Coherent LiDAR With an 8,192-Element Optical Phased Array and Driving Laser

Abstract

We present recent advancements of optical phased array LiDAR with record-performance system demonstrations. First, we give an overview of the technology and how it combines the benefits of coherent LiDAR with the integration capabilities of silicon photonics. Then, an 8192-element optical phased array is shown with individually-addressed elements driven by custom CMOS electronics. This compact chip-scale beam-steering engine is enabled by flip-chip attached ASICs on the photonic integrated circuit. The optical phase shifters and emitters in the array have a 1
$\mu$
m pitch to enable a
$100^{\circ }\times 17^\circ$
field of view. This unprecedented number of active elements forms a near centimeter-scale aperture. Next, a high-performance laser uniquely suited for optical phased array LiDAR is demonstrated. Due to the silicon photonics cavity, it simultaneously supports a large tuning range (60 nm), low linewidth (
$\sim$
50 kHz), and fast linear chirp (1.3 GHz in 17.5
$\mu$
s). Finally, a solid-state coherent LiDAR system is realized with transmit/receive optical phased arrays coupled to an on-chip coherent receiver while being driven by the demonstrated laser. Ranging is shown on diffusive targets while simultaneously extracting velocity at each point in the point cloud. To the best of our knowledge, this single-unit compact system represents the state-of-the-art in optical phased array LiDAR technology.

Full text

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 1
Coherent LiDAR with an 8,192-Element
Optical Phased Array and Driving Laser
Christopher V. Poulton, Matthew J. Byrd, Peter Russo, Benjamin Moss,
Oleg Shatrovoy, Murshed Khandaker, and Michael R. Watts
(Invited Paper)
Abstract—We present recent advancements of optical phased
array LiDAR with record-performance system demonstrations.
First, we give an overview of the technology and how it combines
the benefits of coherent LiDAR with the integration capabilities
of silicon photonics. Then, an 8192-element optical phased array
is shown with individually-addressed elements driven by cus-
tom CMOS electronics. This compact chip-scale beam-steering
engine is enabled by flip-chip attached ASICs on the photonic
integrated circuit. The optical phase shifters and emitters in
the array have a 1µm pitch to enable a 100◦×17◦field of
view. This unprecedented number of active elements forms a
near centimeter-scale aperture. Next, a high-performance laser
uniquely suited for optical phased array LiDAR is demonstrated.
Due to the silicon photonics cavity, it simultaneously supports a
large tuning range (60 nm), low linewidth (∼50 kHz), and fast
linear chirp (1.3 GHz in 17.5 µs). Finally, a solid-state coherent
LiDAR system is realized with transmit/receive optical phased
arrays coupled to an on-chip coherent receiver while being driven
by the demonstrated laser. Ranging is shown on diffusive targets
while simultaneously extracting velocity at each point in the point
cloud. To the best of our knowledge, this single-unit compact
system represents the state-of-the-art in optical phased array
LiDAR technology.
Index Terms—Silicon photonics, optical phased arrays, solid-
state beam steering, lasers, coherent, LiDAR.
I. INTRODUCTION
LIGHT detection and ranging (LiDAR) has become an
ubiquitous sensing technology due to its long range
and high angular resolution in automotive [1], robotics [2],
and consumer electronics [3] applications. Specifically in
autonomous vehicles, LiDAR is expected to be required within
sensor fusion systems present in Level 3 and Level 4 au-
tomation [4]. This application has a long history with LiDAR
starting from the Defense Advanced Research Project Agency
(DARPA) Grand Challenge in 2004 where LiDAR was used
by the front runner Red Team from Columbia University [5].
Through that challenge, and events following afterwards, it
was clear that LiDAR had a place in autonomous vehicle
perception. Since then, the expectation that future robi-taxi
Manuscript received XXX, 2021; revised YYY, 2022; accepted ZZZ,
2022. This work was supported by the Defense Advanced Research Projects
Agency (DARPA) of the United States under the Modular Optical Aperture
Building Blocks (MOABB) program under grant number HR0011-16-C-0108.
(Corresponding author: Christopher Vincent Poulton.)
The authors are with Analog Photonics, Boston, MA 02110 USA
(e-mail: cpoulton@analogphotonics.com; mbyrd@analogphotonics.com;
peter@analogphotonics.com; benm@analogphotonics.com; osha-
trovoy@analogphotonics.com; mkhandaker@analogphotonics.com;
mwatts@analogphotonics.com.
and consumer autonomous vehicles will have on-board LiDAR
has driven the market with numerous emerging companies and
potential solutions [6].
LiDAR detection modalities can be broken up into two
primary categories: (1) Direct detection, also known as time-
of-flight (TOF), and (2) Coherent detection [7]. Currently, the
majority of LiDAR systems utilize TOF due to its ease of im-
plementation with readily available pulsed 905 nm lasers and
single-photon avalanche diodes (SPAD). However, recently
coherent LiDAR has become of interest due to its advantages
of ambient light insensitivity, direct detection of velocity, and
moderate peak powers for eye safety considerations (especially
at 1550 nm). Likewise, the beam steering mechanism of a
LiDAR system can nominally be specified as mechanical or
solid-state [8]. Mechanical solutions, such as a gimbal or
rotating mirror, can decrease reliability, limit scan rate and
regions of interest, and increase overall cost. Therefore, solid-
state beam steering has been a recent focus of LiDAR systems
with potential technologies such as microelectromechanical
systems (MEMS) [9], liquid crystals [10], and optical phased
arrays (OPAs) [11].
Photonic integrated circuits (PICs) containing OPAs enable
solid-state beam steering with near-arbitrary radiation control
[11], [12], [13], [14], [15], [16]. Specifically, silicon photonics
offers an integration platform for that takes advantage of the
mature silicon fabrication and packaging facilities used in the
CMOS electronics market [17]. In addition, silicon photonic
PICs can include high-performance optical components, such
as coherent detectors, enabling the realization of a single-chip
LiDAR system which combines beam steering and detection.
Overall, OPA PICs have the promise to produce inexpensive
and mass-producible solid-state coherent LiDARs on 300 mm
silicon wafers.
In this work, we present state-of-the-art demonstrations in
optical phased arrays, lasers designed for coherent LiDAR,
and an optical phased array LiDAR system. First, we give a
technology overview of optical phased array coherent LiDAR
in silicon photonics. Then, a reticle-sized optical phased array
with 8,192 individually controlled elements is shown. This
record-breaking element count is enabled through wafer-scale
processing and flip-chip CMOS electronics. Eight CMOS
ASICs, each controlling 1,024 elements, are flipped directly
onto the PIC to form a chip-scale beam steering engine which
consumes 2.5 W (300 µW/DAC). The phase shifter and emitter
elements have a tight pitch of 1 µm to maximize steering
range, but still form a near centimeter-scale aperture (8 mm

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 2
×5 mm) due to the large element count. Beam steering is
demonstrated over a 100◦×17◦field of view with a FWHM
diffraction angle of 0.01◦×0.039◦and a 10 dB side-lobe sup-
pression. Next, a laser designed for coherent LiDAR is shown
while being driven by low-noise custom electronics. This
laser simultaneously achieves a large tuning range (60nm),
low linewidth (∼50kHz), and fast linear chirp (1.3 GHz in
17.5 µs). These specifications are accomplished by utilizing a
high-performance silicon photonics cavity coupled to optical
gain material. Finally, a single-unit solid-state coherent LiDAR
system containing optical phased arrays, on-chip coherent
detectors, and custom laser is demonstrated. The on-board
FPGAs drive OPA and laser modules to perform rapid 2D
beam steering with real-time DSP on the return signal to
simultaneously extract range and velocity of targets. These
demonstrations mark firsts for optical phased array LiDAR and
pave the way for the advancement of the technology within
next-generation LiDAR systems.
II. OPTICAL PHA SED AR RAY LIDAR TECHNOLOGY
Optical phased array LiDAR combines a coherent LiDAR
detection modality with the integration of silicon photonics
[Fig. 1]. Coherent LiDAR is a recent LiDAR trend that mimics
the trajectory originally taken by RADAR [18]. It has garnered
interest due to its numerous benefits over traditional time-of-
flight detection, such as shot-noise-limited performance with
coherent receivers and simple linear-mode photodetectors. Ad-
ditionally, coherent detection has intrinsic rejection of external
light interference, such as from sunlight, bright light sources,
and adjacent LiDAR systems, since only received signals
coherent with the local oscillator are amplified. Finally, it
can directly measure both range and velocity at each point
in the point cloud [19] without the need for differentiation
over multiple frames, easing the requirements of perception
algorithms. However, coherent LiDAR is notoriously difficult
to implement in free-space systems due to its single-mode
nature and having a stringent requirement of precisely aligned
low-wavefront-error optics. Silicon photonics is a natural plat-
form to implement coherent LiDAR systems with its single-
mode waveguides and lithographically defined structures. This
technology can support the entire photon path in a LiDAR
system, starting with the generation of light (with integrated
gain material [20], [21]), transmitting and receiving beams
through optical phased arrays, and finally photodetection with
on-chip coherent receivers [22]. Silicon photonic integrated
circuits can be fabricated on 300 mm wafers at commercial
semiconductor foundries [23], [24], [25] in order to produce
an inexpensive integrated LiDAR system that can support large
market volumes.
On the implementation side, optical phased array LiDAR
has three primary technology pillars: (1) Optical phased arrays,
(2) CMOS electronics, and (3) Laser sources. Optical phased
arrays enable solid-state beam steering from a large flat on-
chip aperture. Similar to microwave phased arrays commonly
used in RADAR [26], optical phased arrays achieve beam
steering by tuning the input phases of an arrayed antenna
element. Phase tuning, and thus beam steering, can be achieved
through active phase shifters or by changing wavelength when
grating-based emitters are used. A common optical phased
array architecture is a one-dimensional array of grating-based
antennas, each with a corresponding phase shifter [11]. This
allows for a tight pitch (large steering angle) in one dimension
by active phase tuning, while the other dimension is steered by
altering the input wavelength. Wavelength tuning can limit the
steering range in one dimension and two-dimensional optical
phased arrays have been proposed to mitigate this issue [27],
[28]. Additionally, unique optical phased array architectures
have been shown such as serpentine optical phased arrays
Silicon Photonics
Optical Phased Array LiDAR
Optical Phased Arrays CMOS Electronics Laser Sources
Coherent LiDAR
∆t
fVel
Op�cal
Freq
Beat
Freq
Time
LO
RX
fIF2
fIF1
2fDoppler
fDis
ŝ
Integration
Platform
Detection
Modality
Technology Pillars
Fig. 1. Optical phased array LiDAR as the combination of coherent LiDAR and silicon photonics. Its technology pillars consist of optical phased arrays,
CMOS electronics, and laser sources.

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 3
(SOPA) [29], dual-layer optical phased arrays with silicon and
silicon nitride [30], and sparse end-fire antennas driven by
visible light [31].
Scaling aperture size by increasing OPA element count
is critical for high-performance systems. Passive OPAs have
been demonstrated with >10,000 elements [16], however,
implementing active control has been a challenge. OPAs with
512 active elements have been demonstrated by wirebonding
to external digital-to-analog converters (DACs) [11], [12], and
1,024 elements was achieved with monolithic CMOS pho-
tonics [13]. However, ∼10,000 active elements are required
to form the centimeter-scale apertures needed in long-range
applications such as automotive LiDAR. Though randomly-
pitched elements can mitigate this need while supporting a
large steering range [14], this increases optical loss due to
far field side-lobe “smearing”. Therefore, CMOS electronics
drivers are necessarily, the second technology pillar of optical
phased array LiDAR. Monolithic CMOS [13], [23], wafer-
bonding [15], and flip-chip [32] techniques are options for
integration, with flip-chip being a widely-available low-cost
solution. Besides, with a ∼10,000 element count, it is crucial
for CMOS DAC drivers to have a power consumption on the
order of ∼100 µW/DAC (i.e. single-digit Watt total). Finally,
laser sources are the last technology pillar for an optical phased
array LiDAR system. As will be discussed in Section IV,
these lasers must simultaneously meet the requirements of
wavelength tunability (when one-dimensional OPAs are used),
narrow linewidth, and rapid linear chirp capability.
III. OPT ICA L PHAS ED ARR AY WITH CMOS CONT ROL
As mentioned in Section II, optical phased arrays enable
solid-state beam steering with flat conformal on-chip aper-
tures. Previous work has focused on decreasing element pitch
towards λ/2to enable wide field of views with pushed out
grating lobes. However, it is crucial to simultaneously increase
the element count to maintain a large aperture that supports
small diffraction angles and adequate receiver area. In this
section, we describe an optical phased array which utilizes flip-
chip CMOS electronics in order to scale the element count.
Figure 2(a) shows the 2.8 cm ×2.5 cm PIC containing a one-
dimensional 8,192-element OPA. The light is distributed using
a splitter tree with 8,192 outputs, each output is connected to
a single phase shifter, which is then connected to an emitter.
The optical phase shifters and emitters are placed at a uniform
1µm pitch (∼0.65λ), facilitating 100◦steering in one dimen-
sion. Wavelength-based steering in the second dimension is
achieved through silicon waveguide grating antennas that are
5 mm long and 650 nm wide [33]. This inline architecture
with pitch-matched phase shifters and emitters is inherently
scalable with no footprint overhead associated with sparsely-
pitched phase shifters fanning into tightly-pitched antennas.
Due to the large element count, low power optical phase
shifters are essential for practical devices and therefore electro-
optic phase shifters were utilized. These phase shifters enable
microsecond-level steering rates with no electrical cross talk
and were designed to minimize interaction with the electrical
contacts and coupling to adjacent phase shifters [34].
Outside of the 8 mm ×5 mm emission aperture, eight
identical CMOS die are flip-chip attached to the PIC, each
containing 1,024 DAC outputs. The DAC voltage output is
generated with an R-2R ladder and has 8 bits of resolution
which does not affect the steering resolution. The electrical
traces on the PIC act as an interposer to route each DAC
output to a respective optical phase shifter (8×1,024, 8,192
total), and CMOS control signals to wirebond pads on the
PIC perimeter. The packaged multi-chip module and driving
microcontroller is contained within a housing as shown in
Fig. 2(b). The microcontroller drives the digital control signals,
including the clock, to the CMOS die using a proprietary data
interface. The OPA PIC is fabricated on 300 mm silicon wafers
and an image is shown in Fig. 2(c) where each reticle contains
the 8,192-element OPA. Similar CMOS wafers were fabricated
at a commercial foundry and then bumped, singulated, and
flip-chip attached to PIC. Figure 2(d) shows an SEM image
of one of the CMOS die after attach with an inset of the
bump connections. Each OPA PIC has a total of 16,856 bump
connections which includes DAC outputs, power rails, and
control signals.
After assembly and packaging, the phase distribution of the
elements was calibrated, and beam steering was achieved by
applying a linear front using the phase shifters and a known
phase-voltage relationship [35]. By utilizing this relationship,
DAC non-idealities, such as non-linearity, are effectively cal-
ibrated out and does not greatly affect the performance of
the system including the steering resolution and beam width.
Due to the large steering angle, the unit was first measured
by placing it incident to a wall and observing the scattered
light on an IR camera. Figure 3(a) shows summed IR camera
images of individual spots while performing 2D beam steering
over a 100◦×17◦field of view (120nm wavelength tuning).
Then, a far field imaging setup was used to measure a FWHM
diffraction angle of 0.01◦×0.039◦with a 10 dB side-lobe
suppression ratio [Fig. 3(b)]. Element yield was measured by
attempting to sweep individual DACs and observing changes
CMOS
300mm
Photonics Wafer
2.8cm × 2.5 cm PIC
8192 Elements
Fiber input
CMOS
×8
5mm
(d)(c)
(b)(a)
Fig. 2. (a) OPA PIC with flip-chip CMOS. (b) Photograph of packaged
demonstrator in housing. (c) 300 mm photonics wafer with 8,192-element
OPAs. (d) SEM of CMOS flipped onto PIC with inset of connections.

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 4
OPA
θ [Degrees]
1
Intensity [arb]
-0.1 0.05
0.1
0
-0.05 0
Far Field
(b)
Measured
Theory
0.01
o
/
175
μ
Rad
(c)
Input Power [dBm]
Radiated Power [dBm]
26
22
18
14
25 29 33 37
(d)
(a)
0.6 0.7 0.8 0.9 1
R
2
of Fit to Expected Response
0.6
0.7
0.8
0.9
1
2
Example fit with R2 = ~0.8
>95% with R2 >0.8
100
o
× 17
o
Fig. 3. (a) Sum of spots on wall showing 2D beam steering. (b) Far field and
cross-section in the θdimension. (c) Yield of the 8,192 elements as a function
of R2with example measurement and fit. (d) Measured output radiated beam
power.
in the emission pattern. When fitting to theory (a sinusoid)
and assuming that a fit of R2>0.8 represents a successful con-
nection, the measured yield was >95%, which includes flip-
chip and wirebond connections along with the CMOS and PIC
components [Fig. 3(c)]. The power consumption of the chip-
scale beam steering engine is primarily from the CMOS DAC
drivers as the optical phase shifters themselves consume single
microwatt-level power levels [11]. The measured power of a
single CMOS die was 300 mW (300 µW/DAC), corresponding
to a total power consumption of 2.5 W for all eight die. Finally,
large optical powers are crucial for long-range applications,
and care was taken to minimize any non-linear waveguide
loss throughout the PIC. The power handling capability of
the unit was measured by inputting light from an EDFA and
measuring the radiated beam power coming from the aperture.
Figure 3(d) shows linear operation up to an input power of 5 W
with ∼11 dB insertion loss, including fiber coupling, resulting
in 400 mW being emitted from the aperture.
Finally, Table I shows the performance of recent state-of-
the-art OPA demonstrations. The OPA shown here establishes
new records in element count, steering range, and aperture
size. In addition, the complementing low-power driving CMOS
DACs enable an adaptable beam-steering module that can
be placed in a variety of system form factors for LiDAR
and free-space communication. This demonstration pushes
the boundaries of integrated photonics technology with an
unprecedented number of active components and shows the
potential for complex PICs in challenging applications.
TABLE I
Comparison with Recent Art
Reference # Elements θSteering Aperture Size
This work 8,192 100◦8 mm
[11] 512 56◦0.84 mm
[12] 512 70◦0.67 mm
[13] 1,024 45◦2 mm
IV. LAS ERS FOR OPTICAL PHA SED AR RAY LIDAR
Long-range coherent LiDAR necessitates a particular laser
source due the simultaneous requirements of a narrow
linewidth (long coherence length) and a fast linear frequency
chirp for range detection. Additionally, the use of wavelength
steering in one-dimensional optical phased arrays also requires
that the driving laser is tunable in order to perform 2D beam
steering. These characteristics necessitated the development
of a custom laser solution that simultaneously achieved these
three requirements. A block diagram of the laser cavity is
presented in Fig. 4(a) where a state-of-the-art silicon photonics
process was utilized to create the tuning and feedback elements
in the laser [36]. The silicon photonic wavelength filter enables
a large wavelength range with power efficient tuning, while an
integrated phase shifter allows for rapid frequency chirping
of the laser. Additionally, on-chip feedback sensors allow for
self-contained calibration and collect data on laser health and
performance. The silicon photonics is coupled to optical gain
material in order to form the full laser cavity.
The assembled laser described by Fig. 4(a) was packaged
in a butterfly package to protect against harmful contaminants
and physical damage. The goldbox contains a pigtailed fiber
and its leads are connected to the die pads through wirebonds.
This package was soldered to a breakout PCB to form a
connectorized laser module which is mounted to a custom
driving PCB. The low-noise electronic circuits were developed
to monitor and digitize the feedback signals from the laser
cavity while providing proper driving signals for laser opera-
tion. An FPGA system-on-module with specialized firmware is
connected to the driving PCB which performs laser calibration,
specified wavelength tuning, and linear frequency chirping.
This all inclusive system shown in Fig. 4(b) allows users to
utilize the custom tunable laser for LiDAR measurements and
systems with only power and data connections.
Prior to integration with the LiDAR system in Section V,
the laser module performance was characterized with various
measurements to observe performance metrics that relate to
OPA-based coherent LiDAR. The first measurement was to
characterize the wavelength tunability which directly impacts
the field of view of the LiDAR system. The wavelength filter
within the laser cavity was scanned in order to tune the
wavelength and a look-up-table of voltages was generated
consisting of each valid wavelength point. To enable in-field
and on-the-fly calibration, this step was completed without the
use of external measurement equipment and relied solely on
the on-chip feedback sensors to ensure that each wavelength
point was both unique and single-mode (i.e. large side mode

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 5
(a)
Phase
Shifter
Chirp
Actuation
λ
Filter
Wavelength
Control
Bias
Current
Gain Feedback
Monitor
Signals
(b)
Laser
Driving PCB
Controlling FPGA
Fig. 4. (a) Block diagram of laser designed for coherent LiDAR. (b) Image
of laser in butterfly package driven by custom electronics.
suppression ratio). A overlay of optical spectra from each cal-
ibrated wavelength point as measured by an optical spectrum
analyzer is shown in Fig. 5(a). Once calibrated, the laser is
capable of tuning over approximately 60 nm, maintains a side
mode suppression ratio of about 50 dB, and has nearly uniform
output power over the entire tuning range.
The next important performance metric to measure is the
linewidth of the laser as long coherence lengths are required in
coherent LiDAR systems to produce strong and sharp returns
at long ranges. To characterize the linewidth, the frequency
noise spectrum was measured with a commercial-off-the-
shelf equipment that is based on a homodyne methodology
[Fig. 5(b)]. This plot shows that the frequency noise spectrum
flattens out at around 100 kHz and achieves a white noise
value corresponding to approximately a 50 kHz Lorentzian
linewidth. The frequency noise spectrum presented in Fig. 5(b)
enables several microsecond long integration times without
significant signal degradation at ranges exceeding 100 m.
The final critical performance metric of this laser is the
ability to create a rapid linear frequency chirp. Specifically,
a linear frequency chirp is required to minimize excess loss
and mitigates the need for complex signal processing and re-
timing of return signals. The firmware on the custom driving
PCB FPGA was programmed with an algorithm that optimizes
the control signal waveform used to chirp the laser frequency
by utilizing the output from a reference fiber Mach-Zehnder
interferometer. An example result from one calibrated point
is shown in Fig. 5(c). Without optimization, the chirp rate
deviates significantly from the desired constant value, but the
optimization routine was able to adjust the chirp waveform
to account for these nonlinearities and achieve a constant
chirp rate. A linear chirp rate of 75 THz/s was achieved over
(a)
(b) (c)
103104105106107108
Frequency [Hz]
103
104
105
106
Frequency Noise [Hz2/Hz]
10 15
Time [µs]
50
60
70
80
90
Chirp Rate [THz/s]
Pre-Lineariza�on
Post-Lineariza�on
50
1510 1520 1530 1540 1550 1560 1570
Wavelength [nm]
-60
-50
-40
-30
-20
-10
10
Power [dBm]
0
1.3GHz Excursion
Over 17.5µs
50kHz Lorentzian
Linewidth
60nm
>50dB
Fig. 5. (a) Overlaid laser spectra showing a 60 nm tuning range. (b) Measured
frequency noise of laser. (c) Measured chirp rate of laser after linearizion
algorithm.
17.5 µs which produces a 1.3 GHz chirp excursion. The small
variations from a perfectly constant chirp rate in Fig. 5(c) is
caused by the non-zero laser linewidth and causes ∼0.5 dB
loss at a 50 m target range.
V. LIDAR SYSTE M DEMONSTRATION
The combination of optical phased arrays, driving CMOS
electronics, and lasers enables solid-state coherent LiDAR
systems. A LiDAR system was created by utilizing a photonic
integrated circuit consisting of transmitter and receiver optical
phased arrays (8,192 elements total) coupled to an on-chip
coherent detector. The element phase shifters were driven with
flip-chip CMOS electronics in a similar manner as the optical
phased array shown in Section III. The current output of the
coherent detector was wirebonded to a custom transimpedance
amplifier (TIA) ASIC to provide current-to-voltage conver-
sion, and this signal was digitized using a commercial-off-
the-shelf ADC component. This OPA sub-system was driven
by a FPGA system-on-module which controlled the CMOS
phase shifter drivers, communicated with peripherals such as
the ADC and auxiliary DACs, and performed DSP on the
LiDAR return signal while outputting processed data to the
user.
The laser described in Section IV was used within a laser
sub-system to be combined with the OPA sub-system. This
laser sub-system is similar to Fig. 4(b). The laser output was
coupled to the OPA photonic integrated circuit and both sub-
systems were combined in a single LiDAR head as shown
in Fig. 6. The single-unit LiDAR system is powered with a
standard 12 V connection and is communicated to and from
through a single Ethernet port. The two controlling FPGA
system-on-modules are synchronized to each other through
a number of control lines which enable 2D raster scanning
with phase shifter control from the OPA sub-system and laser
wavelength control from the laser sub-system. Besides, a simi-
lar module could be used for free-space optical communication

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 6
Op�cal Phased Array LiDAR System
Fig. 6. LiDAR system containing optical phased arrays, custom CMOS
electronics, and driving laser source.
with the addition of an optical modulator which could be on
or off-chip.
In order to perform LiDAR, a constant false alarm rate
(CFAR) algorithm is utilized by disabling the transmit beam
and choosing a threshold that results in a desired false alarm
rate.
A∼50mW beam is emitted from the OPA and is raster
scanned through the scene. It takes ∼10µs to load data into
the DACs which then takes ∼1µs to settle afterwards. At each
point the laser has an up-sweep chirp and a down-sweep chirp.
During this time the coherent detector signal is captured by
the FPGA for processing. The on-board DSP is effectively
a simple Fourier Transform of the return signal as no post-
processing is required due to the precise linear chirp of the
laser. Points with magnitude greater than the aforementioned
chosen threshold is returned to the user through Ethernet.
Custom client-side software was written to control the LiDAR
unit and further process and display point cloud data. The
primary processing functions are peak detection for a given
point (i.e. showing the largest return for each point) and
Doppler compensation which extracts range and velocity from
the up-sweep and down-sweep point clouds. This is done
by calculating the up-sweep and down-sweep beat frequency
average (proportional to range) and difference (proportional to
velocity) for each individual point.
Figure 7(a) shows output point cloud data from the unit in
an indoor environment along with an image of the unit and
environment. In this image, the total number of spots in the
point cloud is 154 by 20 and each point takes 30 µs. Here,
peak detection and Doppler compensation is turned on such
that the color of each individual point is representative of its
velocity. Ranged targets include common diffuse objects such
as a person, walls, and chairs where their features can clearly
be seen. Additionally, a 10% Lambertian target is seen at
∼35 m with a strong return. In addition, insets are shown of an
(a)
(b)
10% Lamber�an
Target at ~35m
Image on Different Day
Trees
Sign
Building
Cars
Fig. 7. (a) LiDAR point cloud from system taken inside. Insets show example
people performing activities. (b) Point cloud taken outdoors with an inset of
an image of the location on a different day. Color is representative of velocity
in all point clouds.
individual playing basketball and running on a treadmill where
the velocity detection is clearly beneficial. It is believed that
this capability will be integral for classification algorithms of
perception systems. Finally, the unit was taken outside to test
its performance. As discussed in Section II, background light,
such as sunlight, is expected to create minimum performance
degradation due to the inherent rejection from coherent detec-
tion and the use of a directional receiver. Figure 7(b) shows the
point cloud where cars, a street sign, trees, and a building can
be seen. The inset is an image of the scene taken on a different
day. The system range is currently limited by optical loss in
the transmit and receive path. Other possible improvements
include increasing the receiver aperture and decreasing the
linewidth of the laser.
VI. CONCLUSION
In conclusion, we have presented high-performance demon-
strators related to optical phased array LiDAR technology.
Optical phased array LiDAR combines the benefits of coherent
LiDAR with the integration of silicon photonics. It utilizes
three technology pillars: Optical phased arrays, CMOS elec-
tronics, and laser sources. An 8192-element optical phased ar-
ray was shown that was driven by flip-chip CMOS electronics
and achieved a 100◦×17◦field of view by utilizing 1µm-
pitched phase shifters and emitters. It establishes new records
in element count, steering range, and aperture size, and the
complementing low-power driving CMOS enables an adapt-
able beam-steering module. Overall, it pushes the boundaries
of integrated photonics technology with an unprecedented
number of active components. Additionally, a first-of-its-kind

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 7
laser was demonstrated that was specially designed for optical
phased array LiDAR. Due to its silicon photonics cavity,
it simultaneously achieves a large tuning range (60nm) for
wavelength steering, fast linear chirp capability (1.3GHz in
17.5 µs) for coherent LiDAR, and low linewidth (∼50 kHz)
for long-range sensing. Lastly, optical phased arrays, CMOS
electronics, and lasers were combined to form a single-
unit solid-state coherent LiDAR system that measured range
and velocity simultaneously. Interference rejection from the
coherent receiver and directional optical phased array enabled
detection of diffuse targets outdoors in sunlight. Overall, this
work presents the most advanced demonstrations of optical
phased arrays and OPA-based LiDAR to date, and shows
the capability and increasing maturity of the technology.
Fabricated on 300 mm silicon wafers, it is uniquely situated
as a leading low-cost integrated coherent LiDAR solution.
ACK NOWLEDGME NTS
The authors acknowledge all employees of Analog Pho-
tonics for their help and support in this work. The views,
opinions and/or findings expressed are those of the author and
should not be interpreted as representing the official views or
policies of the Department of Defense or the U.S. Government.
Released under Distribution Statement “A” (Approved for
Public Release, Distribution Unlimited).
REFERENCES
[1] R. Thakur, “Scanning lidar in advanced driver assistance systems and
beyond: Building a road map for next-generation lidar technology,” IEEE
Consumer Electronics Magazine, vol. 5, no. 3, pp. 48–54, 2016.
[2] J. Lalonde, N. V., D. F. Huber, and M. Hebert, “Natural terrain classi-
fication using three-dimensional ladar data for ground robot mobility,”
Journal of Field Robotics, vol. 23, pp. 839–861, 2006.
[3] A. Spreafico, F. Chiabrando, L. T. Los`
e, and F. G. Tonolo, “The ipad pro
built-in lidar sensor: 3d rapid mapping tests and quality assessment,” The
International Archives of Photogrammetry, Remote Sensing and Spatial
Information Sciences, vol. 43, pp. 63–69, 2021.
[4] J. Janai, F. G¨
uney, A. Behl, and A. Geiger, “Computer vision for au-
tonomous vehicles: Problems, datasets and state of the art,” Foundations
and Trends in Computer Graphics and Vision, vol. 12, no. 1–3, pp. 1–
308, 2020.
[5] C. Urmson, J. Anhalt, D. Bartz, M. Clark, T. Galatali, A. Gutierrez,
S. Harbaugh, J. Johnston, H. Kato, P. L. Koon, W. Messner, N. Miller,
A. Mosher, K. Peterson, C. Ragusa, D. Ray, B. K. Smith, J. M.
Snider, S. Spiker, J. C. Struble, J. Ziglar, and W. R. L. Whittaker, “A
robust approach to high-speed navigation for unrehearsed desert terrain,”
Journal of Field Robotics, vol. 23, no. 8, pp. 467 – 508, Aug 2006.
[6] R. Roriz, J. Cabral, and T. Gomes, “Automotive lidar technology: A
survey,” IEEE Transactions on Intelligent Transportation Systems, pp.
1–16, 2021.
[7] S. Royo and M. Ballesta-Garcia, “An overview of lidar imaging systems
for autonomous vehicles,” Applied Sciences, vol. 9, no. 19, 2019.
[8] T. Raj, F. H. Hashim, A. B. Huddin, M. F. Ibrahim, and A. Hussain, “A
survey on lidar scanning mechanisms,” Electronics, vol. 9, no. 5, 2020.
[9] D. Wang, C. Watkins, and H. Xie, “Mems mirrors for lidar: A review,”
Micromachines, vol. 11, no. 5, 2020.
[10] S. R. Davis, S. D. Rommel, D. Gann, B. Luey, J. D. Gamble,
M. Ziemkiewicz, and M. Anderson, “A lightweight, rugged, solid
state laser radar system enabled by non-mechanical electro-optic beam
steerers,” in Laser Radar Technology and Applications XXI, vol. 9832.
SPIE, 2016, pp. 186 – 196.
[11] C. V. Poulton, M. J. Byrd, P. Russo, E. Timurdogan, M. Khandaker,
D. Vermeulen, and M. R. Watts, “Long-range lidar and free-space data
communication with high-performance optical phased arrays,” IEEE
Journal of Selected Topics in Quantum Electronics, vol. 25, no. 5, pp.
1–8, 2019.
[12] S. A. Miller, Y.-C. Chang, C. T. Phare, M. C. Shin, M. Zadka, S. P.
Roberts, B. Stern, X. Ji, A. Mohanty, O. A. J. Gordillo, U. D. Dave,
and M. Lipson, “Large-scale optical phased array using a low-power
multi-pass silicon photonic platform,” Optica, vol. 7, no. 1, pp. 3–6, Jan
2020.
[13] S. Chung, H. Abediasl, and H. Hashemi, “A monolithically integrated
large-scale optical phased array in silicon-on-insulator cmos,” IEEE
Journal of Solid-State Circuits, vol. 53, pp. 275–296, Jan 2018.
[14] D. N. Hutchison, J. Sun, J. K. Doylend, R. Kumar, J. Heck, W. Kim,
C. T. Phare, A. Feshali, and H. Rong, “High-resolution aliasing-free
optical beam steering,” Optica, vol. 3, no. 8, pp. 887–890, Aug 2016.
[15] T. Kim, P. Bhargava, C. V. Poulton, J. Notaros, A. Yaacobi, E. Timurdo-
gan, C. Baiocco, N. Fahrenkopf, S. Kruger, T. Ngai, Y. Timalsina, M. R.
Watts, and V. Stojanovi, “A single-chip optical phased array in a wafer-
scale silicon photonics/cmos 3d-integration platform,” IEEE Journal of
Solid-State Circuits, vol. 54, no. 11, pp. 3061–3074, 2019.
[16] M. Raval, A. Yaacobi, and M. R. Watts, “Integrated visible light phased
array system for autostereoscopic image projection,” Opt. Lett., vol. 43,
no. 15, pp. 3678–3681, Aug 2018.
[17] A. Rahim, T. Spuesens, R. Baets, and W. Bogaerts, “Open-access silicon
photonics: Current status and emerging initiatives,” Proceedings of the
IEEE, vol. 106, no. 12, pp. 2313–2330, 2018.
[18] A. G. Stove, “Linear fmcw radar techniques,” IEE Proceedings - Radar
and Signal Processing, vol. 139, no. 5, pp. 343–350, Oct 1992.
[19] D. Pierrottet, F. Amzajerdian, L. Petway, B. Barnes, G. Lockard, and
M. Rubio, “Linear fmcw laser radar for precision range and vector
velocity measurements,” MRS Proceedings, vol. 1076, 2008.
[20] B. Szelag, K. Hassan, L. Adelmini, E. Ghegin, P. Rodriguez,
F. Nemouchi, P. Brianceau, E. Vermande, A. Schembri, D. Carrara,
P. Cavali, F. Franchin, M.-C. Roure, L. Sanchez, C. Jany, and S. Olivier,
“Hybrid iiiv/silicon technology for laser integration on a 200-mm fully
cmos-compatible silicon photonics platform,” IEEE Journal of Selected
Topics in Quantum Electronics, vol. 25, no. 5, pp. 1–10, 2019.
[21] Z. Wang, A. Abbasi, U. Dave, A. De Groote, S. Kumari, B. Kunert,
C. Merckling, M. Pantouvaki, Y. Shi, B. Tian, K. Van Gasse, J. Ver-
bist, R. Wang, W. Xie, J. Zhang, Y. Zhu, J. Bauwelinck, X. Yin,
Z. Hens, J. Van Campenhout, B. Kuyken, R. Baets, G. Morthier,
D. Van Thourhout, and G. Roelkens, “Novel light source integration
approaches for silicon photonics,” Laser & Photonics Reviews, vol. 11,
no. 4, p. 1700063, 2017.
[22] C. Doerr and L. Chen, “Silicon photonics in optical coherent systems,”
Proceedings of the IEEE, vol. 106, no. 12, pp. 2291–2301, 2018.
[23] K. Giewont, K. Nummy, F. A. Anderson, J. Ayala, T. Barwicz, Y. Bian,
K. K. Dezfulian, D. M. Gill, T. Houghton, S. Hu, B. Peng, M. Rakowski,
S. Rauch, J. C. Rosenberg, A. Sahin, I. Stobert, and A. Stricker, “300-
mm monolithic silicon photonics foundry technology,” IEEE Journal of
Selected Topics in Quantum Electronics, vol. 25, no. 5, pp. 1–11, 2019.
[24] N. M. Fahrenkopf, C. McDonough, G. L. Leake, Z. Su, E. Timurdogan,
and D. D. Coolbaugh, “The aim photonics mpw: A highly accessible
cutting edge technology for rapid prototyping of photonic integrated
circuits,” IEEE Journal of Selected Topics in Quantum Electronics,
vol. 25, no. 5, pp. 1–6, 2019.
[25] A. Rahim, J. Goyvaerts, B. Szelag, J.-M. F´
ed´
eli, P. P. Absil, T. Aalto,
M. Harjanne, C. Littlejohns, G. T. Reed, G. Winzer, S. Lischke, L. Zim-
mermann, D. Knoll, D. H. Geuzebroek, A. Leinse, M. Geiselmann,
M. Zervas, H. Jans, A. Stassen, C. Dom´
ınguez, P. Mu˜
noz, D. Domenech,
A. L. Giesecke, M. C. Lemme, and R. Baets, “Open-access silicon
photonics platforms in europe,” IEEE Journal of Selected Topics in
Quantum Electronics, vol. 25, pp. 1–18, 2019.
[26] A. J. Fenn, D. H. Temme, W. P. Delaney, and W. E. Courtney, “The
development of phased-array radar technology,” Lincoln Laboratory
Journal, vol. 12, no. 2, pp. 321–340, 2000.
[27] J. Sun, E. shah Hosseini, A. Yaacobi, D. B. Cole, G. Leake, D. Cool-
baugh, and M. R. Watts, “Two-dimensional apodized silicon photonic
phased arrays,” Opt. Lett., vol. 39, no. 2, pp. 367–370, Jan 2014.
[28] H. Abediasl and H. Hashemi, “Monolithic optical phased-array
transceiver in a standard soi cmos process,” Opt. Express, vol. 23, no. 5,
pp. 6509–6519, Mar 2015.
[29] N. Dostart, B. Zhang, A. Khilo, M. Brand, K. A. Qubaisi, D. Onural,
D. Feldkhun, K. H. Wagner, and M. A. Popovi´
c, “Serpentine optical
phased arrays for scalable integrated photonic lidar beam steering,”
Optica, vol. 7, no. 6, pp. 726–733, Jun 2020.
[30] P. Wang, G. Luo, Y. Xu, Y. Li, Y. Su, J. Ma, R. Wang, Z. Yang, X. Zhou,
Y. Zhang, and J. Pan, “Design and fabrication of a sin-si dual-layer
optical phased array chip,” Photon. Res., vol. 8, no. 6, pp. 912–919, Jun
2020.

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. XX, NO. YY, ZZ 2022 8
[31] M. C. Shin, A. Mohanty, K. Watson, G. R. Bhatt, C. T. Phare, S. A.
Miller, M. Zadka, B. S. Lee, X. Ji, I. Datta, and M. Lipson, “Chip-scale
blue light phased array,” Opt. Lett., vol. 45, no. 7, pp. 1934–1937, Apr
2020.
[32] K. Li, S. Liu, D. J. Thomson, W. Zhang, X. Yan, F. Meng, C. G.
Littlejohns, H. Du, M. Banakar, M. Ebert, W. Cao, D. Tran, B. Chen,
A. Shakoor, P. Petropoulos, and G. T. Reed, “Electronic–photonic
convergence for silicon photonics transmitters beyond 100 gbps on–off
keying,” Optica, vol. 7, no. 11, pp. 1514–1516, Nov 2020.
[33] K. Shang, C. Qin, Y. Zhang, G. Liu, X. Xiao, S. Feng, and S. B.
Yoo, “Uniform emission, constant wavevector silicon grating surface
emitter for beam steering with ultra-sharp instantaneous field-of-view,”
Opt. Express, vol. 25, no. 17, pp. 19655–19661, Aug 2017.
[34] Z. Zhang, H. Yu, Q. Huang, Z. Zhou, B. Chen, T. Dai, H. Qiu, Y. Wang,
and J. Yang, “High-speed and low-power silicon optical phased array
based on the carrier-depletion mechanism,” IEEE Photonics Technology
Letters, vol. 34, no. 5, pp. 271–274, 2022.
[35] J. Jin, E.-S. Lee, K.-W. Chun, S.-S. Lee, and M.-C. Oh, “Fast-running
beamforming algorithm for optical phased array beam scanners com-
prised of polymeric waveguide devices,” Opt. Express, vol. 30, no. 2,
pp. 768–779, Jan 2022.
[36] H. Guan, A. Novack, T. Galfsky, Y. Ma, S. Fathololoumi, A. Horth,
T. N. Huynh, J. Roman, R. Shi, M. Caverley, Y. Liu, T. Baehr-Jones,
K. Bergman, and M. Hochberg, “Widely-tunable, narrow-linewidth iii-
v/silicon hybrid external-cavity laser for coherent communication,” Opt.
Express, vol. 26, no. 7, pp. 7920–7933, Apr 2018.
Christopher V. Poulton Christopher V. Poulton is currently the Vice President
of Silicon Photonics at Analog Photonics where he leads the silicon photonics
team developing highly integrated chip-scale LiDAR systems. He received
a B.S. degree in Electrical and Computer Engineering from the University
of Colorado Boulder in 2014, and a M.S. degree in Electrical Engineering
from the Massachusetts Institute of Technology (MIT) in 2016. He joined
the Photonic Microsystems Group, MIT, as an NSF Fellow and an MIT
Presidential Fellow. There, he performed under the DARPA E-PHI program
and demonstrated the first coherent LiDAR with silicon photonics and the
first millimeter-scale optical phased array. He has authored or coauthored
more than 50 peer reviewed journal and conference publications along with
several patents. He was the recipient of the Morris Joseph Levin Award for
Outstanding Masterworks and was denoted a DARPA Riser.
Matthew J. Byrd Matthew J. Byrd received the B.S. degree in Electrical
Engineering from Clemson University, Clemson, SC, in 2015 and the M.S.
in Electrical Engineering from the Massachusetts Institute of Technology,
Cambridge, MA in 2017. He joined the Photonics Microsystems Group, MIT,
leading the microwave photonics effort under the DARPA E-PHI program in
2016. Since graduation, he has worked at Analog Photonics as the Director
of Laser Development developing integrated chip-scale LiDAR systems.
Peter Russo Peter Russo is Vice President of LiDAR at Analog Photonics. He
received his Bachelor of Science in Electrical Engineering from University of
Maryland, College Park in 2008. After graduating, he joined BAE Systems as
part of the Engineering Leadership Development Program, through which he
also received his Master of Science in Electrical Engineering from University
of New Hampshire. At BAE Systems, he served as principle investigator on
several active electro-optical systems programs. In 2015, he joined Formlabs,
a 3D-printing startup, as a member of the electro-optical team.
Benjamin Moss Benjamin Moss is the Director of CMOS Development at
Analog Photonics. After he completed his Ph.D. in Electrical Engineering and
Computer Science in 2014 from MIT, with a thesis focused on circuit designs
for driving high-speed integrated silicon photonic modulators, he joined
Professor Watts Photonic Microsystems Group as a Postdoctoral Associate
and helped develop fabless CMOS-Photonics integration technology. He has
been at Analog Photonics since 2015, and recently, he has been leading the
CMOS chip designs for Analog Photonics LiDAR technology for integrated
electronic beam steering. His interests include energy-efficient circuit design
and system and device modeling focused around microphotonics.
Oleg Shatrovoy Oleg Shatrovoy received his Bachelors degree in 2008 and
Master’s degree in 2016, both in Electrical Engineering and from Boston
University, Boston, MA. He has been working on hardware integration and
system characterization for LiDAR since joining Analog Photonics in March,
2019. His previous position was with MIT Lincoln Laboratory where he
contributed to multiple record-power laser systems, a high-data-rate optical
communications link to a lunar satelite, and an infrared spectroscopy system
for functional brain imaging.
Murshed Khandaker Murshed Khandaker received his Bachelors degree
in Electrical and Electronics Engineering from Bangladesh University of
Engineering and Technology in September 1988 and his Masters in Electronics
Engineering from Philips International Institute in June 1990. He has been
working on digital signal processing firmware for LiDAR since joining Analog
Photonics in May, 2017. Prior to joining Analog Photonics, he has worked on
video processing and display drive electronics in Philips Electronics (1992–
2004) and KOPIN Corporation (2005–2017) .
Michael R. Watts Michael R. Watts is the Chief Executive Officer and
Founder of Analog Photonics. He was also an Associate Professor of Electrical
Engineering and Computer Science with tenure at MIT where he was the
principal investigator of the DARPA E-PHI, DODOS, and VIPER programs.
At MIT he demonstrated state-of-the-art silicon photonic devices such as
electro-optic modulators, optical filters, and large-scale optical phased arrays
on the first demonstrated 300 mm silicon photonics platform worldwide.
Prior to MIT, he was a Principal Member of Technical Staff and Sandia
National Labs where he led their silicon photonics effort. He is also the Chief
Technology Officer of AIM photonics, the $600M public-private partnership to
advance the state of photonic manufacturing in the United States. He has a BS
in Electrical Engineering from Tufts, and MS and PhD degrees in Electrical
Engineering from the Massachusetts Institute of Technology.