A Coincidence Counting System for Twelve-photon Entanglement

Experiment

Yi Hu

1,2

, Wei Li

1,2

, Yue-fei Wang

1,2

, Ge Jin

and Xiao Jiang

1,2

Hefei National Laboratory for Physical Sciences at Microscale and Department of Modern Physics,

University of Science and Technology of China, Hefei 230026, China

CAS Center for Excellence and Synergetic Innovation Center in Quantum Information and Quantum Physics,

University of Science and Technology of China, Hefei 230026, China

Keywords:

Multi-photon, Coincidence Counting, Scalability.

Abstract:

Multi-photon entanglement is an important resource for photonic quantum information, and its scale has

reached 6 photons with 18 qubits or 10 photons with 10 qubits. The upcoming challenge will be 12 pho-

tons with 12 qubits. In the entanglement experiments of such a plurality of photons, the coincidence counter

has always been an important tool, and the experiment of 12 photons poses new requirements. Here we report

the upgrading of a coincident counting system that worked well in 6-photon and 10-photon experiments to the

coming 12-photon one. The scalability of the coincident counting system has been shown. By optimizing the

logic in the Field Programmable Gate Array(FPGA) and the LabVIEW program, not only the number of input

channels has been increased for 12 photons, but also the functions of signal alignment and status monitoring

have been improved. The coincidence result can be analysed both in real-time and off-line. The system is

capable to extend to 104 channels at most for channel consuming application.

1 INTRODUCTION

Quantum information science makes great progress

in recent decades, where quantum entangle-

ment(Einstein et al., 1935; Schr

odinger, 1935)

always lies at the heart. At the most fundamental

level, quantum entanglement represents the intrinsic

non-locality(Brunner et al., 2014) and thus serves

to clarifying essential understandings of quantum

physics. On the other hand, quantum entanglement

is the resource of quantum computation(Knill et al.,

2001; Kok et al., 2007). For these reasons, generating

and testing the quantum entanglement of a large

quantity of qubits in experiments becomes one of the

main target of the quantum information science.

To study the quantum entanglement, the system

of linear optics constitutes one nice platform due

to the well-developed techniques to control photonic

qubits. However, the non-deterministic generation

of entangled photon pairs prevents the scaling-up of

the system. In order to overcome this bottleneck, ef-

forts are made to improve the entangled photon pair

source, which is normally from an ultrafast pulsed

laser pumped spontaneous parametric downconver-

sion (SPDC)(Pan et al., 2012; Kwon et al., 2008; Kim,

2003). Recently, with the progresses of the SPDC,

the entanglement of 10 qubits with 10 photons(Wang

et al., 2016; Chen et al., 2017) and 18 qubits with 6

photons (Wang et al., 2018; Wang et al., 2015) are

realized. What’s more, as an optimal SPDC photon

source of 97% heralding efﬁciency and 96% inde-

pendent single photons’ indistinguishability has been

achieved(Zhong et al., 2018), it is hopeful to increase

the number of entangled photons continuously. Now

the next landmark is the 12-photon entangled state. In

this context, for the detection of entanglement of more

photons, the correlated measurement brings new chal-

lenges for the coincidence counting system.

The coincidence counting system usually cor-

responds to a type of instrument that records

the time correlation between single-photon de-

tectors(SPDs)(Hadﬁeld, 2009; Farr, 2012) within

a small time window. Historically, an general

method of coincidence counting called time-to-

amplitude(TAC)(Crotti et al., 2012; Simms, 1961) is

proposed and implemented, in which the arrival times

of different photons are converted to the pulse ampli-

tudes. Though having a high time resolution on the

order of picoseconds, this method consumes a large

amount of resource. An alternative way is to initi-

ate the pattern sampling(Gaertner et al., 2005) into

a latch with trigger signals generated by a electronic

Hu, Y., Li, W., Wang, Y., Jin, G. and Jiang, X.

A Coincidence Counting System for Twelve-photon Entanglement Experiment.

DOI: 10.5220/0007371300390045

In Proceedings of the 7th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2019), pages 39-45

ISBN: 978-989-758-364-3

(a)

Optical Part SPD Part Coincidence counting Part

Pulsed Laser

optics

for

the

entangle-

ment

Fiber

Detector Pre-amplify

I/O boards

pulse

discriminator

Backplane signal transmission

Main board

coincidence

counting

Data processing

Figure 1: Experimental setup. The pulsed laser generates at most 12 entangled photons. The two states of each photon

could be distingushed with two SPDs. The I/O boards collect and discriminate the input signals. I/O boards are inserted in

Backplane to transfer signals to the Main board. The Main board implement coincidence logic and complete statistic analysis

on personal computer.

logic OR gate. This method suffers from an low event

rate of 0.8MHz. Now there are also some commercial

solutions for Time-Corrected Single Photon Counting

(TCSPC)(Wahl et al., 2013) . Nevertheless, the mea-

surement range is restricted in 5ns and can only be

used with double channels. Therefore a home-made

coincidence counting system should be designed for

multi-photon entanglement experiments.

Previously, a coincidence counting system with

48-channel input signals(Zhang et al., 2016) is estab-

lished for 6-photon entanglement experiment. The

system utilizes the internal delay line of Field Pro-

grammable Gate Array (FPGA) to align the input sig-

nals and compress the pulses by logic constrains. Af-

ter these pre-processing of input signals, the clock

phase is shifted to sample signals and the results are

stored in the external random access memory. The

system following the same structure shows the ﬁne

scalability and reconﬁgurability, which supports 104

channels and 1 Gbit data size at most. While in the 12-

photon entanglement experiment, the channel num-

ber and the data volume are both within the maxi-

mal range. So it is practical to upgrade and remodule

the original system to service the new experiment and

meet new requirements.

In this letter, we present a coincidence counting

system with 24 input channels for 12-photon entan-

glement experiment on the basis of the coincidence

counting scheme for 6 photons. This successful up-

grading conﬁrms the ﬁne scalability of the 48-channel

coincidence counting system in the 6-photon entan-

glement experiment.

2 REALIZATION

The holistic structure of the 12-photon entanglement

experiment is shown in Figure 1. The whole ex-

periment can be divided into three parts: the opti-

cal Part, the single photon detector(SPD) part and

the coincidence counting part. In the optical part, a

76MHz laser produces pulses of photons, which go

through the system of linear optics for generating the

12-photon entanglement. Later for each photon its

two states can be distinguished with two SPDs, thus

24 SPDs are needed to record those 16,777,216 (2

)

sorts of coincidence events. These detectors sample

the optical pulses and pre-amplify input signals. Fi-

nally the coincidence counting part relies on a suit of

hardware which consisted of six I/O boards, one Main

board and one Backplane, shown in Figure 2.

2.1 The Hardware Platform

The I/O board is used to capture and transfer the

input signals from detectors. One I/O board has 8

identical channels composed of an I/O port, an am-

plitude, a comparator and a digital potentiometer.

They work together as an ﬂexibly adjustable thresh-

old discriminator uniting the conﬁgurable FPGA(

EP4CE6E22CSN) for adapting the amplitudes of in-

put signals. To reduce crosstalk between channels, the

input and output channels in I/O boards are arranged

in a staggered conﬁguration: the adjacent channels of

the inputs are conﬁgured as output GNDs. So each

I/O board supports 4 channels of input signals and 6

pieces of I/O board are required for the 24 channels

of the 12-photon entanglement.

In the system, the Main board plays a central

role of performing coincidence logical operations,

PHOTOPTICS 2019 - 7th International Conference on Photonics, Optics and Laser Technology

Peripherals

1Gbit

DDR

128

Mbit

Flash

16Mbit

SPI

Flash

Power

User

LED

JTAG

Xilinx FPGA

Spartan 6

XC6SLX16CSG324-2

USB 2. 0

CYPR ESS FX2

64 kb

I2C

EEPROM

USB

Connec to r

24M

Oscill at or

48M

Oscill at or

160 pin

signal

female

connector

Peripherals

USER

LEDs

USB

2.0

Toggle

Switch

8 channe ls

Altera FPGA

EP4CE6E22C8N

JTAG

100M

Oscill at or

Power

Clock si gn al

100 MHz

96 pin

signal

female

connector

Lemo

connec to r

Ampliﬁ er

compar at or

potent io -

meter

(a) I/O board

(b) Main board

8 channe ls

24 chann el s

Clock si gn al

76 MHz

Powe r

Main

I/O

I/O I /O

I/O

I/O I /O

I/OI/OI/O

I/O

Slot0 Slot1 Slot2 Slo t3 S lo t4 Slot5 Slot6 S lo t7 S lo t8Slot13Slot12Slo t11S lo t1 0

I/O

Slot9

Figure 2: (a) The framework of I/O board. (b) The frame-

work of Main board. (c) The framework of Backplane.

system control and data processing. It contains a

low-end spartan6 (XC6SLX16CSG324C) FPGA, a

USB2.0 (CY7C68013A) and a 1 Gbit low-power

DDR (MT46h64M16LFCK). The FPGA realizes the

architecture of a WISHBONE bus. The DDR is put

into use to store the excess data when the data size ex-

ceeds the volume of on-chip memories. And the read-

ing and writing of the DDR by FPGA is completed

by the MCB(Memory Control Block). The address

space of the WISHBONE bus is divided into sev-

eral blocks for DDR, Block RAM on-chip and Regis-

ter groups. Accessing these different addresses leads

to the operations for corresponding memory devices.

Here all the devices communicate with the personal

computer(PC) via the USB.

The I/O and Main boards are all plugged into the

Backplane and then integrated in 3U chassis. The

Backplane has 14 slots in total: the slot0 is for the

Main board and the other 13 slots are for the I/O

boards. The slot1 to slot13 each has 8 data wires con-

necting with the slot0 individually. The output signals

from I/O boards are collected at the Main board by the

backplane data buses.

2.2 The New Challenges

In the coincidence counting system for the photon en-

tanglement, there are 4 key elements: signal align-

ment, pulse width, signal sampling and data process-

ing. In order to carry out these tasks, a logic architec-

ture is designed in the FPGA of Main board and the

corresponding software is programmed. In the 12-

photon entanglement experiment, we face with sim-

ilar difﬁculties in general. Besides, new challenges

arise due to the increase of photon number and also

new requirements from users. Firstly, the new sys-

tem is expected to show more sorts of counting re-

sults in real-time and enable the control of the motion

controller(ESP301) to scan the length of optical path

of entangled photons. Furthermore, arbitrary type of

coincidence events needs to be read out while previ-

ously only some speciﬁc coincidence events can be

shown. Another tough task is to adopt the different

address deﬁnition for different number of photons. In

other words, the modules involving the access to the

addresses of coincidence events all need to be rede-

ﬁned. Here we make partial adjustments to complete

the new architecture depending on the existing one,

which is Shown in Figure 3.

2.2.1 Techniques of Pulse Scan

Signals would have different time delays accumulated

in different optical and electronic paths. In the past,

after comparing these signals with a oscilloscope, the

alignment of signals is performed by changing the

lengths of cables manually, which is repetitive and

onerous work. In the 6-photon entanglement exper-

iment, techniques of pulse scan are adopted to align

the input signals. Shifting the phase of clock step by

step, the relation between the single photon count and

the shifted clock phase is obtained, based on which

the arrival time of signals can be determined. Then all

the other signals would be aligned with the last arriv-

ing one by the internal delay block. Thus, the manual

adjustment of the cables is no longer necessary.

In our new coincidence system, we modify and

improve these pulse scan techniques. First of all, we

scan the single-channel count instead of the single-

photon count. The single channel count includes

not only the single-photon count but also the multi-

photon counts involving this channel together, which

indicates the counting would get started as soon as the

detector has response. The counting result in this way

is more accurate than before since the multi-photon

coincidence events are not missing in the statistics.

This is also better cost-saving during the test because

we can obtain the whole channel scan curves simulta-

neously in a clock period with a signal source. While

A Coincidence Counting System for Twelve-photon Entanglement Experiment

CH1

Ch2

CH1

CH2

CH1

CH2

WISHBONE BUS

From synchronous CLK

Signalx24

DCM

Internal

Delay

Pulse

Reshaping

Latch

FIFO

Data

Redistributor

Internal

RAM

External

DDR

Group

FPGA

USB

Scan

and

Align

Acquire

and

Monitor

Scan

and

Trace

Data

Read-Out

Shifted CLK

(a) Logic architecture of Main board

Labview

Stepper Motor

RS232

(b)Software architecture for Main board

Software

CH1

CH2

CH24

CH1

CH2

CH24

CH1

CH2

CH24

CH1

CH2

CH24

Figure 3: (a)The logic architecture of the Main board. The internal delay block adjust the delay line of FPGA to align the

input pulses. The pulse reshaping block narrow the pulsed width. The clock phase can be shifted by step and the sampled

data is distributed to different memory. (b)The software architecture for Main board. The software access the address of the

memories by WiSHBONE bus and the software are programmed by Labview.

in the past, to produce the single photon event of the

all channels, we have to set up an additional test sys-

tem to send signals in sequence every other cycle.

2.2.2 The Scheme of Data Storing

After the alignment of signals, the pulse reshaping

block is then used to compress the pulse width and

as a result the effects of inter-symbol interference and

dark count are eliminated. This block is kept un-

changed compared to the preceding coincidence sys-

tem. The narrowed pulses will be registered into

FIFO(First in, First out) at the rising edge of syn-

chronous clock which shifts the phase to the center

of the pulses. These data in the FIFO are then writ-

ten into different memory locations according to their

coincidence types. Thus, the scheme of address cod-

ing is determined by the number of entangled photons

and the size of data in the experiment.

In the 12-photon experiment, only a single degree

of freedom, ie. the polarization, of each photon is

taken into account, so two detectors would respond to

the state of one photon. For this reason, we can use 2

bits to describe completely the photon state. Totally,

24 bits are needed to record all sorts of 12-photon co-

incidence event, so each bit array can just take the

place of the memory address to store the correspond-

ing coincidence counting. Moreover, since the experi-

ment of multi-photon entanglement normally lasts for

several days and leads to a huge amount of data, a 64-

bit counter is deﬁned which ensures enough space for

all counting results and any overﬂow can be avoided.

The total storage space in need is 1 Gbits(2

*64 =

= 1 G) which exceeds the volume of the internal

block RAMs, so the excess data should be distributed

to the external DDR. In the experiment, the appear-

ance of photons in pulses follows a Poisson distri-

bution. The single photon and 2-fold events occupy

the primary body, which respectively take 24 and 528

sorts of coincidence. Though the multi-fold (more

than 2-fold) events is rare, their types of coincidence

increase exponentially. So we put multi-fold events

in DDR while single photon and 2-fold coincidence

events are stored in the block RAMs with 10 bits of

address.

In addition, register groups distinguished as the

to the system. The register-in group collects read-out

information from FPGA to PC like the status indicator

and the single-channel coincidence counting results,

while the register-out group carries the write-in mes-

sages from PC to FPGA including the control com-

mands. The register groups can offer favourable com-

munication medium for PC and the counting system.

PHOTOPTICS 2019 - 7th International Conference on Photonics, Optics and Laser Technology

(a)

(b)

0 2 4 6 8 10

Clock Shifting (ns)

2x10

4x10

Count

2x10

4x10

Count

Before Alignment

After Alignment

0 2 4 6 8 10

Clock Shifting (ns)

Figure 4: Pulse scan of 24 channels before and after align-

ment.(a)Before the delay adjusting, the arrival time of dif-

ferent channels differ greatly. (b)After the delay adjusting,

the arrival time of different channels are almost uniform.

2.2.3 The Relative Software

All the programs controlling our coincidence count-

ing are coded with Labview. One program named

“Scan and Align” helps the users to set parameters

concerning the alignment of the signals such as delay

value and clock phase shifting. It can access the reg-

ister groups and scan the single channel coincidence

counting.

After the alignment, another program named “Ac-

quire and Monitor” is used to conﬁgure the process to

acquire and monitor the coincidence results. Particu-

larly, with this program the measurement can be dis-

played in real-time, which makes it easier to take con-

trol over the experiment. To realize this function, we

select the wanted type of coincidence events and read

out the counting results from the memories. How-

ever, the exponential increase in the time cost to tra-

verse all sorts of coincidence would retard the refresh-

ing of the program. So in order to retain the refresh

rate high than 1s

−1

, we calculate the approximate re-

sult for only including the appointed N-fold and the

next (N+1)-fold coincidence counting and neglecting

all the others.

Finally, the program named “Scan and Trace”

serves to control the motion controller used for the

optimization of the photon coherence. The controller

moves step by step to obtain a curve of the count rate

which helps to adjust the optical path to the best state.

The last program named “Data Read-Out” can read

out the data of arbitrary appointed coincidence events.

Figure 5: The working interface of the Labview-based con-

trolling program. The test results are displayed on the inter-

face and agree with the estimate from the Possion equation,

demonstrating the right working status of the system.

3 TEST AND RESULT

To test the alignment of output signals in the experi-

ment, we send simulating signals to all the 24 SPDs

and then compare waveforms of the output signals be-

fore and after alignment, as shown in Figure 4. It can

be seen that the maximum time difference before the

delay alignment is about 3ns, but after adjusting the

delay line, the time difference is around 0.4ns. The

alignment operation provides a narrow time coinci-

dence window and a wide steady sampling window.

What is more, the pulse width can be deduced from

the test. When the count is stable, the width of the ﬂat

top can be considered as the signal pulse width. In the

experiment, the pulse width is about 3ns.

One of the user interface about displaying in real

time during experiment is shown in Figure 5. In the

ﬁgure, each row represents a sort of coincidence event

( “1”,“2” represent the detector response, “0” repre-

sent opposite.). We choose to display coincidence

of photon pairs from either the same entanglement

source or different source. The users ﬁrst count the

every pair source event and then estimate the coinci-

dence count by the Possion equation. When the result

of computation agrees with that displayed in real time,

the counting system is demonstrated to work reliably.

With our coincidence counting system, we are

able to optimize the photon coherence at the detectors

on the basis of a Hong-Ou-Mandel(HOM)-type inter-

ference test. The coincidence counting rates of four

photon detectors recording the four output channels

of the HOM experiment are measured, as the length

of one interfering path is scanned via a mirror tunable

with a motion controller. The optimal position of the

mirror is approached when the detectors the coinci-

dence of two transmitting photons and two reﬂecting

photons reach its maximal value and at the same time

the coincidence of one transmitting photon and one

A Coincidence Counting System for Twelve-photon Entanglement Experiment

reﬂecting photon is the minimal, as shown in Figure

0 1 0 0 2 0 0 3 0 0

4 0

6 0

8 0

1 0 0

C o u n t / s e c

P o s i t i o n ( µm )

Figure 6: The coincidence counting rates of four photon

detectors recording the four output channels of the HOM

experiment.

4 CONCLUSIONS

Based on the original system applied in 6-photons 18-

qubit experiment, we develop a reconﬁgurable coinci-

dence counting system with 24 input channels for 12-

photon entanglement experiment. Compared to the

previous scheme, we apply the former architecture

and perform a plenty of improvements. We adjust the

delay to align the input pulses and after alignment, the

maximum time deviations between output signals is

no more than 0.4ns. We place pulse reshaping blocks

to compress the input signals and achieve a average

pulse width of 3ns. Furthermore, we adopt the differ-

ent address coding method to count and distribute the

data to different memories. The maximum bandwidth

supports 90Mhz and the maximum data size supports

over 1Gbit. Finally, this coincidence counting system

is demonstrated to be feasible in the experiment of

12-photon entanglement.

In the future, some issues are still requires further

optimization. With the quantity of channels increas-

ing, the crosstalk between channels will get worse.

Next we would take more attempts to improve sig-

nal integrity. Besides, since the backplane bandwidth

is not high enough, the single-terminal routing can

be changed to the differential routing to increase the

bandwidth to GHz.

ACKNOWLEDGEMENTS

This work has been ﬁnancially supported by the Na-

tional Natural Science Foundation of China (Grant

Nos.61575185 and 61308014), and the CAS Key

Technology Talent Program. The author would like

to thank Zheng-da Li and Rui Zhang for his feedbacks

and discussions. We especially thank Prof. Jian-Wei

Pan for his guidance and support.

REFERENCES

Brunner, N., Cavalcanti, D., Pironio, S., Scarani, V., and

Wehner, S. (2014). Bell nonlocality. Rev. Mod. Phys.,

86:419–478.

Chen, L.-K., Li, Z.-D., Yao, X.-C., Huang, M., Li, W., Lu,

H., Yuan, X., Zhang, Y.-B., Jiang, X., Peng, C.-Z., Li,

L., Liu, N.-L., Ma, X., Lu, C.-Y., Chen, Y.-A., and

Pan, J.-W. (2017). Observation of ten-photon entan-

glement using thin bib3o6 crystals. Optica, 4(1):77–

83.

Crotti, M., Rech, I., and Ghioni, M. (2012). Four channel,

40 ps resolution, fully integrated time-to-amplitude

converter for time-resolved photon counting. IEEE

Journal of Solid-State Circuits, 47(3):699–708.

Einstein, A., Podolsky, B., and Rosen, N. (1935). Can

quantum-mechanical description of physical reality be

considered complete? Phys. Rev., 47:777–780.

Farr, W. H. (2012). Overview of single photon detection

technologies. In IEEE Photonics Conference 2012,

pages 20–21.

Gaertner, S., Weinfurter, H., and Kurtsiefer, C. (2005). Fast

and compact multichannel photon coincidence unit for

quantum information processing. Review of Scientiﬁc

Instruments, 76(12):123108.

Hadﬁeld, R. H. (2009). Single-photon detectors for optical

quantum information applications. Nature Photonics,

3:696.

Kim, Y.-H. (2003). Quantum interference with beam-

like type-ii spontaneous parametric down-conversion.

Phys. Rev. A, 68:013804.

Knill, E., Laﬂamme, R., and Milburn, G. J. (2001). A

scheme for efﬁcient quantum computation with linear

optics. Nature, 409:46.

Kok, P., Munro, W. J., Nemoto, K., Ralph, T. C., Dowling,

J. P., and Milburn, G. J. (2007). Linear optical quan-

tum computing with photonic qubits. Rev. Mod. Phys.,

79:135–174.

Kwon, O., Cho, Y.-W., and Kim, Y.-H. (2008). Single-mode

coupling efﬁciencies of type-ii spontaneous paramet-

ric down-conversion: Collinear, noncollinear, and

beamlike phase matching. Phys. Rev. A, 78:053825.

Pan, J.-W., Chen, Z.-B., Lu, C.-Y., Weinfurter, H.,

Zeilinger, A., and

Zukowski, M. (2012). Multipho-

ton entanglement and interferometry. Rev. Mod. Phys.,

84:777–838.

Schr

odinger, E. (1935). Die gegenw

artige situation in der

quantenmechanik. Naturwissenschaften, 23(48):807–

812.

Simms, P. C. (1961). Fast coincidence system based on a

transistorized time-to-amplitude converter. Review of

Scientiﬁc Instruments, 32(8):894–898.

Wahl, M., R

ohlicke, T., Rahn, H.-J., Erdmann, R., Kell,

G., Ahlrichs, A., Kernbach, M., Schell, A. W., and

PHOTOPTICS 2019 - 7th International Conference on Photonics, Optics and Laser Technology

Benson, O. (2013). Integrated multichannel pho-

ton timing instrument with very short dead time and

high throughput. Review of Scientiﬁc Instruments,

84(4):043102.

Wang, X.-L., Cai, X.-D., Su, Z.-E., Chen, M.-C., Wu, D.,

Li, L., Liu, N.-L., Lu, C.-Y., and Pan, J.-W. (2015).

Quantum teleportation of multiple degrees of freedom

of a single photon. Nature, 518:516.

Wang, X.-L., Chen, L.-K., Li, W., Huang, H.-L., Liu, C.,

Chen, C., Luo, Y.-H., Su, Z.-E., Wu, D., Li, Z.-D., Lu,

H., Hu, Y., Jiang, X., Peng, C.-Z., Li, L., Liu, N.-L.,

Chen, Y.-A., Lu, C.-Y., and Pan, J.-W. (2016). Ex-

perimental ten-photon entanglement. Phys. Rev. Lett.,

117:210502.

Wang, X.-L., Luo, Y.-H., Huang, H.-L., Chen, M.-C., Su,

Z.-E., Liu, C., Chen, C., Li, W., Fang, Y.-Q., Jiang, X.,

Zhang, J., Li, L., Liu, N.-L., Lu, C.-Y., and Pan, J.-W.

(2018). 18-qubit entanglement with six photons’ three

degrees of freedom. Phys. Rev. Lett., 120:260502.

Zhang, C., Li, W., Hu, Y., Yang, T., Jin, G., and Jiang, X.

(2016). 48-channel coincidence counting system for

multiphoton experiment. Review of Scientiﬁc Instru-

ments, 87(11):113107.

Zhong, H.-S., Li, Y., Li, W., Peng, L.-C., Su, Z.-E., Hu,

Y., He, Y.-M., Ding, X., Zhang, W.-J., Li, H., Zhang,

L., Wang, Z., You, L.-X., Wang, X.-L., Jiang, X.,

Li, L., Chen, Y.-A., Liu, N.-L., Lu, C.-Y., and Pan,

J.-W. (2018). 12-photon entanglement and scalable

scattershot boson sampling with optimal entangled-

photon pairs from parametric down-conversion. ArXiv

e-prints.

A Coincidence Counting System for Twelve-photon Entanglement Experiment