Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
FP7 & H2O2O Network Technologies Concertation DayBrussels, March 1st
Sergi Figuerola ([email protected]) Dimitra Simeonidou ([email protected])
Agenda CaON afternoon sessionPart 1: Results of FP7 optical projects [7’ presentations]• FP7 LIGHTNESS: (http://www.ict‐lightness.eu/) (Nicola Calabreta )
• FP7 COCONUT: (http://www.ict‐coconut.eu/) (Ernesto Ciaramella)
• FP7 DISCUS: (http://www.discuss‐project.eu/) (David Payne)
• FP7 FOX‐C (http://www.ict‐fox‐c.eu/) (Ioannis Tomkos)
• FP7 STRAUSS: (http://www.ict‐strauss.eu/en/) (Raul Muñoz)
• FP7 COSIGN (http://www.fp7‐cosign.eu/partners/) (Lars Ditman)
• FP7 COMBO (http://www.ict‐combo.eu/) (Jean‐charles Point)
Part 2: Updates of the progress H2020 optical projects [7’ presentations]
• H2020 ACINO (http://www.acino.eu/): Application centric IP/Optical network orchestration (Domenico Siracusa)
• H2020 iCIRRUS (http://www.icirrus‐5gnet.eu/) : Intelligent converged network consolidating radio and optical access around user equipment (Nathan Gomes)
• H2020 INSPACE (www.ict‐inspace.eu/) : Spatial spectral flexible optical networking (Ioannis Tomkos)
• H2020 ORCHESTRA (www.orchestraproject.eu) : Optical performance monitoring enabling flexible networking(Manos Varvarigos)
• H2020 ROAM (http://www.roam‐project.eu/): Revolutionising optical fibre transmission and networking using the Orbital Angular Momentum of light (Antonella Bogoni)
Part 3: 5G opportunities for the optical community (round table discussion)
Part 1: Results of FP7 optical projects
LIGHTNESS LOW LATENCY AND HIGH THROUGHPUT DYNAMIC NETWORK
INFRASTRUCTURES FOR HIGH PERFORMANCE DATACENTRE INTERCONNECTS
• All optical hybrid OPS/OCS programmable data plane
• AoD-based intra-cluster and inter-cluster optical back plane
• Programmable NICs move electronics down to the servers
• SDN controller– OpenDaylightplatform
• Extensions for OpenFlow based provisioning of the LIGHTNESS hybrid DC fabric.
• OpenFlow Agent for translating commands/information
DATA & CONTROL PLANE HIGHLIGHTS
• Designed and demonstrated a novel all-optical data plane for delivering high capacity, low latency, and programmable DCN• OPS -- High port count, low latency, used for switching short-lived packet flow• OCS -- Accommodate long-lived traffic with ultra-low latency• Optical ToR switch -- Offer low latency, high energy and cost efficient• Programmable NIC -- Enable high bandwidth and low latency intra-rack server-to-server full mesh direct
interconnect, and eliminate the electronics in the ToR switch • Programmable synthesis of DC architecture to adapt dynamically to the application and traffic
• Implementation of the unified SDN control plane for full optical and hybrid OCS/OPS programanble data centres• Fully aligned with ONF SDN architecture 1.0• OpenFlow-based abstraction of the optical data plane
• LIGHTNESS OpenDaylight SDN controller• Full set of new OpenDaylight features for programmable NIC, OPS, AoD, optical ToR logics• Extended OpenFlow protocol @southbound for control of the LIGHTNESS optical devices• Dedicated control agents to make the optical devices OpenFlow enabled
• Full network virtualization solution to be used with the LIGHTNESS hybrid optical data centre fabric• VDC composition for multi-tenant virtual network provisioning• Support for multicast/unicast and OPS/OCS virtual topologies
6 E. Ciaramella, COCONUT Achievements
COCONUT achievements and prospectsCOst‐effective COhereNt ultra‐dense‐WDM‐PON for lambda‐To‐the‐user access
Demonstrated low‐cost coherent detection for access networks• narrow spacing (6.25 GHz at 1.25 Gbs)• high power budget (compatibility with splitter‐based PON, around ‐50 dBm
sensitivity!)• proven feasibility of long‐reach 4x10 Gb/s
Large demo of COCONUT network on Feb. 5th 2016• various COCONUT 1.25 Gbs and 10 Gbs solutions+ GPON on a installed fibre
network in Pisa• various use‐cases (GbEthernet, 4K video, Wi‐Fi hot spot etc.)• significant impact on media (see www.ict‐coconut.eu )• selected solutions already presented to FSAN (Atlanta Workshop)
Next step: scaling solutions to effectively support 5G front‐haul
email: [email protected]
7 E. Ciaramella, COCONUT Achievements
COCONUT network demonstration
Pisa Network
SMF SpoolFeeder
1xN
ONU (ASK) 1.25 GbESSSA
RX(heteroPSK)1.25 GbEUPC
C‐OLT
HR‐OSA
CO
ERI WiFiInternet
IP‐TVInterfac
e(Promax
)
HD TV
TV Broadcas
t(Promax)
TX (homoPSK)UPCTX/RX
(homo PSK)FPGA 1.25
GbE UPC
CE
MultimediaServer
Separate SetupIII‐V labs devices(D‐EML and FM*)
Separate SetupMAC Layer
Functionalities
E‐PONOLTPRO
TX (hetero PSK)
1.25 GbEUPC
232221
24
22212019
17
46 7
8
ONU (ASK) 1.25 GbESSSA
4x 10 Gb/sRX –PRBSSSA
ONU (ASK) 1.25 GbESSSA
TX/RX (ASK) 1.25 GbESSSA
TX‐UDWDM 8 ch
TX ASK 4x 10 Gb/s
TXSSA
3’
5
3
2
1
6
3
5
2
1
E‐PONONU
CE
DISCUS – A solution to the 3 major problems facing future broadband networks:
1. Financial - The cost of network provision:– Large mismatch between revenues and the cost of providing capacity.
2. Environmental -The huge growth in power used by the worlds telecommunications networks.
3. Social - The need to avoid a “digital divide”. – Provide the same service capability for customers in sparse rural
communities as dense urban areas.– With minimal government subsidies!
An end to end architecture for a high performance, economic and low power future broadband network
Summary of technical achievementsLR-PON is a key enabler for the required architecture change.
• It enables the flat core which has been shown to be the lowest cost solution for the future network
• TDMA+DWA provides future proof bandwidth allocation to all users.
• DWA solutions analysed and preferred solution without needing AMCC or synchronised quiet windows identified.
• LR-PON protocol channels can co-exist with 100Gb/s DP-QPSK s .
• 40Gb/s upgrade using bit-interleaved protocol.
• Technology gaps investigated: burst-mode EDC, burst mode FEC, tuneable filter and low cost tuneable laser, burst mode FEC (not in DISCUS) demonstrated.
• SOA amplifier options can open up other optical windows on same LR-PON fibre infrastructure.
10Concertation Meeting, Brussels ‐ 01/03/2016
FOX‐C (Flexible Optical Cross‐connect Nodes enabling next generation flexible optical networking)
Main goal: • The development of an
Add/Drop node processor with ultra fine switching granularity.
Additional goals:• The evaluation of compatible
transmitter receiver schemes• The techno‐economic
evaluation and the study of the networking benefits and feasibility
• The experimental demonstration of the overall concept
All objectives successfully completed by the end of 2015
High Spectral Resolution filter elementResolution <1GHz, Addressability 400MHzAWG‐based design extendable to a WSS
Final FOX‐C demo at OrangeLabs
Implementation of Nyquist‐WDM andMulti‐band OFDMSuper‐channel TxRxat 1Tb/s with up to 400Gb/s per sub‐Ch
Extensive set of both Transmission and Switching node performance results
11Concertation Meeting, Brussels ‐ 01/03/2016
Benefits offered by FOX‐C solution
The ability to transparently adapt lower rate tributaries into elastic super‐channels• The developed switching technologies with record sub‐1GHz resolution allow low rate
tributaries to be added/dropped transparently with minimum spectral guard band• Concept is proven particularly beneficial for cases like inter‐data centre communications and
extended metro‐access connections to core where a large number of lower rate 1‐10 Gbpstributaries are present.
The resource optimization of the current infrastructures• Extensive network evaluation and comparison results shown that FOX‐C switching scheme
can extend the blocking‐free lifetime of current networks and therefore alleviate the immediate need for further investments in network capacity.
The direct compatibility with the mixed‐line rate EON approach• The two stage node design offers a combination of the typical EON approach and the added
benefits of FOX‐C multi‐granular switching approach• The HSR switching feature is practically an added feature to an elastic switch and is
implemented as an additional switching stage in the FOX‐C node
The significant cost and power consumption savings• The savings originate from the adoption of all‐optical traffic grooming and are particularly
evident when compared with the OTN‐based electronic approach and for increased traffic loads as expected in the near future.
The STRAUSS project focuses on the design, development and integration of: Highly integrated and scalable software defined optical transceivers supporting
bandwidth variable multi‐flows for flexible Ethernet transmission based on OFDM and DMT.
Flexi‐grid DWDM optical circuit switching node architectures for long haul transport.
Cost/energy efficient and extremely fast‐performing switching node architectures based on variable‐capacity and fixed‐length optical packet switching technology for access and aggregation networks, and OPS/OCS integrated interface.
A virtualization layer for dynamic and on‐demand partitioning of the optical infrastructure, offering virtual optical networks .
Control plane architectures based on either legacy (e.g. GMPLS) and new (e.g. OpenFlowbased) approaches for the control and management of virtual optical networks.
A service and network orchestration layer for the interworking and coordination of heterogeneous control planes and transport technologies to offer end‐to‐end Ethernet transport services.
StraussScalable and efficient orchestration of Ethernet services using software-defined and
flexible optical networks
STRAUSS achievements I
Detailed system architecture developed in the STRAUSS project
Overall architecture
Applicability of the STRAUSS architecture to datacenter connectivity
Data Center
Core OPS Switches
Flexi‐grid DWDM network
ToREthernet Switches
Aggregation OPS Switches
OpenFlowController
GMPLS Controller
GMPLS Controller
GMPLS Controller
Data Center
Core OPS Switches
ToREthernet Switches
Aggregation OPS Switches
OpenFlowController
AS‐PCE
Multi‐domainNetwork Orchestration
Cloud Controler
Global Cloud and Network OrchestrationVirtual IT resources Connectivity between virtual IT resources
COSIGN reference modelAdvanced Optical Hardware and SDN for Next-Generation Data Centres
• Integrate the DCN resources into a common resource control framework
• SDN based framework selected for integration with computational and storage/memory resources
• Multiple platforms and interface ”standards” under investigation.
• COSIGN short and medium term Data Plane Infrastructure
– Electro-optical DCN with hybrid topology– Electronic switches as ToRs and leafs– Optical switches for dynamic circuit creation– Virtualized edge with OVS
ToR
Leaf
Circuit Switch
VM
VM
VM
VM
VM
VM
VM
VM
VM
VM
• COSIGN Controller– ODL controls the core network– Guests are interconnected
through with OVS-terminated overlays
Infrastructure Controller
OverlayController
ServerVirtu
lSwitch
Virtual
Switch
Nova
te
Nova Compu
te
ServerVirtu
lSwitch
Virtual
Switch
Nova
te
Nova Compu
te
ServerCompute Agent
Opto‐
Switch
Opto‐Electronic
Switch
Opto‐
Switch
Opto‐Electronic
Switch
Opto‐Electronic Switch
Optical SwitchOptical SwitchOptical Switch
PacketTunnel Light
Advanced Scenarios – vApp and VDC Use Cases
1) Virtualized application in a public cloud
• User– Application owner who has decided to
become cloud tenant and run his app as a cloud service
• User triggered operations– Create application template (AppTempate) – Deploy application defined by the template– Undeploy a running application– Optional – stop, pause, resume, health
check, repair, …
2) Virtual Data Center• User
– Network-savvy researcher who needs experimentation environment for his work
– Cloud service owner with strict requirements to interconnect (type/quality/topology); can be the cloud provider who offers IaaS over shared infrastructure
• User triggered operations– Create VDC template – Provision VDC resources as defined by the
template– Optional – destroy, health check, modify, …
COSIGN architecture outcome
This presentation is property of the COMBO Consortium and shall not be distributed or reproduced without the formal approval of the Project Board
COMBOCOnvergence of fixed and Mobile BrOadband access/aggregation networks
■ COMBO targets structural and functional convergence of fixed and mobile networks• Integration of backhaul/fronthaul with fixed access / aggregation• Integration of unified fixed and mobile functions in key functional blocks
and functional entities■ COMBO defines Next Generation Point of Presence (NG-POP) as the location
for the subscriber IP edge of fixed, Wi-Fi and mobile networks, in centralized and distributed scenarios
■ COMBO defines Universal Access Gateway (UAG) as a functional entityunifying user access to fixed and mobile networks• UAG data planes include the common subscriber
IP edge for fixed, mobile and Wi-Fi and are located in NG-POPs■ innovation aspects: universal authentication, universal data path management,
inclusion of caching in QoS, tight coupling between wifi and 4G, optimisation of optical infrastructure for 4-5G
■ outcome: full demonstrator including network function in SDN environnementand infrastructure in Lannion April 28
This presentation is property of the COMBO Consortium and shall not be distributed or reproduced without the formal approval of the Project Board
COMBO: Distributed NG-PoP integrated demonstration – a key ongoing work
19
Part 2: Updates of the progress H2020 optical projects
ACINO: Application‐centric Transport Network Orchestration
Objective: differentiate the service offered to each application at each layer of the transport network
Basic use‐case: app‐centric IP/Optical service provisioning• Example: two applications
– Tactile Internet (very‐low latency, high availability and reliability) – VM migration (very high bandwidth)
• Compilation of intents into two different multi‐layer application‐centric services, e.g.:
– Tactile Internet: setup dedicated optical channel, fast IP survivability– VM migr: re‐use existing IP tunnel, slow optical survivability
• Network orchestrator config IP and optical devices– Open‐source development with ONOS framework
• App‐centric and orchestration approaches can be extended to 5G transport (i.e., multiple techs in Xhaul)
FP7 and H2020 Network Technologies Concertation Day
Insert sub‐title here
The iCIRRUS Ethernet fronthaul
• New functional splits between centralised and remote sites, enable lower bit‐rates and statistical multiplexing gains through aggregation/switching units
• iCIRRUS also considers UE functions such as cellular‐assisted D2D and mobile cloud computing in the virtualised RAN
intelligent Converged network consolidating Radio and optical access aRound USer equipment
Insert sub‐title here
• Ethernet mapping• Includes primitives/control traffic
• Delay and delay variation, including through aggregators/switches
• Requirements different for different functional splits• Effects of different, other types of traffic
• Aggregation (synchronous) and low‐cost, high‐bit‐rate transport
• OAM, specifying SLAs over new fronthaul – probing the network for performance parameters and performance optimisation
Challenges (optical fronthaul)
Concertation meeting – Brussels, 1 March 2016
INSPACE (Spatial‐Spectral Flexible Optical Networking: Enabling Solutions for a Simplified and Efficient SDM)
INSPACE solutions provide:• Support of ultra‐high capacity
… by expanding to the space dimension
• Increased flexibility in the allocation of demands … by efficiently switching at both the space
and spectrum domains
Key developments:• Joint spatial switch and fractional joint
switching design solutions• Spatial mux‐demux elements• Advanced transmission and DSP
algorithms• RSpace‐SpectrumA optimization
algorithms• SDN based controller
Additional studies• Experimental demonstration of a fully
controlled data plane with space switching capabilities
• Network cost evaluation and feasibility studies
High port count WSS for joint switching of spatial modesWith a fibre array of 316 (functional) fibres, a 3‐mode(115) spatial spectral high port count WSS has been designed/fabricated
Design and fabrication of FMF (left) and MCF (right) breakout
Initial network performance comparison results among Joint Independent and Fractional switching – Cost studies soon
Network Abstraction Module (NAM)
North‐bound Communications Manager (NCM)
Topology Service (TS)
TDB
South‐bound Protocol Manager (SPM) #1
Optical NodeCP Agent
Optical NodeCP Agent
Optical NodeCP Agent
Client Application
…Client Application
Client Application
TED Manager (TM)
PCE / RSSA Engine (PRE)
Virtualization Engine (VE)
Connection Manager (CM)
CDB
VDB
South‐bound Protocol Manager (SPM) #2
1
512 13
4 6 2 8 3
971514
10 16
1711
Design and implementation of SDN based controller for Spatial ‐Spectral resource allocation
Motivation
ICT 6 – 2014: Smart optical and wireless network technologies
“Address the limitations of current optical transmission technologies”
Reducing the margins improves efficiency and reduces or postpones investments BER issues (soft‐failures) arise and there are no current mechanisms to solve
Optical networks are designed under worst case assumptions using “End‐of‐life margins” Equipment (fibers, amps) ageing Interference (Nonlinear impairments) Maintenance interventions
Motivation
ICT 6 – 2014: Smart optical and wireless network technologies
“Address the lack of dynamic control and management of optical network resources within and across operator's domains for lower cost and more flexible use of resources”
Physical layer monitoring information is barely used in the network lifecycle Planning‐provisioning inefficiencies are never corrected Soft‐failures are treated as black or white
Limited knowledge of the failure cause, and only one control remedy available: protection
Flex‐grid and tunable transceivers provide vast optimization options, but cannot be efficiently configured without physical layer feedback
ORCHESTRA proposes to close the loop between physical layer and the control plane
Interacting with the physical layer enables a more dynamically controlled network that can be used in a more flexible and efficient way
Vision
An optical network has to be observable before it canbecome controllable and be subject to optimization
• ORCHESTRA proposes to close the control loop by enabling physical layer observability
• Observability relies on the coherent receivers that are extended, almost for free, to operate as software defined impairment optical performance monitors (soft‐OPM)
• Physical layer information of single or multiple soft‐OPMs is used to take better optimization decisions
• Re‐acting dynamically on the network to increase its efficiency
1. Develop an advanced DSP‐based physical‐layer multi‐impairment monitoringalgorithm suite
2. Develop a holistic approach to Quality of Transmission (QoT) determination inall network lightpaths using information from distributed software‐definedoptical performance monitors (soft‐OPMs) and advanced correlation algorithms
3. Develop a hierarchical control and monitoring infrastructure providing activeand passive monitoring capabilities with rapid and effective reactions todegradations and failures
4. Develop dynamic optimization procedures for fault management and networkre‐optimization
5. Lower the barriers of resource sharing among operators’ domains through theefficient monitoring of alien lightpaths and accurate physical layer SLAs
6. Demonstrate dynamic and highly efficient flexible network operation enabledby software‐defined optical performance monitoring
ORCHESTRA vision: “An optical network has to be observable before it can become controllable and be subject to optimization”
Objectives
1. R. J. Essiambre, et al., ‘Capacity Limits of Optical Fiber Networks’, J. Lightw. Technol., vol. 28, no.4, pp. 662‐701, 2010.2. A. Ellis, et al., “Approaching the non‐linear Shannon limit” J. Lightw. Technol., vol. 28, no.4, pp. 423‐433, 2010. 3. H. Takara, et al., “1.01‐Pb/s (12 SDM/222 WDM/456 Gb/s) Crosstalk‐managed Transmission with 91.4‐b/s/Hz Aggregate Spectral Efficiency,”
presented at the ECOC’2012, Amsterdam, The Netherlands, Jun. 20124. V.A.J.M. Sleiffer, et al.,“73.7 Tb/s (96 x 3 x 256 Gb/s) mode division multiplexed DP16QAM transmission with inline MMEDFA,” presented at
the ECOC’2012, Amsterdam, The Netherlands, Jun. 2012
1. R. J. Essiambre, et al., ‘Capacity Limits of Optical Fiber Networks’, J. Lightw. Technol., vol. 28, no.4, pp. 662‐701, 2010.2. A. Ellis, et al., “Approaching the non‐linear Shannon limit” J. Lightw. Technol., vol. 28, no.4, pp. 423‐433, 2010. 3. H. Takara, et al., “1.01‐Pb/s (12 SDM/222 WDM/456 Gb/s) Crosstalk‐managed Transmission with 91.4‐b/s/Hz Aggregate Spectral Efficiency,”
presented at the ECOC’2012, Amsterdam, The Netherlands, Jun. 20124. V.A.J.M. Sleiffer, et al.,“73.7 Tb/s (96 x 3 x 256 Gb/s) mode division multiplexed DP16QAM transmission with inline MMEDFA,” presented at
the ECOC’2012, Amsterdam, The Netherlands, Jun. 2012
Fiberbandwidth
Fiberbandwidth
inter‐core coupling major technicalchallenge,
MIMO processing full use of the 11.4 THz
C+L band 1Tb/s over 50 km, using a
12 cores(demonstrated up to
49 cores)
6‐mode 73.7 Tb/s system over 119 km
FYI – Call for Papers. Special Issue on:
Optical Networking for 5G Mobile and Wireless Communications at Journal of Optical
Communications and Networking(JOCN)
Submission Deadline: 31 May 2016
http://www.osapublishing.org/jocn/journal/jocn/feature_announce/on.cfm
Thanks, you’re welcome to the CaON session