Chapter 55.

Optical and Quantum Communications

Optical and Quantum Communications

RLE Group
Optical and Quantum Communications Group
Academic and Research Staff
Professor Jeffrey H. Shapiro, Dr. Franco N. C. Wong, Dr. Lorenzo Maccone, Dr. Ranjith Nair,
Dr. Ral Garca-Patrn, Dr. Julien Le Gout, Dr. Maria Tengner, Dr. Mankei Tsang,
Dr. Valentina Schettini
Graduate Students
Christopher G. Blake, Fabrizio Guerrieri, Nicholas D. Hardy, Bhaskar Mookerji,
Veronika Stelmakh, Dheera Venkatraman, Wenbang Xu, Tian Zhong
The central theme of our programs has been to advance the understanding of optical and
quantum communication, radar, and sensing systems. Broadly speaking, this has entailed: (1)
developing system-analytic models for important propagation, detection, and communication
scenarios; (2) using these models to derive the fundamental limits on system performance; and
(3) identifying, and establishing through experimentation the feasibility of, techniques and devices
which can be used to approach these performance limits.
Sponsors
Air Force Research Laboratory Contract FA8750-09-C-0194
Army Research Office Grant W911NF-05-1-019
NIST - Grant 70NANB7H6186
Office of Naval Research - Contract N00014-08-1-1247
Office of Naval Research - Contract N66001-09-1-2028
U.S. Department of Interior Contract NBCHC00671
W. M. Keck Foundation Center for Extreme Quantum Information Theory
National Science Foundation Grant DGE-0801525
We are embarked on research in the area of quantum information technology whose goal is to
enable the quantum-mechanical information transmission, storage, and processing needed for
future applications in quantum computing and quantum communication. Our theoretical work is
currently focused on the fundamental limits on classical information transmission that are due to
the quantum noise of bosonic channels, and on the use of quantum resources in precision
measurement and imaging applications. Our main experimental work is focused on generation
and application of entanglement sources with high brightness and wavelength tunability. In
addition, we are interested in novel entanglement sources and their applications in quantum logic
gates, enhanced quantum measurements, quantum imaging, quantum protocols for
entanglement distillation, and quantum cryptography.
Fiber-Optic Distribution of Polarization Entangled Photons The capability to efficiently generate
and distribute high-quality entangled photons is key to many applications of photonic quantum
information processing, such as quantum key distribution and linear optics quantum computing.
In recent years most entanglement sources have been based on spontaneous parametric
downconversion (SPDC) in quasi-phase-matched bulk nonlinear crystals such as periodicallypoled potassium titanyl phosphate (PPKTP) or periodically-poled lithium niobate (PPLN) [1]. We
have recently reported a photon-pair source using a PPKTP waveguide that is significantly more
efficient than bulk crystal sources [2]. During the past year, in collaboration with Professor Karl
Berggren, we have used the PPKTP waveguide source to generate high quality polarizationentangled photons and distribute them over a 200-meter fiber-optic link [3]. The polarization
entanglement quality was found to be stable for ~30 min. without active control of the fiber-optic

55-1

Chapter 55. Optical and Quantum Communications

link, suggesting potential applications in short-distance fiber-optic distribution of polarization
entanglement, for instance, in networking between photonic and atomic qubit systems.
SPDC generation efficiency in bulk crystals is typically in the range of 10-12 to 10-8, depending on
the type of crystal, the crystal length, collection angle and bandwidth. Nonlinear waveguides, on
the other hand, have been shown to have a significantly higher SPDC efficiency [2,4]. In our
waveguide work [2] we show theoretically that the enhancement originates from the transverse
index profile of a nonlinear crystal waveguide that imposes an effective transverse momentum on
the phase-matching conditions. The added transverse momentum leads to a broader transverse
spatial bandwidth of the signal and idler outputs, which in turn explains the much higher spectral
brightness of a waveguide SPDC source compared with a bulk-crystal SPDC source.
Figure 1 shows the schematic of our experimental setup for demonstrating polarization
entanglement distribution over a 200-m fiber-optic link between adjacent MIT Buildings 36 and 38
[3]. The 16-mm long fiber-coupled Rb-indiffused PPKTP waveguide was fabricated by AdvR, Inc.
for type-II phase matching with orthogonally-polarized degenerate outputs at 1316 nm. The
signal and idler outputs were separated by a polarizing beam splitter (PBS) and recombined at a
50-50 non-polarizing beam splitter (NPBS) that postselectively generated a pair of polarizationentangled photons in separate single-mode optical fibers. The optical link consisted of two 200-m
single-mode (SMF-28) optical fibers between the source laboratory in Building 36 and the
Berggren laboratory housing a superconducting nanowire single-photon detector (SNSPD) [5]
located in Building 38. See the chapter Quantum Nanostructures and Nanofabrication in this
Report for details of the design, fabrication, and characteristics of SNSPDs. The two entangled
photons were polarization analyzed at the remote location before coincidence detection using a
single SNSPD with a time-multiplexing scheme.

Figure 1. Schematic of experimental setup for polarization entanglement generation in a PPKTP

waveguide, distribution over a 200-m fiber-optic link, and polarization analysis followed by time-multiplexed
two-photon coincidence detection with a single SNSPD at the remote location.

Figure 2 shows the two-photon quantum-interference measurements of the distributed

polarization-entangled photons in the horizontal-vertical (H-V) and antidiagonal-diagonal (A-D)
polarization bases. To avoid degradation due to multiple-pair events, the measurements were
made at a pump power of only 25 W. We obtained visibilities of 98.2% and 97.2% in the H-V
and A-D bases, respectively, without subtraction of accidental coincidences. Compared with a
separate quantum-interference measurement with polarization analysis before the photons were
sent through the fiber link (with visibilities of 98.3% and 97.2%), the fiber-optic transmission of the
entangled photons showed no visibility degradation over the 200-m fiber link.

55-2 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

Figure 2. Two-photon quantum interference of distributed polarization-entangled photons in the H-V basis
(filled diamonds) and the A-D basis (open squares) at 25 W pump power without subtraction of accidentals
[3]. Solid curves are sinusoidal fits.

Polarization entanglement may degrade in fiber propagation owing to depolarization mechanisms

such as temperature fluctuation and mechanical vibration. We investigated the quality of the
distributed entanglement over time by repeatedly measuring the two-photon quantuminterference visibility in both the H-V and A-D bases over a duration of 150 min. Figure 3 shows
the results in which the time origin refers to the starting point when the polarizations were set
correctly and the system was left unattended thereafter. We found that the visibility remained
high at greater than 97% for ~30 min. without active polarization control [3]. This suggests that
high-quality polarization entanglement can be easily distributed over short distances, such as
between laboratories on campus, for an extended period of time if simple active polarization
control is implemented. The compact waveguide source with its high SPDC generation efficiency
and high entanglement quality will play an increasing role in future photonic sources for quantum
information science such as linear-optics quantum computing.

Figure 3. Time evolution of two-photon quantum-interference visibility in the H-V (filled diamonds) and A-D
(open squares) bases for distributed polarization entanglement over two 200-m unattended fibers.

Imaging with Phase-Sensitive Light We have been exploring the use of phase-sensitive light in a
variety of imaging scenarios in both quantum and quantum-mimetic imaging scenarios. A pair of
Gaussian-state light beams that possess a phase-sensitive cross correlation can be produced by
continuous-wave (cw) spontaneous parametric downconversion (SPDC) with vacuum-state signal
and idler inputs [1,6,7]. The low-flux limit of cw SPDC can then be approximated by a vacuum

55-3

Chapter 55. Optical and Quantum Communications

state plus a frequency-entangled biphoton. Many quantum imaging scenarios have been
characterized both theoretically and experimentally in terms of postselected biphoton
detection, e.g., quantum optical coherence tomography [8,9], ghost imaging [10], and two-photon
imaging [11,12]. The primary objective of our work has been to clearly delineate the boundary
between classical and quantum behavior in these and other imaging scenarios and to use this
understanding to develop new, and more robust imaging schemes that offer advantages over
classical techniques. What follows is a brief summary of our ongoing work in ghost imaging.
Ghost imaging is the acquisition of the transmittance pattern of an object through intensity
correlation measurements, and it has been demonstrated with both thermal (classical) light and
biphoton (quantum) light [10,13-15]. We have used our coherence theory [16] for Gaussian-state
sources which encompasses both thermal light and biphoton-state light as special cases to
show that almost all the characteristics of quantum ghost imaging are due to the phase-sensitive
cross correlation between the signal and reference beams [17,18]. The particular ghost-imaging
setup that we considered is shown in Fig. 4. For this arrangement we showed that thermal light,
classical phase-sensitive light, and quantum phase-sensitive light all yield ghost images in both
near-field and far-field operation. The same image inversion that has been seen in the quantum
phase-sensitive light case, but not the thermal light case, turns out to be present for ghost
imaging with classical phase-sensitive light. If the ghost-imagers source fields are constrained to
have specific phase-insensitive auto-correlations, then quantum light offers a spatial resolution
advantage in the sources near field and improved field-of-view in the far field. The principal
advantage of quantum ghost imaging, however, comes from the near-absence of any background
term in the ghost image. We have reported a comprehensive analysis of the signal-to-noise ratio
(SNR) behavior obtained with thermal light, classical phase-sensitive light, and quantum phasesensitive light [19], as well as conceiving two new configurations for ghost imaging [20].

Figure 4. Schematic for transmissive ghost imaging. The signal and reference are broadband light beams
with either a phase-insensitive or phase-sensitive cross correlation. After propagation over an L-m-long free
space path, the signal beam illuminates a scanning pinhole detector and the reference illuminates an object
transmittance mask followed by a large-area (bucket) detector. Cross correlating the resulting (shot-noise
limited) photocurrents as the pinhole detector is scanned yields the ghost image.

Our SNR analysis permits, for the first time, a meaningful comparison between quantum-state
and thermal-state ghost imaging performance with respect to their image acquisition time. i.e., the
integration time required to achieve a desired SNR value for the image. For the important case of
far-field broadband entangled-state imaging versus far-field narrowband thermal-state imaging we
find that neither one enjoys a universal advantage, viz., depending on the parameter values
involved either the quantum or the classical-state system may have the shorter image acquisition
time [19].
The correlation-based theory we have developed for ghost imaging has recently led us to
conceive two novel configurations for ghost imaging: spatial-light modulator (SLM) ghost imaging
and computational ghost imaging [20], as shown in Figs. 5 and 6, respectively. In SLM ghost
imaging we transmit a cw laser beam through a spatial light modulator that imposes an

55-4 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

independent deterministic phase shift on each pixel such that the output field mimics a source of
low spatial coherence. The rest of the setup is the same as a thermal-light lensless ghost imager.
Our analysis shows that this arrangement yields a far-field ghost image with essentially the same
field-of-view and spatial resolution characteristics as found previously for thermal-state ghost
imaging. In particular, the field of view is inversely proportional to the effective coherence length
at the output of the SLM and the spatial resolution in inversely proportional to the beam size at
the output of the SLM.

Figure 5. Configuration for spatial light modulator ghost imaging. The output from a cw laser is passed
through a spatial light modulator driven by deterministic waveforms that impose different phase shifts on
each pixel such that the output field mimics a source of low spatial coherence. The remainder of the setup
is the same as a thermal-light lensless ghost imager.

Figure 6. Configuration for computational ghost imaging. The output from a cw laser is passed through a
spatial light modulator driven by deterministic waveforms that impose different phase shifts on each pixel
such that the output field mimics a source of low spatial coherence. The remainder of the setup is the same
as a thermal-light lensless ghost imager except that the reference path is derived by computing the freespace diffraction integral of the output field obtained from the spatial light modulator.

The transition from SLM ghost imaging to computational ghost imaging arises from the realization
that we can precompute the reference field arriving at the high-resolution detector, in this case,
because it is due to free-space diffraction of the deterministic light field obtained from passing the
cw laser beam through the spatial light modulator. Aside from eliminating the need for a high
spatial-resolution reference-path detector, computational ghost imaging allows the reference field
to be precomputed for a range of path lengths, hence by correlating the bucket detectors output
with these precomputed quantities permits range sectioning to be performed using a single data
collection, something that is not possible in conventional ghost imaging. In this regard it is worth

55-5

Chapter 55. Optical and Quantum Communications

noting that proof-of-principle experiments have already demonstrated the basic features we
predicted for computational ghost imaging [21].
During the past year we have begun a theoretical study of reflective ghost imaging, which, as
shown in Fig. 7, is the configuration that is needed for standoff sensing. Unlike the transmissive
case, ghost imaging done in reflection must cope with the effects of rough-surface scattering from
the target being imaged. So far we have shown [22] that the target-induced speckle created by
this rough-surface scattering does not affect the spatial resolution or image contrast behavior of
the ghost image formed with pseudothermal (classical, phase-insensitive) light, but it does set an
upper limit to the SNR that is not present in the transmissive case.

Figure 7. Configuration for pseudothermal reflective ghost imaging. The output from a cw laser is passed
through a rotating ground-glass diffuser and then split into identical signal and reference beams. The signal
is collected by a CCD camera after L-m-long free-space diffraction. The reference illuminates a roughsurfaced target at range L and the reflected light is measured by a single-pixel (bucket) detector. The ghost
image is formed by cross correlating the photocurrents from the CCD array and the bucket detector.

During the past year we have begun an experimental program to explore ghost imaging. Figure 8
shows the experimental configuration for SLM ghost imaging using two SLMs that allows both
phase-sensitive and phase-insensitive cross correlation to be implemented. The input beam is a
cw laser at 795 nm in a single spatial mode that is split into a reference beam and a signal beam.
The two beams undergo correlated phase-pattern modification imposed by the two spatial light
modulators. For phase-insensitive (-sensitive) cross correlation, the two SLMs have equal-phase
(anti-phase) patterns that are computer generated and updated in real time at a rate of 2 Hz. The
random phase patterns create random speckle patterns for the signal and reference beams in the
far field. The object is placed in the signal path and its transmitted light is collected by a singlepixel bucket detector. A CCD camera is placed in the reference beam path at a distance equal to
the SLM-to-object distance. Intensity correlations between the outputs of the bucket detector and
the CCD camera pixels are averaged and processed to generate the ghost image. The average
light level gives rise to a featureless background that is subtracted to improve the image contrast.

55-6 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

Object

Bucket
detector

Signal beam
SLM1
BS

Input
beam Correlator

SLM2
Reference beam

CCD
camera

Figure 8. Schematic of experimental setup for spatial-light modulator (SLM) ghost imaging using two SLMs.
A beam splitter (BS) splits the input laser at 795 nm into a reference beam and a signal beam, and the two
SLMs impart correlated phase patterns onto each beam. The phase patterns can be equal-phase or antiphase to yield phase-insensitive or phase-sensitive cross correlation, respectively. Intensity correlation
between the bucket detector output and the pixels of the CCD camera yields the ghost image.

We have made preliminary SLM ghost imaging measurements to demonstrate the feasibility of
the two-SLM technique and to show for the first time that ghost imaging can be performed using a
classical light source with phase-sensitive cross correlation. The object was a square that was
placed with an offset relative to the center of the signal beam (the bright spot in Fig. 9). We used
a beam radius of ~200 m at the SLM and we grouped 2 x 2 pixels of the SLM as one superpixel
(30 x 30 m). The distance between the SLM and the object (or the CCD camera) was ~1 meter
and satisfied the far-field requirement for ghost imaging formation for both phase-insensitive and
phase-sensitive cross correlations [17]. Figure 9(a) shows the speckle pattern recorded by the
CCD camera for one of the equal-phase random patterns, showing clearly that the object cannot
be discerned in a single frame. Figure 9(b) shows the ghost image obtained from phaseinsensitive cross correlation between the signal and reference beams after averaging for ~2600
frames of equal-phase phase patterns. The center spot indicating the center of the beam and the
vertical line passing through it seem to be artifacts of the CCD camera imaging optics, as they
also appear in a single frame in Fig. 9(a). For phase-sensitive cross correlation using anti-phase
patterns between the signal and reference beams, Fig. 9(c) shows an inverted image of the
square, as predicted by theory for far-field phase-sensitive ghost image formation [17]. Note that
the two images have similar signal-to-noise ratios and resolutions.

(a)

(b)

(c)

Figure 9. (a) Single frame of speckle pattern recorded by CCD camera, showing no discernable image of
the object. (b) Phase-insensitively cross correlated ghost image of the light-grey square that is placed offset
from the center of the beam. (c) Phase-sensitively cross correlated ghost image of the same square that is
inverted relative to the beam, as predicted by theory for far-field ghost image formation.

Sub-Rayleigh Imaging The Rayleigh diffraction bound sets the minimum separation for two point
objects to be distinguishable in a conventional imaging system. Due to diffraction, the image of a

55-7

Chapter 55. Optical and Quantum Communications

point source through a lens of diameter d and focal length F is an Airy disk with a separation R
between the center to the first zero given by R = 1.22F/d. Feature sizes less than R at the
image plane cannot be resolved. This is the principle dictating that large-diameter telescopes
(larger d) are needed to improve resolution (smaller R).
We have previously proposed an active imaging technique to go beyond the Rayleigh resolution
limit by using focused-beam illumination of an object and employing N-photon detection [23].
Consider a point source with an Airy-disk imaged output whose peak has a maximum average
detected photon number Nmax. The N-photon detection strategy sets a detection threshold N >
Nmax such that a measurement of exactly N photons constitutes a value of 1 for a pixel; otherwise,
the pixel registers a zero value. For N > Nmax the pixels near the center of the Airy distribution are
more likely to register an N-photon event than those at the wings. Assuming a Poisson
distribution of the detected photocounts, the N-photodetection distribution is simply the N-th
power of the Airy distribution, hence sharpening the spatial distribution of the image and locating
the center of the point-source image more accurately.
The sub-Rayleigh resolution obtained for a point source by N-photon detection can be applied to
a spatially-extended object if we illuminate the object point by point and measure the
corresponding N-photocount output. The point-like illumination can be realized using a focused
laser beam with its beam diameter at the object defining the size of the point and hence placing
a lower bound on the ultimate resolution of this technique. It should be noted that the usual fullobject illumination cannot lead to sub-Rayleigh resolution even if N-photon detection is utilized.
We have recently demonstrated this sub-Rayleigh imaging technique in collaboration with Prof.
Franco Zappa and visiting graduate student Fabrizio Guerrieri of Politecnico di Milano and Dr.
Simone Tisa of Micro Photon Devices. Our collaborators provided the crucial 32 x 32-pixel
single-photon counting array to enable N-photon imaging measurements [24]. Part of a U.S. Air
Force resolution target mask was used as the object for imaging, as indicated by the arrow in Fig.
10(a). The imaging apparatus had a small aperture (~2 mm diameter) that served to impose the
Rayleigh limit on the image of three pairs of alternately clear and opaque stripes, each of 660 m
width at the image plane. Full-object illumination yielded an image in which the stripes cannot be
resolved due to the Rayleigh diffraction limit, estimated to be 1.86 mm, as shown in Fig. 10(b).
By focusing the illuminating laser at 532 nm to a 20-m radius at the object and scanning the
beam spot in an arbitrary fashion to cover the spatial extent of the object, we obtained the Nphotocount image in Fig. 10(c), showing clearly the three stripes of the mask. The resolution
improvement is in good agreement with the theoretical prediction of (N-Nmax)1/2 = 3, for N = 23
and Nmax = 14. In Fig. 10(c) a few of the pixels had a high number of N-photon events and
obscured the clarity of the image. By capping the event occurrence to a maximum of 800 (out of
8000 frames per illumination spot) to make the lower-count pixels more visible, we obtained the
3D intensity profile in Fig. 10(d) that reveals the three stripes very clearly. This new imaging
method uses a classical light source, tight focusing on the object, and N-photon detection to yield
sub-Rayleigh resolution. It may find useful applications in imaging situations in which precise
raster scanning is not possible.

55-8 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

32

100

3
4
5
6

Distance (x 100 m)

4
5
6

24

40
8

20
0
0

8
16
24
Distance (x 100
m)

x 10

3.0
24

2.5
2.0

16

1.5
1.0

N-photon
events
1000
500
0
5
Pi

0.5

10
15
20
25
30

l
xe

m
nu

(c)

be

8
16
24

Distance (x 100 m)

32

sub-Rayleigh bound
stripe size
Rayleigh bound

32

stripe size
Rayleigh bound

(b)

32

Distance (x 100 m)

60

16

(a)

80

25
20
15 umber
10 ixel n
P

30

(d)

Figure 10. (a) Red arrow indicates part of an US Air Force resolution target to be imaged. (b) Blurred image
obtained conventionally using full illumination. (c) Sub-Rayleigh image using focused illumination and N =
23. (d) 3D intensity profile of (c) with the stripes clearly revealed by clipping event counts at 800.

Quantum Illumination Loss and noise can quickly destroy entanglement, so it has commonly
been thought that there is little reason to employ entangled light sources in such scenarios. Lloyd
[25], however, showed that quantum illumination can reap substantial benefits, from the use of
entanglement in target detection, despite the presence of loss-destroying loss and noise. In
Lloyds quantum-illumination paradigm, a photonic source creates d-mode maximally entangled
signal and ancilla beams each containing a single photon. The signal beam irradiates a target
region containing a very weak thermal-noise bath with an average of b << 1 photons per mode
in which a low-reflectivity object might be embedded. The light received from this region
together with the retained ancilla beam is then used to decide whether the object is present or
absent. Lloyd showed that quantum illumination, with the optimum joint measurement on the
received light and the ancilla, achieves a much higher signal-to-background ratio than that
realized by optimum quantum reception of light received in response to transmission of a single
unentangled photon.
The analysis in [25] was confined, for the most part, to the vacuum plus single-photon manifold,
wherein at most one photon arrives at the receiver during the measurement interval regardless of
whether the object of interest is absent or present in the target region. We remedied that

55-9

Chapter 55. Optical and Quantum Communications

deficiency by providing a full Gaussian-state treatment of quantum-illumination target detection
[26] by employing the exact quantum statistical model for the entangled signal and idler beams
obtained from cw SPDC in the absence of pump depletion [1] in conjunction with the standard
model for the lossy, bosonic channel [27]. We showed that in a very lossy, very noisy
environment, a low-brightness quantum-illumination system enjoys a substantial improvement in
the effective signal-to-background ratio which translates into a very large reduction in the
target-detection error probability in comparison to that achieved by a coherent-state transmitter
of the same average photon number. Just as Lloyd found in [25], the SPDC quantum illumination
advantage that we have derived accrues despite there being no entanglement between the light
that is received from the target region and the retained idler. Quantum illumination is thus the first
example of an entanglement-based performance gain, in a full bosonic-channel setting, that
survives entanglement-killing loss and noise.
During the past year, we extended our work on quantum illumination in two significant ways.
First, we showed that quantum illumination offers a performance advantage in one-versus-two
point-target resolution [28] that it similar to what it provides in one-versus-none target detection.
Figure 11(a) shows the one-dimensional geometry that we considered. The problem is to
determine the minimum angular separation at which we can make a reliable 0.03 error
probability decision as to whether one or two point objects, which reflect equal power back to
the receiver, are present. Figure 11(b) shows that the use of a cw-SPDC quantum illumination
system outperforms a coherent-state (laser) system of the same average transmitted power.

Figure 11. (a) One-dimensional geometry for one-versus-two target resolution problem. Under hypothesis
H1 there is a single on-axis specular point target. Under hypothesis H2 there are two, identical, in-phase
specular point targets symmetrically disposed at angle about the axis. A quantum-illumination receiver is
employed with a diameter-D entrance pupil. (b) Performance of the quantum-illumination system (red)
versus a coherent-state system of the same average transmitted power and wavelength (blue), showing the
normalized angle relative to the diffraction limit /D that each can resolve. NS is the average signal
photon-number per mode that is transmitted; NB is the average background photon-number per mode that is
received; M is the number of temporal modes employed; and is the roundtrip transmissivity.

Quantum Illumination-based Secure Communication Our most recent theoretical work on SPDC
quantum illumination has been the proposal of a novel two-way secure optical communication
protocol that is immune to passive eavesdropping [29]. The basic setup for our analysis is shown
in Fig. 12. Alice generates multi-temporal mode entangled signal and idler light beams using a
cw SPDC source. She sends the signal to Bob over a lossy channel while retaining the idler.
Each T-sec-long transmission (one bit) from Alice comprises M = WT >> 1 signal-idler mode
pairs, where W is the bandwidth of the signal and idler fields. Each of these modes has a mean
photon number NS = NI << 1. Bob encodes the desired information by modulating the received
signal phase using binary phase-shift keying (BPSK). He then amplifies the signal with a phaseinsensitive optical amplifier to compensate for loss and add a significant amount of noise
before sending it back to Alice, again over a lossy channel. Alice makes a joint measurement on

55-10 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

the returned signal (plus noise) and the retained idler to extract Bob's information. We assume
that eavesdropper Eve obtains all the photons lost en route from Alice to Bob and from Bob to
Alice, and that Eve has access to an optimal quantum receiver while Alice only has access to a
receiver we know how to build, viz., the optical parametric amplifier (OPA)-based receiver
described in [30]. Alice then enjoys several orders of magnitude better error probability than Eve,
as seen in Fig. 13, which plots upper and lower bounds on the error probability of Eves optimum
quantum receiver and upper bounds on the error probability of Alices optimum quantum receiver
whose implementation is unknown and her OPA receiver. Alice's performance advantage
relative to Eve originates from the stronger-than-classical phase-sensitive cross correlation
between the signal and idler created by the SPDC source, which gives her enhanced sensitivity
despite the fact that the entanglement between the noisy returned signal and the idler has been
destroyed by loss and noise. From Fig. 13 we have that with W = 1 THz and T = 20 ns we can
get 50 Mbit/s communication over 50 km of 0.1 dB/km loss fiber (when other losses can be
neglected) with Alices OPA receiver having an error probability of less than 5.1 x 10-7 and Eves
optimum quantum receiver having an error probability bounded between 0.28 and 0.46.

Figure 12. Two-way communication protocol using quantum illumination that is immune to passive
eavesdropping.

Figure 13. Theoretical performance error probability bounds versus number of modes employed for the
two-way communication protocol using quantum illumination. The parameters assumed are: NS = 0.004;
= 0.1; and G = NB = 104.

55-11

Chapter 55. Optical and Quantum Communications

We are in the process of performing a proof-of-principle experiment of this secure communication
protocol. In our fiber-based implementation, shown in Fig. 14, the broadband output from a cwpumped periodically-poled MgO-doped lithium niobate (PP-MgO:LN) crystal SPDC source was
coupled into a single-mode fiber and separated by a coarse wavelength division multiplexer
(CWDM) into signal and idler beams centered at 1550 and 1570 nm, respectively. For ~100 mW
of pump we measured 180 pW of signal at the CWDM output (bandwidth of ~16 nm). After signal
amplification of 40 dB in an erbium-doped fiber amplifier (EDFA) we estimated a signal power of
200 nW plus 1.6 mW of noise at its output. This very significant amount of noise frustrates wouldbe eavesdroppers who gain very little information without the conjugate idler. Our initial
measurements were to test the weakly-pumped OPA receiver in its ability to extract the weak
encoded signal. Figure 15 shows the OPA-receiver signal for a square-wave (0) input phase
modulation at 20 kHz in the (a) time domain and (b) frequency domain. The measured signal
strength was ~10x smaller than expected (100x less in signal-to-noise ratio), which was caused
by dispersion of the 0.25-ps (16-nm bandwidth) signal pulse through ~70-m of standard fiber and
a smaller amount of the idler pulse through a combination of standard and low-dispersion fibers in
the idler arm. We verified that no OPA-receiver signal was measurable when the idler light (~100
pW) was blocked, clearly suggesting that the error probability would be high (~0.5) without the
idler field. Work is ongoing to: compensate the dispersion with a pair of gratings and hence
better recover the dispersion-impaired signal; implement standard communication protocol; and
construct an Eve with which to verify the bit error rate disparity between Alice and Eve.

Figure 14. Experimental setup for quantum illumination-based secure communication. The eavesdropping
channel, schematically denoted by Eve, is not implemented in the initial experiment.

Figure 15. Measured OPA-receiver signal in the (a) time domain, and (b) frequency domain, for a 0 input
square-wave phase modulation at 20 kHz.

Quantum-Enhanced Laser Radar Operation A key feature of a remote sensing system is its
ability to obtain detailed spatial information about targets of interest, in both transverse and
longitudinal (range) dimensions. High-resolution spatial information is essential for such tasks as
target classification, image processing, and tracking of multiple closely-spaced targets. For

55-12 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

modes-range (1100 km) terrestrial applications under clear-weather conditions laser radar
systems offer superior spatial resolution when compared to microwave radars, owing to their use
of much shorter wavelengths. When atmospheric turbulence can be neglected, the spatial
resolution of such a system is generally limited by the Rayleigh resolution of its receiving optics
and the signal-to-noise ratio. We have analyzed two ways in which quantum effects can be used
to improve the spatial resolution of a laser radar system that uses conventional, floodlight laser
illumination and a soft-aperture entrance pupil in its receiver [31,32], see Fig. 16. Squeezedvacuum injection (SVI), as proposed in [33], reduces the vacuum noise incurred on the highspatial-frequency target information that has been attenuated by the soft aperture. SVI requires
the use of homodyne detection, so the laser radar receiver is only sensitive to the quadrature in
which the noise reduction has occurred. The effectiveness of this noise reduction is, however,
severely restricted by inefficiency in that homodyne measurement. Thus, we proposed the use of
phase-sensitive amplification (PSA) after the SVI stage and before homodyne detection. PSA
enables noise-free amplification of a single field quadrature and hence allows any homodyne
inefficiency to be overcome.

Figure 16. Diagram of quantum-enhanced laser radar receiver. Point targets (one or two) at range L are
shown at the left. Baseband field operators are shown for the received field and the squeezed-vacuum
injected field impinging from the left and the right, respectively, on a soft-aperture pupil function. The
combination of these two fields propagated to an image plane undergo phase-sensitive amplification
and homodyne detection on a continuum array.

Figure 17 shows simulated intensity images for our quantum-enhanced laser radar receiver when
the planar target is the US Air Force resolution chart shown in (a) that gives rise to fullydeveloped speckle. We have assumed a target range L = 1 km, a 15 m x 15 m square target
region, 1550 nm laser wavelength, a 4-mm-waist Gaussian soft-aperture pupil inside an 8-mmdiameter hard aperture imaged onto a continuum homodyne-detection array. Figure 17(b) shows
the image of the resolution chart after blurring by transmission through the soft aperture. This
corresponds to the image in the limit of high SNR and averaging a large number of intensity
images with independent speckle. The images in Fig. 17(c)-(f) show detected images averaging
over 100 intensity images with independent speckle fluctuations and 25% homodyne efficiency.
Figure 17(c) shows the baseline image, i.e., no SVI and no PSA. Figure 17(d) shows the result of
adding SVI enhancement with 15 dB of quadrature-noise squeezing to the baseline receiver.
Here we see that the low homodyne efficiency has rendered the SVI ineffective. Figure 17(e)
shows the result of adding 15 dB of PSA gain to the baseline configuration with no SVI. In this
case there is some improvement in the resolution. Figure 17(f) shows the combined value of 15
dB of SVI plus 15 dB of PSA. This figure shows a substantial improvement in image quality over
the baseline, SVI-only, and PSA-only images.
In continuing work, we have been relaxing a number of idealizations that were made in [33,34].
Specifically, we have replaced the continuum homodyne array with an array comprised of a finite
number of discrete detectors, and we have been working to incorporate more realistic modedecomposition models for the nonlinear optical devices used in the SVI and PSA. So far, our
results still indicate that SVI and PSA continue to offer spatial-resolution performance advantages
when added to a baseline soft-aperture, homodyne-detection laser radar receiver.

55-13

Chapter 55. Optical and Quantum Communications

Figure 17. Computer simulated laser radar images. (a) US Air Force resolution chart used as the target.
(b) Soft-aperture blurred version of the US Air Force target, i.e., the high SNR image in the absence of
speckle. (c) Baseline image of the US Air Force resolution chart obtained with no quantum enhancements.
(d) The image obtained when 15 dB of SVI is added to the baseline configuration. (e) The image obtained
when 15 dB of PSA is added to the baseline configuration. (f) The image obtained when 15 dB of SVI and
15 dB of PSA is added to the baseline configuration. In (c)-(f) the displayed image is the result of averaging
100 frames with independent speckle and the homodyne array is 25% efficient.

References
1. F. N. C. Wong, J. H. Shapiro, and T. Kim, Efficient Generation of Polarization-Entangled
Photons in a Nonlinear Crystal, Laser Phys. 16, 1517-1524 (2006).
2. T. Zhong, F. N. C. Wong, T. D. Roberts, and P. Battle, "High Performance Photon-Pair
Source based on a Fiber-coupled Periodically Poled KTiOPO4 Waveguide," Opt. Express 17,
12019-12030 (2009).
3. T. Zhong, X. Hu, F. N. C. Wong, K. K. Berggren, T. D. Roberts, and P. Battle, "High-Quality
Fiber-Optic Polarization Entanglement Distribution at 1.3 m Telecom Wavelength," Opt. Lett.
35, 1392-1394 (2010).
4. M. Fiorentino, S. M. Spillane, R. G. Beausoleil, T. D. Roberts, P. Battle, and M. W. Munro,
Spontaneous Parametric Down-Conversion in Periodically Poled KTP Waveguides and Bulk
Crystals, Opt. Express 15, 7479-7488 (2007).
5. X. Hu, T. Zhong, J. E. White, E. A. Dauler, F. Najafi, C. H. Herder, F. N. C. Wong, and K. K.
Berggren, Fiber-Coupled Nanowire Photon Counter at 1550 nm with 24% System Detection
Efficiency, Opt. Lett. 34, 3607-3609 (2009).
6. J. H. Shapiro and N. C. Wong, An Ultrabright Narrowband Source of Polarization-Entangled
Photon Pairs, J. Opt. B: Quantum Semiclass. Opt. 2, L1-L4 (2000).

55-14 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

7. J. H. Shapiro and K.-X. Sun, Semiclassical versus Quantum Behavior in Fourth-Order
Interference, J. Opt. Soc. Am. B 11, 1130-1141 (1994).
8. A. F. Abouraddy, M. B. Nasr, B. E. A. Saleh, A. V. Sergienko, and M. C. Teich, Quantum
Optical Coherence Tomography with Dispersion Cancellation, Phys. Rev. A 65 053817
(2002).
9. M. B. Nasr, B. E. A. Saleh, A. V. Sergienko, and M. C. Teich, Demonstration of DispersionCanceled Quantum-Optical Coherence Tomography, Phys. Rev. Lett. 91, 083601 (2003).
10. T. B. Pittman, Y. H. Shih, D. V. Strekalov, and A. V. Sergienko, Optical Imaging by Means of
Two-Photon Quantum Entanglement, Phys. Rev. A 52, R3429-R3432 (1995).
11. M. DAngelo, M. V. Chekhova, and Y. Shih, ``Two-Photon Diffraction and Quantum
Lithography, Phys. Rev. Lett. 87, 013602 (2001).
12. Y. Shih, Quantum Imaging, IEEE J. Sel. Top. Quantum Electron. 13, 1016-1030 (2007).
13. F. Ferri, D. Magatti, A Gatti, M. Bache, E. Brambilla, and L. A. Lugiato, High Resolution
Ghost Image and Ghost Diffraction Experiments with Thermal Light, Phys. Rev. Lett. 94,
183602 (2005).
14. G. Scarcelli, V. Berardi, and Y. Shih, Can Two-Photon Correlation of Chaotic Light Be
Considered as Correlation of Intensity Fluctuations?, Phys. Rev. Lett. 96, 063602 (2006).
15. R. Meyers, K. Deacon, and Y. Shih, Ghost-imaging Experiment by Measuring Reflected
Photons, Phys. Rev. A 77, 041801(R) (2008).
16. B. I. Erkmen and J. H. Shapiro, Optical Coherence Theory for Phase-Sensitive Light, Proc.
SPIE 6305, 6305G (2006).
17. B. I. Erkmen and J. H. Shapiro, Unified Theory of Ghost Imaging with Gaussian-State Light,
Phys. Rev. A 77, 043809 (2008).
18. B. I. Erkmen and J. H. Shapiro, Gaussian-State Theory of Two-Photon Imaging, Phys. Rev.
A 78, 023835 (2008).
19. B. I. Erkmen and J. H. Shapiro, Signal-to-Noise Ratio of Gaussian-State Ghost Imaging,
Phys. Rev. A 79 023833 (2009).
20. J. H. Shapiro, Computational Ghost Imaging, Phys. Rev. A 78, 061802(R) (2008).
21. Y Bromberg, O. Katz, and Y. Silberberg, Ghost Imaging with a Single Detector, Phys. Rev.
A 79, 053840 (2009).
22. N. D. Hardy and J. H. Shapiro, Ghost Imaging in Reflection: Resolution, Contrast, and
Signal-to-Noise Ratio, Proc. SPIE 7815, 78150P (2010).
23. V. Giovannetti, S. Lloyd, L. Maccone, and J. H. Shapiro, Sub-Rayleigh-Diffraction-Bound
Quantum Imaging, Phys. Rev. A 79, 013827 (2009).
24. F. Guerrieri, L. Maccone, F. N. C. Wong, J. H. Shapiro, S. Tisa, and F. Zappa, Sub-Rayleigh
Imaging via N-Photon Detection, Phys. Rev. Lett. 105, 163602 (2010).
25. S. Lloyd, Enhanced Sensitivity of Photodetection via Quantum Illumination, Science 312,
1463-1465 (2008).

55-15

Chapter 55. Optical and Quantum Communications

26. S.-H. Tan, B. I. Erkmen, V. Giovannetti, S. Guha, S. Lloyd, L. Maccone, S. Pirandola, and J.
H. Shapiro, Quantum Illumination with Gaussian States, Phys. Rev. Lett. 101, 253601
(2008).
27. V. Giovannetti, S. Guha, S. Lloyd, L. Maccone, J. H. Shapiro, B. J. Yen, and H. P. Yuen,
Classical Capacity of Free-Space Optical Communication, in Quantum Information,
Statistics, Probability ed. O. Hirota (Rinton Press, New Jersey, 2004).
28. J. H. Shapiro and S. Guha, Enhanced Standoff Optical Sensing Resolution Using Quantum
Illumination, Single Photon Workshop 2009, Boulder, CO, November 3-6, 2009.
29. J. H. Shapiro, Defeating Passive Eavesdropping with Quantum Illumination, Phys. Rev. A
80, 022320 (2009).
30. S. Guha and B. I. Erkmen, Quantum-Illumination Receivers for Target Detection, Phys.
Rev. A 80, 052310 (2009).
31. Z. Dutton, J. H. Shapiro, and S. Guha, LADAR Resolution Improvement using Receivers
Enhanced with Squeezed-Vacuum Injection and Phase-Sensitive Amplification, J. Opt. Soc.
Am. B 27, A63 (2010).
32. Z. Dutton, J. H. Shapiro, and S. Guha, LADAR Resolution Improvement using Receivers
Enhanced with Squeezed-Vacuum Injection and Phase-Sensitive Amplification: Erratum, J.
Opt. Soc. Am. B 27, 2007 (2010).
33. P, Kumar, V. Grigoryan, and M. Vasilyev, Noise-Free Amplification: Towards Quantum
Laser Radar, 14th Coherent Laser Radar Conference, Snowmass CO, July 9-13 (2007).

Publications
Journal Articles, Published
X. Hu, T. Zhong, J. E. White, E. A. Dauler, F. Najafi, C. H. Herder, F. N. C. Wong, and K. K.
Berggren, Fiber-Coupled Nanowire Photon Counter at 1550 nm with 24% System Detection
Efficiency, Opt. Lett. 34, 3607-3609 (2009).
J. H. Shapiro and S. Lloyd, Quantum Illumination versus Coherent-State Target Detection, New
J. Phys. 11, 063045 (2009).
J. H. Shapiro, The Quantum Theory of Optical Communications, IEEE J. Sel. Topics Quantum
Electron. 15, 1547-1569 (2009).
J. H. Shapiro, Corrections to The Quantum Theory of Optical Communications, IEEE J. Sel.
Topics Quantum Electron. 16, 698 (2010).
J. H. Shapiro, Defeating Passive Eavesdropping with Quantum Illumination, Phys. Rev. A 80,
022320 (2009).
Z. Dutton, J. H. Shapiro, and S. Guha, LADAR Resolution Improvement using Receivers
Enhanced with Squeezed-Vacuum Injection and Phase-Sensitive Amplification, J. Opt. Soc. Am.
B 27, A63 (2010).
Z. Dutton, J. H. Shapiro, and S. Guha, LADAR Resolution Improvement using Receivers
Enhanced with Squeezed-Vacuum Injection and Phase-Sensitive Amplification: Erratum, J. Opt.
Soc. Am. B 27, 2007 (2010).

55-16 RLE Progress Report 152

Chapter 55. Optical and Quantum Communications

J. Le Gout, D. Venkatraman, F. N. C. Wong, and J. H. Shapiro, Experimental Realization of

Phase-Conjugate Optical Coherence Tomography, Opt. Lett. 35, 1001-1003 (2010).
J. H. Shapiro, Dispersion Cancellation with Phase-Sensitive Gaussian-State Light, Phys. Rev. A
81, 023824 (2010).
T. Zhong, X. Hu, F. N. C. Wong, K. K. Berggren, T. D. Roberts, and P. Battle, High-Quality FiberOptic Polarization Entanglement Distribution at 1.3 m Telecom Wavelength, Opt. Lett. 35,
1392-1394 (2010).
F. Guerrieri, L. Maccone, F. N. C. Wong, J. H. Shapiro, S. Tisa, and F. Zappa, Sub-Rayleigh
Imaging via N-Photon Detection, Phys. Rev. Lett. 105, 163602 (2010).
Meeting Papers, Published
J. H. Shapiro, Defeating Passive Eavesdropping with Quantum Illumination, Frontiers in Optics
Meeting, Rochester, NY, October 11-15, 2009.
X. Hu, T. Zhong, F. Najafi, F. Marcili, C. Herder, F. N. C. Wong, E. A. Dauler, and K. K. Berggren,
Efficiently Coupling Light to Superconducting Nanowire Single-Photon Detectors, Single Photon
Workshop 2009, Boulder, CO, November 3-6, 2009.
F. N. C. Wong, Generation of Single Spatiotemporal Mode Photons, Single Photon Workshop
2009, Boulder, CO, November 3-6, 2009.
T. Zhong, X. Hu, F. N. C. Wong, C. Herder, F. Najafi, K. K. Berggren, T. D. Roberts, and P.
Battle, High Quality Photonic Polarization Entanglement Distribution at 1.3-m Telecom
Wavelength, Single Photon Workshop 2009, Boulder, CO, November 3-6, 2009.
J. H. Shapiro and S. Guha, Enhanced Standoff Optical Sensing Resolution Using Quantum
Illumination, Single Photon Workshop 2009, Boulder, CO, November 3-6, 2009.
J. Le Gout, D. Venkatraman, F. N. C. Wong, and J. H. Shapiro, Experimental Realization of
Phase-Conjugate Optical Coherence Tomography, 18th International Laser Physics Workshop,
Barcelona, Spain, July 13-17, 2009.
J. H. Shapiro, Z. Dutton, and S. Guha, Improving Laser Radar Spatial Resolution by SqueezedVacuum Injection, 18th International Laser Physics Workshop, Barcelona, Spain, July 13-17,
2009.
R. Garca-Patrn, F. N. C. Wong, and J. H. Shapiro, Optimal Individual Attack on BB84 Quantum
Key Distribution using Single-Photon Two-Qubit Quantum Logic, SPIE Proceedings on Quantum
Information and Computation VIII, Orlando, FL, April 5-9, 2010.
J. H. Shapiro, Dispersion Cancellation with Phase-Sensitive Gaussian-State Light, Digest of
Quantum Electronics and Laser Science Conference, San Jose, CA, May 16-21, 2010.
F. Guerrieri, L. Maccone, F. N. C. Wong, J. H. Shapiro, S. Tisa, and F. Zappa, Sub-Rayleigh
Imaging via N-Photon Detection, Digest of Quantum Electronics and Laser Science Conference,
San Jose, CA, May 16-21, 2010.
A. H. Nejadmalayeri, J. A. Cox, F. N. C. Wong, T. D. Roberts, P. Battle, and Franz X. Krtner,
Phase Noise Measurement of Mode Locked Lasers using Guided Wave PPKTP Balanced Cross

55-17

Chapter 55. Optical and Quantum Communications

Correlators, Digest of Conference on Lasers and Electro-Optics, San Jose, CA, May 16-21,
2010.
N. D. Hardy and J. H. Shapiro, Ghost Imaging in Reflection: Resolution, Contrast, and Signal-toNoise Ratio, SPIE Proceedings on Quantum Communications and Quantum Imaging VIII, San
Diego, CA, August 4-5, 2010.
Theses
B. Mookerji, Generation of Fiber-Coupled, Nondegenerate, Polarization-Entangled Photons for
Quantum Communication, S.B. thesis, Department of Physics, MIT 2009.
T. Zhong, High Performance Photon-Pair Source based on a Fiber-Coupled Periodically Poled
KTiOPO4 Waveguide, S.M. thesis, Department of Electrical Engineering and Computer Science,
MIT 2009.

55-18 RLE Progress Report 152