Transcript Josep Prat

EU-Japan Workshop, Brussels 18-April-2013
Research on Next Generation Optical
Access Networks
Josep Prat
Universitat Politècnica de Catalunya (UPC)
Barcelona
1
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
Optical access evolution
WDM
xWDM
CAPACITY, cost
(per user)
ngPON 1
ngPON 2, 3 ..
FTTH
WDM-PON
1G
100M
- PON
ultra-dense WDMPON E (OFDM) /
Coh
FTTH-PtP
HFC
10M
xDSL
FTTH
XG/G/E-PON
WDM&TDM-PON
ADSL
TIME
POTs
2013
2008
ACCORDANCE
SARDANA
EU projects:
EUROFOS NoE
MUSE IP
2
COCONUT
ALPHA IP
EU-Japan Workshop, Brussels 18-April-2013
OASE IP
DISCUS IP
[email protected]
PON Optical spectrum
FSAN-ITU Estandards:
?
XG-PON
G-PON
upstream
upstream
G-PON
downstrea
m
1260 1270 1280 1290 1300 1310 1320 1330 1340 1350 1360
NG
2
Vide
o
NG
2
1480 1490 1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 1610
possible bands for ngPON3
10
10
(with compatibility) :
3
X
G
40 nm
•
Video: IP
40 nm / 6.25 G = 800 channels !!
•
EDFA
CAPACITY x 100 !!
•
Low loss fiber
•
High availability of comp.
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
E/O BW & power Efficiency
Segment 1:
Copper-based Access
(DSL)
ONU 1
Segment 2:
Hybrid Optical/Wireless
to Segment 4
to Segment 3
to Segment 2
to Segment 5
to Segment 1
BS 1
(ONU 2)
ONU 3
ONU 1
Segment 3:
OFDMA/DSCA PON
ONU 3
from Segment 1
from Segment 5
from Segment 2
from Segment 3
CO
from Segment 4
ONU 2
BS 1
BS 3
ONU 1
ONU 3
ONU 2
EF = Rb / BWD
Segment 4:
Pure Wireless
(e.g. WiMax, LTE)
BS 2
Segment 5:
Legacy (e.g. GPON, EPON)
and NG Optical Access
1000%
RoF
OFDM
100%
Electrical
Spectrum
Efficiency
•
Statistical WDM
multiplexing
•
Time-switched phase
diversity&modulation
ud-WDM
10%
1%
GPON
WDM/TDM
1:K
D1,…, Dm
RN1
Central
Office
XG-PON
Casa
RN2

U
1,…,
Casa
U2N
Upstream Signals
Casa
TDM TREE
RNi
1:K
RNj
U1,…, U2N
Edificio 1
PowerComsp.:
= a C Vd2 fr + Io 10Vt/S Vd
x Activity %
4
RNN
0.1%
1%
10%
RSOA
Factory
PIN/APD
1:K
Bidirectional Transmission
RNN-1
D m+1,…, D2N Downstream Signals
0.01% 0.1%
ONU
Building 2
WDM RING
Casa
100%
Optical Spectrum Efficiency
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
PON alternatives
GPON XGPON
Aggregate BW
2.5/1.25
10G/2.5
TWDMPON
WDM-PON
40G/10G
>40G
Depends on
type
Good if combined
with WDM
>40G
1 AWG port
assigned
N subcarriers
1G-10G Good
if +TDM
Up-stream or
tunable laser
Upstream, PAPR,
optical amplif. or
coherent
Coherent RX,
random/tunable
lasers ?
If ONU only
handles assign.
subc.
User rate = line
rate
Granularity
(min BW/user)
Tunable TRX
ONUs
Complexity, cost
Energy efficiency
BWuser
/
10G
SCM/OFDM-
PON
BWuser
/
10G
Good if combined
with WDM or
coherent
Optical spectrum
efficiency
best
Power budget
Incremental
scalability (n.,BW)
Compatibility
5
cUDWDM
-PON
Limited by
device, AWG,
or RBS
AWG
EU-Japan Workshop, Brussels 18-April-2013
Sensitivity issue
Wireless
Splitter if
coherent det.
[email protected]
SARDANA project
Scalable
Advanced
Ring-based passive
Dense
Access
Network
Architecture”
Participant name
6
Activity: ICT-1-1.1 - Network of the Future
Grant agreement n.: 217122 (SARDANA)
STREP: 2008-2010, 2.6 MEuro
Short name
Country
1
Universitat Politecnica de
Catalunya
UPC
Spain
2
France Telecom / Orange
France
3
Tellabs
4
Intracom S.A. Telecom Solutions
5
Instituto de Telecomumicações
6
High Institute of Communication
and Information Technology
FT
TELLABS
IntraCOM
IT
ISCOM
7
Research and Education
Laboratory in Information Tech.
AIT
Finland
Greece
Portugal
Italy
Greece
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
SARDANA architecture def.
• Resilient trunk
• Fully passive
• Hybrid:
• WDM Metro ring
• TDM Access trees
D1,…, Dm
RN1
• Cascadable remote nodes
• New adoption of remotelypumped amplification
• Colourless ONU
• RSOA
• Tunable laser
• 10G-2.5G (1G-100Mb /user)
• 100 Km
• 1000 users
CO
ONU
ON
U
1:K
RN2
ONU
ON
U
U 1,…, U2N
Upstream Signals
TDM TREE
RNi
1:K
WDM RING
RNN
ONU
ON
U
ONU
ONU
ON
U
RNj
U1,…, U2N
ONU
ON
U
PIN/APD
1:K
RNN-1
RSOA
Bidirectional Transmission
D m+1,…, D2N Downstream Signals
ONU
ON
U
• Multi-operator
• Based on xGPON, but
transparent.
• IP trafic
7
SARDANA PON
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
SARDANA inter-operability & multi-operability
8
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
3 testbed demonstrations
1.
Stakeholder demo in Espoo (Finland) in
October 2010 (integration and tests)
2.
Field trial in Lannion (France), in January over
the metropolitan 12x 18 Km ring cable
3.
Public demo at the Exhibit of FTTH Council
Europe 2011, in Milan in February 2011.
9
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
ACCORDANCE STREP project
A Converged Copper-Optical-RaDio OFDMA-based
Access Network with High Capacity and FlExibility
Consortium
Objectives

Definition of a novel Access Network architecture
achieving convergence among heterogeneous
technologies (optical, wireless, copper).

Propose low-cost, low-complexity concepts to achieve
ultra high data rates in the access network (up to
100Gbps aggregate and more than 10Gbps in each
segment).

Introduction of flexible bandwidth allocation concepts
using dynamic FDM and OFDM sub-carrier
assignment.

Provision of smooth migration from and coexistence
with legacy access solutions.

Demonstration of the ACCORDANCE concepts using
experimental test beds.
10
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
OFDM-PON
(ACCORDANCE STREP project)
Envisioned Converged Network Architecture
Segment 1:
Copper-based Access
(DSL)
ONU 1
Segment 2:
Hybrid Optical/Wireless
to Segment 4
to Segment 3
to Segment 2
to Segment 5
to Segment 1
BS 1
(ONU 2)
ONU 3
ONU 1
Segment 3:
OFDMA/DSCA PON
Integrated
DEMO
ONU 3
from Segment 1
from Segment 5
from Segment 2
from Segment 3
CO
from Segment 4
ONU 2
BS 1
BS 3
ONU 1
ONU 3
ONU 2
11
Segment 4:
Pure Wireless
(e.g. WiMax, LTE)
BS 2
Segment 5:
Legacy (e.g. GPON, EPON)
and NG Optical Access
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
OFDM-PON system
OLT
ONU
DS signal DS aux carrier
Clocks synthesis
low-phase
noise laser
source
oIQ
modulator
twin MZM
twin driver
6
to OLT
to ONUs
T
low-phase
noise laser
source
low-phase
noise laser
source
G2
tx subchannel select
seed link
el.
up-conversion
6
ONU MAC
DAC
6
IFFT
FEC
R
G1
linear
EAMdriver
FFT
Mo d
ADC
6
REAM
6
switch
ADC
FEC
DSP
256FFT
electrical
tuning
section
R
OLT-MAC
pol-div,
pol-div,
90° PBS
90°
hybrid
hybrid
PIN TIA
ADC
switch
T
clock synthesis
FEC
6
optical
pre-amp
Demod
DAC
25 Gsps
DAC
25 Gsps
Rx subchannel select
carrier
offset
control
ADC
FEC
DSP
256IFFT
low-phase
noise laser
source
DAC
6
Seed signal combiner:
DS/US signals
Seed
3dB
DS/US/Seed signals
splitter-stage
Indicative experimental results by ACCORDANCE
partners:
R. Schmogrow et al., “101.5 Gbit/s Real-Time OFDM Transmitter with
16QAM Modulated Subcarriers” , OFC’11
OFC’13: P. Schindler et al. OW1A.5
12
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
Flexible BW Allocation to deal with:
the Power difference between (ONU1-ONU2)
for proper detection at BER=10-3, leaving ONU1 with B(Q)PSK modulation
ONU2 BW occupation [%]
40
30
20
10
50
0
50
ONU2 BW occupation [%]
40
30
20
10
0
f
e
d
b
a
c
a
b
c
d
e
f
The high BW granularity allows OFDMA to deal with differential link loss among users.
13
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
Example of ACCORDANCE MAC Scheduling
time
sub-carrier 1
ONU 2
ONU 5
sub-carrier 2
ONU 1
ONU 1
ONU 4
ONU 6
ONU 3
ONU 6
ONU 3
Inter-Frame Gap
ONU 1
Various options possible:
•
Multiple subcarriers per ONU (hybrid
OFDMA/TDMA)
•
Multiple ONUs per subcarrier (TDMA)
…
Inter-Frame Gap
ONU 2
sub-carrier N
OFDMA Frame
(duration = few ms)
PHY/MAC Cross-layer optimization
14
•
Use of Adaptive Subcarrier Modulation (ASM)
•
Use a different m-QAM for each ONU based on
transmission performance
•
MAC takes into account different bitrates for a
subcarrier each time, depending on the allocation
•
Lower delay for same traffic, i.e. capacity is
effectively increased
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
COCONUT project: Coherent UD-WDM
• Can we improve the spectral efficiency in both optical&electrical domains?
• and the BW?
• and optical fine granularity ?
• and the Power budget?
Proposal:
Coherent UD-WDM with random-wavelength DFB lasers
I’’ Q’’ I’
t0
Q’
t0+T t0+3T/2 t0+2T
t
+
ONU 1
ONU i
OLT
SW
t0+T/2
Optical
Input
TX
+
array
I&Q
processing
Data
Data
DS
CLK
Phase
Scrambler
Local
Laser
hRX
TX
array
hRX
Data
US
us ds
hr-OSA
15
ONU n
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
COst-effective COhereNt Ultra-dense-WDMPON for lambda-To-the-user access
• STREP 318515, ICT Objectives 1.1 a) b)
• 1st-Nov + 36 months
• Coordinator:
– SSupSA (Ernesto Ciaramella)
– Vice-coordinator: UPC (Josep Prat)
• GOAL: efficient TTH
– 1G-to-the-Home + 10G-to-the-Antenna
– By: ultra-dense WDM-PON, by:
• Statistical WDM multiplexing
• Simplified coherent transmission
16
th
10 Concertation
Meeting,
EU-Japan
Workshop, Brussels
18-April-2013
[email protected]
What Ligh Source at Home for WDM ?
• Requirements:
– All the same : colourless (wavelength agnostic)
• for feasible provisioning and high volumes fabrication
– Broadband: > 1GBit/s
• Possible solutions:
– Tuneable laser : still expensive and non-repetible
– Reflective SOA, wavelength remodulation: limited perform.
– Random-wavelength laser !?
• Statistical multiplexing
• CO adapts to ONU
• Benefit of chaos ..!
17
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
Wavelength contention statistics
•
•
•
•
•
Grid= 6.25GHz
Total BW = 30nm
Uniform laser statistical
distribution
Max. N. users = 600
Thermal 2nm (DFB)
With thermal tuning
Without thermal tuning
(Results 1000 times aver.)
Spectral efficiency:
• Periodical allocation: 16 %
• Statistical allocation: 14%
18
EU-Japan Workshop, Brussels 18-April-2013
~ 1 ONU in 512
to replace
[email protected]
NGPON2/3 arch. features comparison
XG-PON
TWDM
WDM/TDM-PON
OFDM-PON **
UD-WDM
Aggregate BW
User data*
Bit rate
10 GBit/s
50-200 MBit/s
40 GBit/s
50-500 MBit/s
320 GBit/s
50M-1G Bit/s
10 - 100 GBit/s x8
100 MBit/s
512 GBit/s
1 GBit/s
10 GBit/s
10 GBit/s
10 GBit/s
10 - 100 GBit/s
1 GBit/s
ONU Elec. BW
Elec.Spectrum
Eff.(up)
Number of users
Wavelength
Channels
Wavelength
Spacing
Optical Spectrum
Eff. (up)
Passive distance
reach
COST (per user,
1G, relative)
ENERGY C. (per
user, relative)
kHz - 10GHz
kHz - 10GHz
kHz - 10GHz
Fi – Fi+50MHz
MHz - 1 GHz
0.5%
1%
0.5% - 10%
< 200%***
< 200%
32
128
512
128
512
1-2
4-8
32
1-8
512
-
100 GHz
50 - 100 GHz
>50 GHz
3 - 12 GHz
-
10%
20-40%
1% - 20%
20 - 60%
20 km
20 Km
60 km
20 km
100 km
1
2
3 – 0.8
26 - 1
10 – 0.8
1
1
2 – 0.8
10 – 0.7
5 - 0.4
compatibility
protection, metroaccess converg.
wired&wireless
convergence, SCDBA
user independency
splitter
Tunable ONU
Colourless ONU,
Optical remote pump
power
Up-stream OBI,
PAPR, Sensitiv.
Wavel.tuning,
polarization
Added features
Issues
(typical-mean values)
19
* Symmetrical PON; ** 16QAM; *** Only odd carriers used.
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
Possible colaboration research topics
• Definition of requirements, test of solutions and standardization steps
towards NGPON3, implementing the concept of lambda-to-the-home
TTH
– from 100 Mb to 10 Gb to each home or antenna site
• Development of low cost phase-modulated lasers for elastic coherent
udWDM-PON.
• Global Convergence of udWDM wired access, wireless and metro
networks, in physical, MAC and control plane layers
– current collaboration between NEC and Accordance/AIT
• All optical processing with OCDMA multiplexing, transmission and
GMPLS routing
– topic in joint research by NICT, RomaTre, Aveiro Univ.
• Extension of the concepts and techniques to smaller ranges:
– CORE -> ACCESS -> LOCAL WIRED AND WIRELESS
20
EU-Japan Workshop, Brussels 18-April-2013
[email protected]
THANKS!
Acknowledgements:
European 7th Framework
European 7th Framework
European 7th Framework
European 7th Framework
21
Programme project SARDANA (www.ict-sardana.eu)
Programme project ACCORDANCE (www.ict-accordance.eu)
Programme Network of Excellence EUROFOS (www.euro-fos.eu)
Programme project COCONUT (www.ict-coconut.eu)
EU-Japan Workshop, Brussels 18-April-2013
[email protected]