiode was demonstrated with a 3 dB bandwidth
of 8 GHz, saturation current of 320 mA, maximum RF output
power of 1.5 W, and OIP3 of up to 47 dBm [40].
The inherent wide-bandwidth capabilities of microwave photonic components as well as the attractive size and weight benefits of optical fibers make microwave photonic links of significant interest to military systems for microwave antenna remoting
and signal distribution. Key figures of merit for a microwavephotonic link are:
1) Noise figure (NF)The measure of degradation of the
signal-to-noise ratio from the input microwave signal to
the output microwave signal.
2) SFDRThe ratio of the input signal and the induced
spurious signals.
3) Microwave frequency of operation.
4) Instantaneous bandwidth.
These key figures of merit are dependent, in part, on the microwave photonic components described in the previous section.
However, they can also be influenced by other system design
considerations, such as the use of photonic down-conversion,
optical phase-locked loops, and reconfigurable optical filtering,
as explored in the DARPA programs TROPHY, PHOR-FRONT,
and PhASER, respectively.

Calculated link noise figure as a function of laser RIN.

This section will review the key figures of merit of microwave

photonic links and review results from several DARPA-funded
A. Noise Figure
The noise figure is a critically important metric of the link.
While low noise figure can often be achieved through the use of
an electronic low-noise amplifier at the front of an RF receiver
chain, this diminishes the considerable bandwidth advantage
of using a microwave-photonic receiver and is impractical for
many antenna-remoting applications. The noise figure of a link
is given by [12]:
NF = 10 log 1 +
gi nin

where nin is the input noise (nin = kT B), nadd is the noise
added by the link, and gi is the intrinsic gain of the analog link.
For a MZI, gi is given by
Tf f P1 Rs
gi =
rd2 .

Recognizing that there are two resistors that contribute to the

system noise, the noise figure can be rewritten as [1]
RL % 2
+ iSN + it
N F = 10log 1 + constant +
gi kT RIN
where i2RIN , i2SN , and i2t are the noise-like currents associated
with laser RIN, shot noise, and thermal noise, respectively, and
constant term refers to added thermal noise from the modulation device circuit. This term has a marked frequency dependence, which can play a significant role in links using modulators
with traveling-wave electrodes (this term was first recognized
in [49] and is analyzed in greater detail in [1]).
Laser RIN tends to be the dominant noise current for microwave photonic links operating with high average photodetector current [12]. Fig. 8 shows the calculated noise figure
as a function of laser RIN for a link with high optical power
(P1 = 30 dBm), and low optical loss through the electrooptic modulator (Tf f = 0.5). The three curves denote modulators with V = 4 V, V = 2 V, and V = 1 V, respectively. It
should be noted that it is challenging to achieve a laser RIN of



Fig. 10.

Photonic mixer architecture [19].

C. The Role of Modulator Bias on SFDR

Fig. 9.

Calculated RF Transfer curve for a microwave photonic link.



Laser Power, P 1
Laser RIN
Modulator transmission, T f f
Detector responsivity, r d
Modulator load resistance, R m
Detector load resistance, R l
Noise figure

170 dB/Hz
5 dB
1.7 V
0.9 A/W
10 dB
10 MHz

less than 170 dB/Hz. For this reason, it is difficult to achieve

a noise figure of less than 10 dB without using a noise-figurereducing technique, such as balanced detection [13] or using a
low noise amplifier at the input of the MZI to set the link noise
For external modulation, amplitude-modulated links of the
type depicted in Fig. 2, the component that fundamentally limits the dynamic range performance of the photonic link is the
electrooptic modulator. When biased at quadrature, the cosine
response shown in the modulator transfer function (see Eq. (3))
will generate third-order nonlinearities as the input signals become large. Equations (3), (5), and (6) can be used to calculate
the microwave photonic links transfer function. Fig. 9 shows a
calculated transfer function for a link with component metrics
shown in Table I.
The SFDR of a microwave photonic link without predistortion
is limited by the third-order intermodulation distortion as is
given by [50]
[IIP3 NF (kT B)]
where IIP3 is the input third-order distortion intercept point. For
an electrooptic modulator in a MZI configuration, IIP3 increases
with increasing V [50]. For the example shown in Fig. 9, IIP3
was calculated to be 13.7 dBm and the SFDR was calculated to
be 71 dB for a 10 MHz instantaneous bandwidth.

The modulator bias conditions role in SFDR has been the

subject of several DARPA-funded programs. One DARPAfunded seedling program recognized that shifting the bias angle
( in Eq. (4)) away from quadrature has no effect on third-order
input power intercept but increases the even-order distortions
significantly. It was experimentally shown that the normalized
SFDR at an operational frequency of 2 GHz could be increased
by 9 dB to 122 dB Hz2/3 simply by adjusting the bias angle 168 from the maximum transmission point to a value of
= 78 and increasing the laser power to provide an equivalent
amount of photocurrent at the detector [50]. Additional experiments demonstrated operation of the link at 18 GHz with an
SFDR of 118 dB Hz2/3 . However, it is important to recognize
that these SFDR values assume that the bandwidth is limited to
less than an octave, since the second-order distortion terms have
not been minimized.
D. Photonic Downconversion
As described previously, photonic links can achieve large
SFDR and gain while reducing noise figure. For links spanning
tens of meters in length and with high microwave frequencies,
photonic links can meet or exceed the performance of conventional coaxial cable while reducing size and weight. In these
systems, the dynamic range of the receiver chain tends to be
limited by the microwave mixer.
To push system performance, the TROPHY program designed and demonstrated photonic mixers for wideband downconversion [19], [51]. Fig. 10 illustrates the architecture used
to implement photonic mixing. The system separately modulates the RF signal and local oscillator onto the split optical
source. The optical carrier and lower sidebands are removed in
the optical regime by using fiber Bragg grating filters. This step
is critical to the performance. By removing sidebands and harmonics before mixing on the photodiode, the spurs are greatly
reduced. The SFDR of the microwave photonic link with photonic downconversion was compared with the SFDR from a
conventional microwave mixer. In every spur measurement, the
photonic mixer outperformed the conventional microwave mixer
by at least 30 dB [51]. As empirically demonstrated, mixing in
the optical regime provided significantly reduced spurious signals over state-of-the-art electronic mixers while maintaining
operation over wider instantaneous bandwidths. The balanced
detection also rejected common mode noise such as RIN and
ASE beat noise. Using commercially available components, the


Fig. 11.


Microwave photonic link employing an optical phase-locked loop

TROPHY team was able to demonstrate photonic downconversion at X-Band with an SFDR near 120 dB Hz2/3 [52].
E. Phase Modulation and Optical Phased-Locked Loops
The DARPA PHOR-FRONT program explored the use of
phase modulators and optical phase-locked loops (OPLL) as a
means of increasing the SFDR of a microwave-photonic link. A
phase modulator formed in lithium niobate is intrinsically linear
[53], [54]. PHOR-FRONT explored optical phase-locked loops
as linear demodulators. In one implementation of the PHORFRONT concept, an attenuation-counterpropagation (ACP) inloop phase modulator was used to modulate the phase of the
LO laser and a directional coupler to serve as a phase detector,
as shown schematically in Fig. 11. Two photodetectors monitor
the phase difference and provide a feedback signal to the ACP
in-loop modulator to essentially match the phase modulation
placed on the optical signal by the incoming RF signal [53].
Initial measurements of the SFDR of the microwave-photonic
link were carried out with a commercial-grade lithium niobate phase modulator (V = 1.95 V) as the transmitter phase
modulator and an ACP-OPLL. At a modulation frequency of
100 MHz, the link was shown to have an SFDR of 135 dB
Hz2/3 [53]. This paper projects even higher SFDR if the optical power is increased, detector responsivity is increased and
ACP phase modulator is improved. Calculations suggest that an
SFDR of 147 dB Hz2/3 could be achieved with a 3 dB bandwidth of 500 MHz. These calculations further predict a link
noise figure of 2.2 dB [53].
A key challenge with optical phase locked loops is that the
latency around the loop limits the RF operational frequency.
Integrating the detectors and phase modulators into a common
substrate can reduce the latency and increase the operational frequency. In a second implementation of the PHOR-FRONT optical phased-lock loop, a coherent receiver comprised a balanced
photodetector, a two-by-two waveguide multimode interference
coupler, and the integrated tracking phase modulators [55]. The
photonic chip used an adjacent electronic chip comprising a
transconductance amplifier in its feedback path to drive the integrated phase modulator. The receiver was shown to have a
bandwidth of 1.45 GHz. The SFDR was measured to be 125 dB
Hz2/3 at a frequency of 300 MHz, 121 dB Hz2/3 at 500 MHz
and 113 dB Hz2/3 at 1 GHz.
PHOR-FRONT also explored a self-homodyne coherent receiver that relies on linear optical phase encoding of the RF
signal and uses a 90 optical hybrid for in-phase/quadrature
phase demodulation [56], as shown schematically in Fig. 12.

Fig. 12.

Self-homodyne coherent receiver [56].

The system used a fiber stretcher in one arm of the receiver to

provide interferometric bias control. The SFDR was measured
to be 126.8 dB Hz2/3 at test frequencies near 1 GHz. The noise
figure was measured at 18.6 dB. Calculations suggest that this
architecture could achieve an SFDR of 135.3 dB Hz2/3 with a
shot-noise-limited optical source [56].
This same in-phase/quadrature (I/Q) phase demodulation was
also used to demonstrate photonic downconversion. Here, an
additional phase modulator was added to one arm to provide the
LO. With the two test tones set at 10.0195 GHz and 10.0225 GHz
and the LO phase modulator driven at 5 GHz, the photonic
downconversion and I/Q demodulation demonstrated an SFDR
exceeding 105 dB Hz2/3 [36]. With the two test tones set at
3.0195 GHz and 3.0225 GHz and the LO phase modulator driven
at 1.5 GHz, the photonic downconversion and I/Q demodulation
demonstrated an SFDR exceeding 105 dB Hz2/3 [36].
F. Microwave-Photonic Filters
Microwave photonic filters are of substantial interest as an
essential component to a wide variety of practical microwave
photonic system configurations. Moreover, the high Q values attainable with photonics could enable filtering performance that
is not practical in the electronic domain [57]. Additionally, reconfigurability of microwave photonic filters is desired because
of its potential to enable more dynamic microwave photonic systems and to serve as a universal filtering component in photonic
systems. To meet this need, the DARPA PhASER program leveraged emerging photonic integrated circuit (PIC) technologies
to develop reconfigurable microwave photonic filters with high
dynamic range and wide bandwidths. The PhASER program advocated an integrated photonic unit cell concept, which would
enable scaling to very high-order filters. PhASER investigated
PIC-based unit cells both with and without integrated optical
gain. Optical gain enables implementation of more complex filters with a greater number of unit cells, but this can result in
added noise and distortion, and it requires the use of a fabrication process supporting optical gain (typically InP-based). As
part of the PhASER program, reconfigurable microwave filters
were demonstrated on a number of platforms, with and without
optical gain. In the InP-InGaAsP platform, a unit cell with reconfigurable poles and zeroes, based on an asymmetrical MZI
with feedback, was demonstrated [58]. This unit cell was cascaded to produce a fourth-order filter with passband tunability
of 1.95.4 GHz with stopband rejection greater than 32 dB. The
linearity of these filters was investigated, and optimization of


Fig. 13.


Filtered coherent microwave photonic link [63].

the demonstrated structure is expected to enable SFDR of up to

117 dB Hz2/3 for filters in the 12 GHz range [59], Also of
interest was the demonstration of a widely tunable filter unit cell
on a hybrid silicon platform with integrated InP-based materials
for optical gain [60]. These unit cells were shown to demonstrate poles and zeroes simultaneously with unit cell transfer
characteristics that closely matched theory.
PhASER also included development of microwave photonic
filters on silicon-based platforms without optical gain. This
study included the development of fully reconfigurable microwave photonic filters demonstrating both poles and zeroes
and comprising high-Q silicon microdisk resonators [61]. These
fourth-order filters were formed on a silicon-on-insulator platform with tunable 3 dB bandwidths from 900 MHz to 5 GHz
and out-of-band rejection of 38 dB.
In order to minimize the need for gain in optical filters, development of low-loss waveguides is of interest for improving
this technology. Optical loss in waveguides was the focus of
the DARPA Integrated Photonic Delay (iPhoD) program, which
made significant progress in reducing the loss of on-chip optical
waveguides, demonstrating losses of 0.08 dB/m through a 27 m
integrated waveguide, with losses of 0.037 dB/m achieved in
waveguide resonators [62].
These filters can be used to increase the SFDR of microwave
photonic coherent links as shown schematically in Fig. 13 [63].
The optical filtering allows only a single sideband of the modulated optical signal to reach the optical detector. The two electrooptic MZI modulators were biased for carrier suppression.
An optical filter with 350 MHz bandwidth centered at the first
LO sideband ensures that the only signals that reach the detector are the first sidebands and the LO. An SFDR of 116 dB
Hz2/3 was measured when the optical power to each electrooptic
MZM was set to 20 dBm [63].
The SFDR was further improved using predistortion compensation of the nonlinearities of the microwave photonic link by
inserting a compensating modulator in the LO path [63]. The
third-order distortion was suppressed by 20 dB, resulting in a
13 dB reduction in SFDR [64]. The intermodulation distortion
as a function of input RF power was shown to have a slope of
five rather than three, indicating that the SFDR was now limited
by fifth-order distortion [63]. Like many other linearization approaches, these SFDR improvements are only valid over limited
bandwidths, as discussed in [65].
The state-of-the-art in microwave photonic link SFDR as
a function of frequency has been summarized in previous reviews, in particular Cox et al. [65]. This paper showed that links

Fig. 14. SFDR in 1 Hz noise bandwidth as a function of measured performance frequency. IMDD is an abbreviation for intensity modulation, direct

employing directly modulated laser diodes had high dynamic

range (up to 125 dB Hz2/3 [66]) at frequencies less than 1 GHz
but suffered significantly as the frequencies increased. This paper further showed that links employing laser diodes and standard lithium niobate modulators could achieve SFDRs of 112 dB
Hz2/3 at frequencies between 2 to 18 GHz [67][69]. Links employing electrooptic modulators biased near the off-state were
shown to have an even higher SFDR of 115 dB Hz2/3 at frequencies up to 20 GHz [70], although operating the modulator
near the off-state emphasizes the second-order distortion, thus
limiting the links operation to bandwidths of less than one octave.
Fig. 14 shows the published SFDR in 1 Hz noise bandwidths
for the representative values from [65] along with the values
summarized in this paper. In this chart, the black markers denote SFDR values from [65], and the red markers denote the
SFDR values from recent DARPA-funded efforts reviewed in
this paper. The filled triangles denote external modulators biased
at quadrature and the open triangles denote external modulators
biased near the modulator off-state and operating over less than
one octave. In both cases, recent DARPA investments have resulted in a 3 to 5 dB improvement in SFDR.
The filled squares show the SFDR of links that use electrooptic phase modulators and optical phase-locked loops for demodulation. These links have shown exceptionally high SFDR
(>130 dB Hz2/3 ) at low frequencies (<1 GHz) but have limited
frequency response due to the closed-loop latency of the optical
phase-locked loops.
The red diamonds show the measured SFDR for experimental
results that use photonic downconversion and optical filtering.
One particularly noteworthy result is the measurement carried
out at 5 GHz with an SFDR of 129 dB Hz2/3 that involved
photonic downconversion, photonic filtering, and photonic predistortion compensation [64].




Fig. 15.



Frequency band
Instantaneous bandwidth
Dynamic range
Receiver sensitivity

> 50 GHz
> 5 GHz
> 60 dB
< 90 dBm

Microwave-photonic-based channelized receiver.


Over the last few years, significant progress has been made in
the field of microwave photonics. However, microwave photonic
systems have seen limited transition to existing military equipment. This could change as communications systems, radar systems, and electronic warfare systems increase in frequency to the
millimeter-wave portion of the spectrum (30 GHz300 GHz).
Optical fiber continues to be an attractive alternative to heavy
cables and waveguides, particularly at high frequencies.
The SFDRs reviewed in this paper compare favorably to
electronic microwave receivers, and military system designers
are eager to have broadband systems with normalized SFDRs
greater than 120 dB Hz2/3 . However, system designers are also
interested in instantaneous bandwidth and receiver sensitivity.
Table II summarizes the system performance metrics of an attractive millimeter-wave receiver. There are no known ways to
achieve these metrics simultaneously.
The tradeoff between bandwidth and receiver sensitivity often leads to the use of channelized receivers where multiple
channels, each operating over a narrow bandwidth, work in
parallel to increase the receivers instantaneous bandwidth. For
microwave receivers, this usually means replicating the bandpass filter, the LNA, the cables, the mixer, and the digitizer,
leading to a bulky and heavy receiver system. For microwave
photonic systems, channelized systems can share the broadband electrooptic modulator and optical fiber. Furthermore, the
channelization can occur in the optical domain near the receiver,
leading to a significant reduction in complexity and weight [63].
Fig. 15 shows a schematic of a microwave photonic circuit that
uses channelization to split a 5 GHz instantaneous bandwidth at
70 GHz into fifty 100 MHz bandwidth channels. Here, each link
achieves high dynamic range and improved sensitivity by limiting the bandwidth of each channel. Photonic downconversion at
the photodetector can then mix the millimeter-wave-modulated

optical signal down to an intermediate frequency or baseband,

where it is digitized with a low-sample-rate analog-to-digital
converter (ADC).
While the implementation of the planar lightwave circuit
shown is conceptually straightforward using optical interleavers,
arrayed waveguide gratings, directional couplers, and tunable optical filters, the actual implementation poses significant
1) Arrayed-waveguide-grating-type channelizer or interleaver with 100 MHz output ports are well beyond what
is commercially available.
2) Narrowband 100 MHz optical filters would have high optical loss and limited out-of-band rejection.
3) Minimizing the microwave phase distortion across the
channelized bands would be challenging.
Military systems designers are also concerned with system
size and cost. Lithium niobate modulators are the predominant
modulator of the microwave photonics industry. While many
other electrooptic modulators have been studied, none have met
the benchmark performance of lithium niobate when all of the
pertinent performance metrics, including drive voltage, linearity, and fiber-to-fiber optical loss, are taken into consideration.
However, lithium niobate modulators require interaction lengths
of >10 cm to achieve low drive voltages, making them bulky
and expensive. Lowering the size and cost of microwave photonic systems will require integration of photonic and electronic
Continued development of silicon photonics offers significant
potential for leveraging highly advanced CMOS fabrication infrastructure to achieve levels of photonics integration which
cannot be achieved by other means. The DARPA ElectronicPhotonic Integrated Circuit (EPIC) program combined electronic and photonic integration on a common silicon substrate
[72]. EPIC explored modulation, filtering, and downconversion
in a silicon photonic circuit [73]. While the SFDR of 94.3 dB
Hz2/3 measured at 11 GHz did not meet the performance of the
microwave photonic links reviewed in this paper, it is impressive
for a reconfigurable silicon photonic circuit.
Another DARPA effort, the Electronic-Photonic Heterogeneous Integration (E-PHI) program, is currently underway to
develop the design and manufacturing tools for electronicphotonic heterogeneously integrated systems. The goal of this
effort is to deliver circuits employing cutting edge CMOS circuits co-integrated with active and passive photonic devices.
If successful, E-PHI could have several critical impacts to microwave photonics, including reduced modulator V , very-lowlinewidth semiconductor-based photonic sources with worldclass noise levels for intensity-modulated and coherent links,
integrated microwave signal sources with phase noise levels
surpassing all other compact sources, and major advances in
photonics integration, leading to smaller size and lower system
costs as well as improved functionality for more sophisticated
microwave photonic architectures. For instance, the tight cointegration of electronics and photonics could also reduce the
latency in optical phase-locked loop systems, pushing the operational frequency into the gigahertz levels while maintaining
exceptionally high SFDR.



This paper has reviewed many of the DARPA-funded microwave photonic link developments carried out over the last
ten years. The key takeaways from these investments include
the following:
1) Reducing laser RIN and the V of electrooptic modulators as well as increasing the linearity and power handling
capability of optical detectors has improved link performance, including SFDR and NF.
2) Microwave photonic links with photonic downconversion
can yield SFDRs higher than similar microwave-only links
employing a microwave mixer for downconversion.
3) Optical phase modulation with optical phase-locked loops
can yield exceptionally high SFDRs.
4) Photonic-based predistortion compensation can effectively increase SFDR over limited bandwidths.
DARPAs Microsystems Technology Office has been a leader
in developing microwave photonic components, architectures
and links. Using these technical advances for microwave antenna remoting, photonic downconversion, and channelized receivers could significantly improve the performance metrics for
the militarys advanced receivers, including extending the frequency range, increasing the instantaneous bandwidth and dynamic range as well as improving receiver sensitivity. The next
step is to move microwave photonic technologies from these
experimental efforts and demonstrations to deployed military








The authors would like to acknowledge the past DARPA
program managers within the Microsystems Technology Office
who have developed and supported the innovative programs in
microwave photonics, including Dr. R. Leheny, Dr. S. Pappert,
Dr. M. Haney, Dr. R. Esman, and Dr. J. S. Rodgers. Without
their insights and dedication to microwave photonics, many of
these programs would not have been completed. The authors
also acknowledge Dr. J. S. Rodgers for discussions on electrooptic modulators and Prof. J. Campbell and Prof. A. Beling
of the University of Virginia for discussions on high power




[1] E. I. Ackerman, W. K. Burns, G. E. Betts, J. X. Chen, J. L. Prince,
M. D. Regan, H. V. Roussell, and C. H. Cox, III, RF-over-fiber links with
very low noise figure, J. Lightw. Technol., vol. 26, no. 15, pp. 24412448,
Aug. 2008.
[2] S. Pappert, RF photonics: Status, challenges and opportunities, presented at the IEEE Conf. Avionics Fiber-Optics and Photonics Technology,
San Diego, CA, USA, 2011, Paper TuA4.
[3] R. Esman, S. Pappert, B. Krantz, and G. Gopalakrishnan, Photonics for
microwave generation, transmission and processing, in Proc. Opt. Fiber
Commun., San Diego, CA, USA, Mar. 2009, pp. 2226.
[4] R. C. Williamson and R. D. Esman, RF Photonics, J. Lightw. Technol.,
vol. 26, no. 9, pp. 11451153, May 2008.
[5] Harbour Industries, Inc. (2013, Dec. 18). Specifications for Low
loss coaxial cableSolid center conductors, Part #LL335i. [Online]. Available: http://www.harbourind.com/images/stories/datasheets/
[6] Optical Cable Corporation. (2013, Dec. 18). Specifications for
Military tactical fiber optic cables for extreme environments,



Part #RK981104. [Online]. Available: http://www.occfiber.com/main/

L. A. Coldren and S. W. Corzine, Diode Lasers and Photonic Integrated
Circuits. New York, NY, USA: Wiley, 1995.
H. Shi, D. Cohen, J. Barton, M. Majewski, L. A. Coldren, M. C. Larson,
and G. A. Fish, Relative intensity noise measurements of a widely tunable
sampled-grating DBR laser, IEEE Photon. Technol. Lett., vol. 14, no. 6,
pp. 759761, Jun. 2002.
P. W. Juodawlkis, J. J. Plant, W. Loh, L. J. Missaggia, F. J. ODonnell,
D. C. Oakley, A. Napoleone, J. Klamkin, J. T. Gopinath, D. J. Ripin,
S. Gee, P. J. Delfyett, and J. P. Donnelly, High-power, low-noise 1.5m slab-coupled optical waveguide (SCOW) emitters: Physics, devices,
and applications, IEEE J. Sel. Topics Quantum Electron., vol. 17, no. 6,
pp. 16981714, Nov./Dec. 2011.
Y.-G. Zhao, X. Luo, D. Tran, Q. Hang, P. Weber, T. Hang, R. KnustGraichen, N. Nuttall, R. Cendejas, A. Nikolov, and R. Dutt, High-power
and low-noise DFB semiconductor lasers for RF photonic links, presented
at the IEEE Conf. Avionics Fiber-Optics and Photonics Technology, Cocoa
Beach, FL, USA, 2012.
A. Nikolov, D. Guenther, W. Liu, R. Cendejas, and R. Dutt, Advancements and challenges for photonic components and avionic interconnects,
presented at the IEEE Conf. Avionics Fiber-Optics and Photonics Technology, San Diego, CA, USA, 2013.
C. Cox, Analog Optical Links. Cambridge, U.K.: Cambridge Univ. Press,
E. Ackerman, G. Betts, W. Burns, J. Campbell, C. Cox, N. Duan, J. Prince,
M. Regan, and H. Roussell, Signal-to-noise performance of two analog
photonic links using different noise reduction techniques, in Proc. IEEE
MTT-S Int. Microw. Symp. Dig., Jun. 2007, pp. 5154.
G. Gopalakrishnan, W. Burns, R. McElhanon, C. Bulmer, and
A. Greenblatt, Performance and modeling of broadband LiNbO traveling wave optical intensity modulators, J. Lightw. Technol., vol. 12, no.
10, pp. 18071819, Oct. 1994.
R. W. Ridgway, V. E. Wood, M. R. Seiler, and G. T. Ruck, Integratedoptic modulator operating at millimeter-wave frequencies, in Proc. OSA
Top. Meet. Integr. Guided Wave Opt., 1989, pp. 102105.
W. K. Burns, M. M. Howerton, R. P. Moeller, R. Krahenbuhl, R. W.
McElhanon, and A. S. Greenblatt, Low drive voltage, broad-band LiNbO
modulators with and without etched ridges, J. Lightw. Technol., vol. 17,
no. 12, pp. 25512555, Dec. 1999.
H. Roussell, M. Regan, J. Prince, C. Cox, J. Chen, W. Burns, E. Ackerman,
and J. Campbell, Gain, noise figure, and bandwidth-limited dynamic
range of a low-biased external modulation link, in Proc. IEEE Int. Top.
Meet. Microw. Photon., Victoria, Canada, Oct. 2007, pp. 8487.
E. Ackerman and C. Cox, Optimization of photonic transmit/receive
module performance, presented at the IEEE Int. Topical Meeting Microwave Photonics, Valencia, Spain, Oct. 2009.
C. Middleton and R. DeSalvo, Photonics-based wideband frequency conversion, presented at the Government Microcircuit Applications Critical
Technology Conf., Las Vegas, NV, USA, 2013.
V. Stenger, J. Toney, A. Pollick, J. Busch, J. Scholl, P. Pontius, and
S. Sriram, Engineered thin film lithium niobate substrate for high
gain-bandwidth electro-optic modulators, presented at the Conf. Lasers
and Electro-Optics, San Jose, CA, USA, 2013.
S. Sriram, V. Stenger, J. Toney, N. Smith, A. Pollick, and J. Scholl,
Electro-optic devices in thin film lithium niobate, presented at the IEEE
Conf. Avionics Fiber-Optics and Photonics Technology, Cocoa Beach,
FL, USA, 2012, p. 5.
D. Huang, D. Jin, B. Chen, H. Chen, D. Tolstedt, S. Condon, A. Barklund,
G. Yu, E. Miller, R. Dinu, J. Luo, Z. Shi, T.-D. Kim, S. Huang, and
A. K.-Y. Jen, Highly efficient EO polymers for low V modulators,
Proc. SPIE, vol. 7049, p. 70490H, 2008.
(2013, Dec. 17). Data sheet for Gigoptix LX8401 40Gb/s DPSK MZ Modulator. [Online]. Available: http://www.gigoptix.com/product/73-lx8401
J.-H. Shin, Y.-C. Chang, and N. Dagli, 0.3 V drive voltage GaAs/AlGaAs
substrate removed Mach-Zehnder intensity modulators, Appl. Phys. Lett.,
vol.92, p. 201103, 2008.
S. Dogru, J.H. Shin, and N. Dagli, InGaAlAs/InAlAs multi quantum
well substrate removed electro-optic modulators, in Proc. IEEE Photon.
Conf., 2011, pp. 73940.
W. Green, M. Rooks, L. Sekaric, and Y. Vlasov, Ultra-compact, low
RF power, 10 Gb/s silicon Mach-Zehnder modulator, Opt. Exp., vol. 15,
pp. 1710617113, 2007.
A. Tauke-Pedretti, M. Dummer, M. Sysak, J. Barton, J. Klamkin, J. Raring,
and L. Coldren, Separate absorption and modulation Mach-Zehnder












wavelength converter, J. Lightw. Technol., vol. 26, no. 1, pp. 9198,

Jan. 2008.
H. Chen, J. Peters, and J. Bowers, Forty Gb/s hybrid silicon MachZehnder modulator with low chirp, Opt. Exp., vol. 19, pp. 14551460,
R. Griffin and A. Carter, Advances in InP optical modulators, presented at the Asia Optical Fiber Communication and Optoelectronic Expo.
Conf., Optical Society of America Technical Digest (CD), 2008, Shanghai,
China, Paper SuF6.
H. Nguyen, Y. Sakai, M. Shinkawa, N. Ishikura, and T. Baba, 10 Gb/s
operation of photonic crystal silicon optical modulators, Opt. Exp.,
vol. 19, pp. 1300013007, 2011.
(2013, Dec.17). Data sheet for Gigoptix LX8220 40 Gb/s DQPSK Modulator. [Online]. Available: http://www.gigoptix.com/product/86-lx8220
(2013, Dec. 17). Data sheet for Gigoptix LX8901 100 Gb/s DPSK MZ
Modulator. [Online]. Available: http://www.gigoptix.com/product/185lx8901
(2013, Dec. 17). Data sheet for Fujitsu 40 Gbps DP-QPSK lithium niobate modulator Part #FTM7967EQA. [Online]. Available: http://jp.fujitsu.
(2013, Dec. 17). Data sheet for EOSpace 40+ Gb/s Z-cut fixed
pre-chirp modulator. [Online]. Available: http://www.eospace.com/
(2013, Dec. 17). Data sheet for EOSpace ultra-low loss phase modulator.
[Online]. Available: http://www.eospace.com/phase_modulator.htm
T. R. Clark, Jr., S. R. OConnor, and M. L. Dennis, A phase-modulation
I/Q-demodulation microwave-to-digital photonic link, IEEE Trans. Microw. Theory Techn., vol. 58, no. 11, pp. 30393058, Nov. 2010.
J. S. Rodgers, Technologies for RF photonics in wideband multifunction systems, in Proc. IEEE Avionics, Fiber-Opt. Photon. Conf., 2013,
J. C. Campbell, A. Beling, M. Piels, F. Yang A. Cross, Z. Qiugui J. Peters
J. E. Bowers, and L. Zhi, High-power, high-linearity photodiodes for RF
photonics, in Proc. Conf. Optoelectron. Microelectron. Mater. Devices,
2012, pp. 215216.
Q. Zhou, A. S. Cross, A. Beling, Y. Fu, Z. Lu, and J. C. Campbell, Highpower V-band InGaAs/InP photodiodes, IEEE Photon. Technol. Lett.,
vol. 25, no. 10, pp. 907909, May 2013.
Z. Qiugui, A.S. Cross, F. Yang, A. Beling, B. M. Foley, P. E. Hopkins, and
J. C. Campbell, Balanced InP/InGaAs photodiodes With 1.5-W output
power, IEEE Photon. J., vol. 5, no. 3, p. 6800307, 2013.
N. Duan, X. Wang, L. Ning, H.-D. Liu, and J. C. Campbell, Thermal analysis of high-power InGaAsInP photodiodes, IEEE J. Quantum Electron.,
vol. 42, no. 12, pp. 12551258, Dec. 2006.
N. Duan, L. Ning, S. Demiguel, and J. C. Campbell, High-speed highcurrent InGaAs/InP photodiode with thick depletion region, in Proc.
IEEE 18th Annu. Meet. Lasers Electro-Opt. Soc., Oct. 2222, 2005,
pp. 458459.
N. Li, X. Li, S. Demiguel, X. Zheng, J. C. Campbell, D. A. Tulchinsky,
K. J. Williams, T. D. Isshiki, G. S. Kinsey, and R. Sudharsansan, Highsaturation-current charge-compensated InGaAs-InP uni-traveling-carrier
photodiode, IEEE Photon. Technol. Lett., vol. 16, no. 3, pp. 864866,
Mar. 2004.
K. Sakai, E. Ishimura, M. Nakaji, S. Itakura, Y. Hirano, and T. Aoyagi,
High-current back-illuminated partially depleted-absorber p-i-n photodiode with depleted nonabsorbing region, IEEE Trans. Microw. Theory
Techn., vol. 58, no. 11, pp. 31543160, Nov. 2010.
D. A. Tulchinsky, J. B. Boos, D. Park, P. G. Goetz, W. S. Rabinovich, and
K. J. Williams, High-current photodetectors as efficient, linear, and highpower RF output stages, J. Lightw. Technol., vol. 26, no. 4, pp. 408416,
Feb. 15, 2008.
M. Chtioui, A. Enard, D. Carpentier, S. Bernard, B. Rousseau, F. Lelarge,
F. Pommereau, M. Achouche, High-performance uni-traveling-carrier
photodiodes with a new collector design, IEEE Photon. Technol. Lett.,
vol. 20, no. 13, pp. 11631165, Jul. 1, 2008.
M. Chtioui, F. Lelarge, A. Enard, F. Pommereau, D. Carpentier, A.
Marceaux, F. Van-Dijk, and M. Achouche, High responsivity and high
power UTC and MUTC GaInAs-InP photodiodes, IEEE Photon. Technol.
Lett., vol. 24, no. 4, pp. 318320, Feb. 15, 2012.
J. C. Campbell and A. Beling, unpublished data compilation.
L. T. Nichols, K. J. Williams, R. D. Esman, Optimizing the ultrawideband photonic link, IEEE Trans. Microw. Theory Tech., vol. 45, no. 8,
pp. 13841389, Aug. 1997.

[50] A. Karim and J. Devenport, High dynamic range microwave photonic

links for RF signal transport and RF-IF conversion, J. Lightw. Technol.,
vol. 26, no. 15, pp. 27182724, Aug. 2008.
[51] C. Middleton, A. Mast, and R. DeSalvo, Mixing spur reduction through
photonic-assisted frequency conversion, in Proc. IEEE Conf. Avionics
Fiber-Opt. Photon. Technol., 2012, pp. 2627.
[52] C. Middleton and R. DeSalvo, Scalable photonic-assisted wideband frequency converter, presented at the OSA Signal Processing in Photonic
Communications, Toronto, Canada, 2011, Paper SPMC3.
[53] Y. Li and P. Herczfeld, Coherent PM optical link employing ACP-PPLL,
J. Lightw. Technol., vol. 27, no. 9, pp. 108694, May 2009.
[54] Y. Li, A. Bhardwaj, R. Wang, S. Jin, L. A. Coldren, J. E. Bowers, and P. Herczfeld, A monolithically integrated ACP-OPLL receiver
for RF/Photonic links, IEEE Photon. Technol. Lett., vol. 23, no. 20,
pp. 14751477, Oct. 15, 2011.
[55] A. Ramaswamy, L. A. Johansson, J. Klamkin, H.-F. Chou, C. Sheldon,
M. J. Rodwell, L. A. Coldren, and J. E. Bowers, Integrated coherent
receivers for high-linearity microwave photonic links, J. Lightw. Technol.,
vol. 26, no. 1, pp. 209216, Jan. 2008.
[56] S. R. OConnor, M. L. Dennis, and T. R. Clark, Optimal biasing of a
self-homodyne optically-coherent RF receiver, IEEE Photon. J., vol. 2,
no. 1, pp. 17, Feb. 2010.
[57] R. Won, Microwave photonics shines, Nature Photon., vol. 5, no. 12,
p. 736, 2011.
[58] E. J. Norberg, R. S. Guzzon, J. S. Parker, L. A. Johansson, and L. A.
Coldren, Programmable photonic microwave filters monolithically integrated in InPInGaAsP, J. Lightw. Technol., vol. 29, no. 11, pp. 1611
1619, Jun. 2011.
[59] R. S. Guzzon, E. J. Norberg, L. A. Coldren, Spurious-free dynamic
range in photonic integrated circuit filters with semiconductor optical
amplifiers, IEEE J. Quantum Electron., vol. 48, no. 2, pp. 269278,
Feb. 2012.
[60] H.-W. Chen, A. W. Fang, J. D. Peters, Z. Wang, J. Bovington, D. Liang,
and J. E. Bowers, Integrated microwave photonic filter on a hybrid silicon
platform, IEEE Trans. Microw. Theory Techn., vol. 58, no. 11, pp. 3213
19, Nov. 2010.
[61] P. Alipour, A. A. Eftekhar, A.H. Atabaki, Q. Li, S. Yegnanarayanan, C. K.
Madsen, and A. Adibi, Fully reconfigurable compact RF photonic filters
using high-Q silicon microdisk resonators, Opt. Exp., vol. 19, no. 17,
pp. 1589915907, Aug. 2011.
[62] H. Lee, T. Chen, J. Li, O. Painter, and K. J. Vahala, Ultra-low-loss
optical delay line on a silicon chip, Nature Commun., vol. 3, p. 867,
[63] A. Agarwal, T. Banwell, and T. K. Woodward, Optically filtered microwave photonic links for RF signal processing applications, J. Lightw.
Technol., vol. 29, no. 16, pp. 23942401, 2011.
[64] T. K. Woodward, T. C. Banwell, A. Agarwal, P. Toliver, and R. Menendez,
Signal processing in analog optical links, in Proc. IEEE Conf. Avionics
Fiber-Opt. Photon. Technol., 2009, pp. 1718.
[65] C. H. Cox III, E. I. Ackerman, G. E. Betts, and J. L. Prince, Limits
on the performance of RF-over-fiber links and their impact on device
design, IEEE Trans. Microw. Theory Tech., vol. 54, no. 2, pp. 906920,
Feb. 2006.
[66] S. Pappert, C. Sun, R. Orazi, and T. Weiner, Photonic link technology for
shipboard RF signal distribution, in Proc. SPIE, vol. 3463, pp. 123134,
[67] K. J. Williams, L. T. Nichols, and R. D. Esman, Photodetector nonlinearity on a high-dynamic range 3 GHz fiber optic link, J. Lightw. Technol.,
vol. 16, no. 2, pp. 192199, Feb. 1998.
[68] D. Baldwin, Dynamic range measurements comparing DFB laser links
to Mach-Zehnder modulator optical links for wide-band microwave transmission, in Proc. Photon. Syst. Antenna Appl. Conf., Monterey, CA, USA,
1993, pp. 5154.
[69] G. Betts, C. Cox, and K. Ray, 20 GHz optical analog link using an external
modulator, IEEE Photon. Technol. Lett., vol. 2, no. 12, pp. 923925,
Dec. 1990.
[70] G. Betts, F. ODonnell, J. Donnelly, J. Walpole, S. Groves, Z. Liau,
L. Missaggia, R. Bailey, K. Ray, and L. Johnson, Microwave links using
externally modulated semiconductor laser sources, in Proc. Photon. Syst.
Antenna Appl. Conf., Monterey, CA, USA, 1998, pp. 2831.
[71] M. LaGasse, W. Charczenko, M. Hamilton, and S. Thaniyavarn, Optical
carrier filtering for high dynamic range fiber optic links, Electron. Lett.,
vol. 30, pp. 21572158, Dec. 1994.


[72] J. Shah, DARPAs EPIC program, electronic and photonic integrated

circuits on silicon, presented at the IEEE Int. Conf. Group IV Photonics,
Antwerp, Belgium, Sep. 2005.
[73] K. Y. Tu, M. S. Rasras, D. M. Gill, S. S. Patel, Y. K. Chen, A. E.White,
A. Pomerene, D. Carothers, J. Beattie, M. Beals, J. Michel, and L. C.
Kimerling, Silicon RF-photonic filter and down converter, J. Lightw.
Technol., vol. 28, no. 20, pp. 30193028, Oct. 2010.

Richard W. Ridgway (M80SM05) received the B.S. and the M.S. degrees in
electrical engineering from the University of Michigan, Ann Arbor, MI, USA,
in 1978 and 1979, respectively, and the Ph.D. degree in electrical engineering
from The Ohio State University, Columbus, OH, USA, in 1985.
In July 2011, he joined the U.S. Defense Advanced Research Projects Agency
(DARPA), Arlington, VA, USA, where he is currently a Program Manager for
the Strategic Technology Office. His interests include millimeter-wave communications and microwave photonics. Prior to joining DARPA, he was a Senior
Research Leader at Battelle, Columbus, where for more than 25 years he was
involved in the development of integrated optical components for optical, microwave, and millimeter-wave communication systems. From 2001 to 2007, he
served as Chief Technology Officer for the Battelle-spinout company, Optimer
Photonics, Inc., focused on bringing electrooptic waveguide technology to the
telecommunications industry. He holds 26 US patents.


Carl L. Dohrman (M08SM13) received the B.S. degree from the University
of Illinois, Urbana, IL, USA, in 2002 and the Ph.D. degree from the Massachusetts Institute of Technology, Cambridge, MA, USA, in 2008, both in
materials science and engineering.
Since 2008, he has been a Consultant with Booz Allen Hamilton, Inc., where
he has provided subject matter expertise in microelectronic and photonic materials, devices, and circuits to the Microsystems Technology Office and Strategic
Technology Office of the U.S. Defense Advanced Research Projects Agency. He
has advised on a range of projects covering microwave photonics, photonic integrated circuits, device-level heterogeneous integration technologies, RF/mixed
signal electronics, and nitride optoelectronics. He has more than 30 refereed
journal and conference publications, and one patent.

Joshua A. Conway (M13SM13) received the B.S. degree in physics and the
M.S. degree in electrical engineering from the University of Illinois, Urbana,
IL, USA, in 1999 and 2001, respectively, and the Ph.D. degree in electrical
engineering from the University of California, Los Angeles, CA, USA, in 2006.
Since August 2012, he has served as a Program Manager for the Microsystems
Technology Office at the U.S. Defense Advanced Research Projects Agency
(DARPA), Arlington, VA, USA. His interests include active integrated photonic
devices, RF photonics and advanced imaging systems. Prior to joining DARPA,
he was with Kinsey Technical Services (KTSi), where he was Senior Principal
Engineer of special programs at the Los Angeles Air Force Base. Prior to
joining KTSi, he served on the technical staff of The Aerospace Corporation,
starting in 2003. From 2001 to 2003, he worked at Boeing Satellite Systems,
where he designed, built, and tested fiber-optic subsystems for inter-satellite
laser communication systems.
Dr. Conway has received numerous awards, has authored more than 30
technical papers and conference proceedings and holds eleven patents.