/
FABULOUS FABULOUS

FABULOUS - PowerPoint Presentation

calandra-battersby
calandra-battersby . @calandra-battersby
Follow
397 views
Uploaded On 2017-04-20

FABULOUS - PPT Presentation

FDMA Access By Using Lowcost Optical network Units in Silicon photonics Project outcomes and lessons learned July 2017 SOME THOUGHTS ABOUT SILICON PHOTONICS What do silicon manufacturers think about silicon photonics ID: 539851

onu silicon cost gbps silicon onu gbps cost pon qam photonics optical fabulous rate modulator project wavelengths pieces bit system odn fdm

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "FABULOUS" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

FABULOUS

FDMA Access By Using Low-cost Optical network Units in Silicon photonics. Project outcomes and lessons learned.

July 2017Slide2

SOME THOUGHTS ABOUT SILICON PHOTONICS

What do silicon manufacturers think about silicon photonics?

Silicon Photonics is great, I can give you any component you like, just ask me millions pieces per year!

What do silicon manufacturers think about silicon photonics?

Silicon Photonics is great, I can give you any component you like, just ask me

millions pieces

per year!

What do system manufacturers think about silicon photonics?

It’s cheap and I like it, as long as it doesn’t change my whole system

What do system manufacturers think about silicon photonics?

It’s

cheap

and I like it, as long as it doesn’t change my whole system

CHEAP

MILLIONS PIECESSlide3

LONG-HAUL

: high-bit rates, expensive devices, non million pieces per

year not the best entrance market

DATACOM

TELECOM

FTTx

: already millions of customers for EPON, GPON, XGPON

a good potential for silicon

INTERCONNECT

LET’S FOCUS ON OPTICAL TELECOMSlide4

Two types of services

P2P-WDM: dedicated

wavelengths

used

for high-capacity links,

eg

.

for

4G or 5G BTS

TWDM-PON: 4

wavelengths

(in

first releases) to transport TDM services with an aggregate capacity of 10 Gbps DS and 2,5 Gbps US per wavelength

The PON infrastructure is the same of XGPON, the wavelength plan has to be compatible with GPON, XGPON, RF-video

The first PON standard to rely on DWDM

THE NEW PON STANDARD: G.989Slide5

Low-cost

tunable

lasers (for the ONU)

are still an open issue.

No room for

SiP

optoelectronic devices

(

SiP

not mature for sources).

TDM: in its timeslot, the user has to go

max speed (10 Gbps DS, 2,5 Gbps US).

Most of the electronic is ready but it has

to be fast.

IS G.989 SUITABLE FOR LOW-COST CPE?Slide6

THE FABULOUS EU PROJECTSlide7

EU STREP contract 318704

7

FABULOUS AT-A-GLANCE

4,3M€

TOTAL COST

2,9M€

FROM THE EU

1

st

October

2012

PROJECT START

36+9

months

DURATIONSlide8

WE DON’T WANT OPTICAL SOURCES AT THE ONU

REFLECTIVE PON

Most of WDM complexity and cost (

tunable

laser) left at the OLT

No uncontrolled wavelengths

at ONU switch-on

PROs

HERE COMES THE SILICON!

Challenging power budget:

coherent detection at the OLT

Modulator and

tunable

optical

filter

at

the ONU

CONs

Modulator and

tunable

optical

filter at the ONU

LET’S RETHINK NG-PON2 FROM SCRATCHSlide9

WE WANT THE ONU ELECTRONICS TO RUN

AT THE USERS’ BAUD RATE

FREQUENCY DIVISION MULTIPLEXING

Continuous data-stream

Long FEC, M-QAM and DSP easier to use

with respect to burst-mode systems

PROs

Many degrees of freedom for bit-rate adaption

on per-user basis

Linear optoelectronics needed

DAC and ADC

CONs

Frequency

mixers

LET’S RETHINK NG-PON2 FROM SCRATCHSlide10

Reflective WDM PON based on FDM / FDMA

SUMMARIZING…Slide11

The ONU

reflects

the seed signal, modulates itand performs 90° polarization rotation

Purpose

of the

project

is

also

to integrate the ONU on a

Silicon

Photonics PIC

…SUMMARIZING…Slide12

Polarization rotation allows simplified coherent detection at the OLT

…SUMMARIZINGSlide13

WITH DISCRETE COMPONENTSSlide14

1

wavelength

Modulation format

Electrical bandwidth per channel

Net bit-rate per user

Maximum ODN loss

64-QAM660 MHz3 Gbps

23.0 dB64-QAM

1320 MHz

6 Gbps23.0 dB

16-QAM

330 MHz

1 Gbps31.0 dB16-QAM1650 MHz5 Gbps30.5 dB16-QAM3300 MHz10 Gbps

28.0 dBQPSK66 MHz

100 Mbps41.0 dB

QPSK330 MHz500 Mbps

38.0 dB

QPSK

1650 MHz

2.5

Gbps

34.5 dB

QPSK

3300 MHz

5 Gbps

32.5 dB

BPSK

330 MHz

250 Mbps

40.5 dB

OOK

6250 MHz

10 Gbps

38.5 dB

~11GHz of modulator bandwidth

BENCHMARKING WITH DISCRETE COMPONENTSSlide15

Number of ONUs

Bit-rate per ONUMaximum ODN lossTheoretical splitter insertion lossOptical margin324 Gb/s

27 dB15 dB12 dB642 Gb/s28 dB

18 dB

10 dB

1281 Gb/s

31 dB (Class N2)21 dB

10 dB

Number of ONUs

Bit-rate per ONU

Maximum ODN loss

Theoretical splitter insertion lossOptical margin

128375 Mb/s37 dB (Class E2)21 dB16 dB256187.5 Mb/s38 dB (Class E2)24 dB14 dB

51293.75 Mb/s

39 dB (Class E2)27 dB

12 dB

4 wavelengths @32 Gbps

4 wavelengths @10 Gbps

BENCHMARKING WITH DISCRETE COMPONENTSSlide16

Modulator bandwidth

Number of ONUs per λNet Bit-rate per ONU and modulation formatTotal upstream net Bit-rate per λMaximum ODN loss11 GHz (reference)32

1 Gbps (16-QAM)32 Gbps31.4 dB22 GHz322 Gbps (16-QAM

)

64 Gbps

29.3 dB22 GHz

322.5Gbps (32-QAM)

80 Gbps25.1 dB

22 GHz

641 Gbps (16-QAM

)

64 Gbps30.1 dB

33 GHz323 Gbps (16-QAM)96 Gbps28.1 dB33 GHz322.25 Gbps (8-QAM)72 Gbps

32.0 dB33 GHz

641.5Gbps (16-QAM)

96 Gbps28.7 dB

33 GHz

64

1.1Gbps (8-QAM

)

72 Gbps

33.0 dB

WITH A BETTER MODULATORSlide17

Preliminary integrated ONU

version:

polarizing beam splitter, coupler,

segmented MZ intensity modulator, CMOS electrical driver flip-chipped on top

THE FIRST SILICON DEVICESlide18

Integrated on a test-board for system experiments with an external bi-directionally used SOA

THE FIRST SILICON DEVICESlide19

EXPERIMENTAL RESULTS AT 500Mb/s PER USER

Not yet suitable for PON but of potential interest for Passive Optical LANSlide20

SOA: A FUNDAMENTAL BLOCK

InP

over Si enables amplification in PIC:

fundamental for complex functionsSlide21

ABOUT 100$

(80% DUE TO PACKAGING)

Is it so important to have a low-cost ONU?

Most of the cost of PON is in the infrastructure!

NG-PON2 is designed to be compatible with existing

infrastructure

A miniaturized low-cost device could be built-in in PC, video-sets, etc.

Considering:

>1.000.000 pieces/year

the

size of the

SiP

chip, the

SOA,

the driver, commercial

FDM IC

IS IT LOW COST?Slide22

the effectiveness of a self-coherent reflective WDM/FDM PON architecture in complying to ODN loss standard requirements

a target cost for the FABULOUS ONU lower than 100€, suitable for home devices

the feasibility of complex Photonic Integrated Circuits and of low-speed DSP with FDM approach

We have demonstrated

SCIENTIFIC OUTCOMES

We need to improve the PIC performances via amplificationSlide23

we co-develop systems and components:

it is not only substituting and integrating

we lower packaging costs

we can rely on amplification in the PIC; hybrid integration seems an efficient way

Let’s wait for the silicon laser!

SiP

can definitely benefit optical access provided that:

LESSONS LEARNEDSlide24

FURTHER READINGS

We published 13 Journals (some for open access) and over 30 conference papers. Slide25

The research leading to these results has received funding from the European Community's Seventh Framework

Programme FP7/2007-2013 under grant agreement n°318704, titled:

25

ACKNOWLEDGMENTS

To contact the coordinator:

info@fabulous-project.eu

WEB site:

www.fabulous-project.eu

FABULOUS:

“FDMA Access By Using Low-cost Optical Network Units in Silicon Photonics”