Next Gen Optical Network: Scalability, Flexibility and Security
•Download as PPTX, PDF•
2 likes•180 views
Secutity, Flexibility and Scalability are key for next gen optical networks. Through SDN control a full L3 to L1 resource coordination can be achieved.
Recommended
More Related Content
What's hot
What's hot (20)
Similar to Next Gen Optical Network: Scalability, Flexibility and Security
Similar to Next Gen Optical Network: Scalability, Flexibility and Security (20)
Recently uploaded
Recently uploaded (20)
Next Gen Optical Network: Scalability, Flexibility and Security
- 1. Security, Flexibility and Scalability in NG optical networks Francesco Fucelli 09.05.2018
- 2. © 2018 ADVA Optical Networking. All rights reserved. Confidential.22 The digital transformation Source: CISCO VNI report, 2017
- 3. © 2018 ADVA Optical Networking. All rights reserved. Confidential.33 Flexibility and Scalability of Optical Networks
- 4. © 2018 ADVA Optical Networking. All rights reserved. Confidential.44 OSNR performance increasingly critical, flexible spectrum mandatory Unprecedented network flexibility Software defined optics Configure constellation & baud rate for selected data rate • 100-600G • 50G steps Optimize for given channel • based on receive OSNR & spectral shaping due to ROADM cascades Coherent channel bit rate (Gbit/s) Maximumspectralefficiency(bits/s/Hz) 3.6 Tb/s = 3 x 2 carriers x 600Gb/s = 6 x 75 – 100 GHz Spectrum / bandwidth OSNR
- 5. © 2018 ADVA Optical Networking. All rights reserved. Confidential.55 Versatile System Reach – from LH to Metro 4000 km 200G QPSK 1500 km 300G 8-64QAM 800 km 400G 16-64QAM 120 km 600G Metro: fiber capacity Regional: maximum capacity per fiber Long Haul: Maximum distance Optimizing flexible reach vs capacity
- 6. © 2018 ADVA Optical Networking. All rights reserved. Confidential.66 Elastic Channel Spacing – Flex Grid Source: Innovative Future Optical Transport Network Technologies - Toshio Morioka†, Masahiko Jinno, Hidehiko Takara, Hirokazu Kubota
- 7. © 2018 ADVA Optical Networking. All rights reserved. Confidential.77 Control Layer Transport SDN Architecture Application Layer NBI / APIs Cloud Orchestrator Network Orchestrator StorageCompute Parent Controller Domain Controller Domain Controller Domain Controller Infrastructure Layer SBI Domain 1 Domain 2 Domain 3 NE NE NE NE NE NE NE NE NE • Diverse applications • Planning, optimization, services, etc. • Common framework • Multi‐vendor NW SW • Routing, Resiliency • Standard, programmatic interfaces across layers • Open/common device data models Common APIs Standard Interfaces Open Platform
- 8. © 2018 ADVA Optical Networking. All rights reserved. Confidential.88 SDN Cloud Architecture Network Hypervisor Multi-domain / Multi-vendor Transport SDN Control & Orchestration Cloud SDN Controller DC Orchestration Storage Compute Network REST/NETCONF Open APIs IP Domain Controller Cloud Orchestration and Automation Data Center Network Optical Network Programmability and Automation
- 9. © 2018 ADVA Optical Networking. All rights reserved. Confidential.99 SDN use cases examples Improving resource utilization IP Offloading Improving service availability Self provisioning bandwidth
- 10. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1010 SDN use cases examples
- 11. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1111 Innovation through disaggregation Transponders TranspondersTerminal ILA ROADM Terminal Disaggregated APIAPIAPI API API API Open Line System Aggregated Terminal ILA TerminalTransponders Transponders Transponders Transponders End to End System ROADM Partially Disaggregated
- 12. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1212 Optical Open Line System objectives • Financial benefits • Choice, no vendor lock-in • Innovation • Interoperability and flexibility • Software control and coordination
- 13. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1313 Security
- 14. © 2018 ADVA Optical Networking. All rights reserved. Confidential.141414 © 2018 ADVA Optical Networking. All rights reserved. Confidential.14 When you transport information from A to B… How valuable is your information to you? What is the damage in reputation and cost to you IF the information ends up in the wrong hands? in Industry, Finance, Government, Health Care… Easy insurance: Encrypt your data transmission! What about your data streams?
- 15. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1515 Data center environment & security SAN, HPC Switches DWDM Transport Server Storage Server Storage Y-Bridge for service activities Fiber coupling device Apps Apps DWDM Transport SAN, HPC Switches …and what about the fiber connection? There are multiple ways to access fiber
- 16. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1616 High speed encryption modes Cisco Overlay Transport Virtualization (OTV) +82 Bytes MacSec +32 Bytes Cisco TrustSec +40 Bytes Bulk mode (0 Bytes) • Hop-by-hop only • Ethernet only • Overhead creates latency and throughput issues • Point-to-point • Protocol and I/F agnostic (Ethernet, FC, IB, Sonet/SDH) • Integrated solution with lowest latency • Huge overhead • IP VPN services • Cisco Nexus
- 17. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1717 Encryption performance Framesize / Bytes Throughput And why on Layer 1? • Protocol and data rate agnostic • Lowest latency • 100% throughput • Operational simplicity Comparison of maximum throughput
- 18. © 2018 ADVA Optical Networking. All rights reserved. Confidential.1818 Cloud and mobility are radically transforming our connected world. Virtualization and software are keys to differentiated solutions, but hardware will remain strategically important. Brian Protiva, ADVA co-founder and CEO
- 19. Thank you IMPORTANT NOTICE The content of this presentation is strictly confidential. ADVA Optical Networking is the exclusive owner or licensee of the content, material, and information in this presentation. Any reproduction, publication or reprint, in whole or in part, is strictly prohibited. The information in this presentation may not be accurate, complete or up to date, and is provided without warranties or representations of any kind, either express or implied. ADVA Optical Networking shall not be responsible for and disclaims any liability for any loss or damages, including without limitation, direct, indirect, incidental, consequential and special damages, alleged to have been caused by or in connection with using and/or relying on the information contained in this presentation. Copyright © for the entire content of this presentation: ADVA Optical Networking. ffucelli@advaoptical.com
- 20. © 2018 ADVA Optical Networking. All rights reserved. Confidential.2020 One minute online in 2017 >243,000 photos uploaded >65,000 photos uploaded >3.8 million search requests >800,000 files uploaded >87,000 hours of video watched >16,550 video views 60 seconds >400 hours of video uploaded >2 million minutes of calls >156 million emails sent >18,000 matches >25,000 posts on tumblr >120 new accounts >210,000 snaps uploaded >350,000 tweets sent >29 million messages processed Data based on: statista, Digital Economy Compass 2018
- 21. © 2018 ADVA Optical Networking. All rights reserved. Confidential.2323 Delivering cloud-native connectivity Capacity CloudConnect™ terascale DCI networking Performance Service assurance and precise synchronization Intelligence Openness and elasticity by SDN control Efficiency Automation and programmability for the self-driving network Security ConnectGuard™ encryption of data in motion Scalability Ensemble virtualization and NFV hosting
Editor's Notes
- The enormous traffic increase observed during the past few years, as a result of the rapid adoption of broadband connectivity and the emergence of new bandwidth-hungry and Quality of Service (QoS) demanding applications and services, has set new requirements for the underlying network infrastructures. This is true for all parts of the network, either closer to the end users of towards the core network. Capacity improvements in optical networks have till now been addressed mostly by leveraging on the use of the channels available within the ITU-T Wavelength Division Multiplexing (WDM) grid to transmit signals employing fixed modulation formats and bit-rate serial interfaces (fixed transponders). It is evident that there is not much space available for further capacity growth, as the achieved spectral efficiencies are very close to the ultimate limits set by Shannon’s law.
- Current optical transport networks based on Wavelength Division Multiplexing (WDM) technologies, transport the information in wavelengths that are usually assigned in a permanent and static manner in the network planning phase. The wavelength assignment in this type of transmission systems, considering only the optical layer, presents an inefficiency problem due to the coarse granularity of a wavelength, which is to become more significant in the near future as the WDM networks are evolved in wavelength capacity, from 10 to 40, and then to 100 Gb/s and beyond. there is room to increase the efficiency by making the network more flexible during resource allocation, i.e. by introducing a finer transport/switching granularity that would allow for a better adjustment of the allocated capacity according to the user requests.
- Technology developement will not stop at 200G per wavelength. So what coming after that? In principle the two key parameters we have are constellation or in other words how much information does a single symbol that we code carry and baud rate – how many symbol per second can we transmit? With the next generation of technology we will be able to optimize both of these key factors again. You see baud rates going up to 66G and constellation going from 4 level up to 64 levels – 64QAM. With the new interfaces and chipset we will be able to provide many more options and every option can have a different design goal…like for example max. capacity on a single wavelength…or max. length x bandwidth product in a regional network which can be something like 16QAM with 45G baud.
- In today’s networks interconnecting hyper-scale data centres, there is no one size fits all. ICPs and CNPs have radically different demands and need solutions that reflect their specific requirements. Spectral efficiency, price, space and power are four critical building blocks that must be optimized to meet your key objectives. Whether you are in the backbone or in the metro, an optical system shall offers modulation formats to suit your specific needs, ranging from spectrally highly-efficient coherent detection to low-power, low-footprint direct- detect solutions.
- The essence of the elastic optical path network concept is adaptive spectrum allocation to an optical path. One benefit that the elastic optical path network yields is spectral savings achieved by exploiting spectral resources that had not been fully utilized. This results in an increase in network capacity. Let us consider an example in which we transport mixed-rate traffic. First, for client traffic that does not fill the entire capacity of a wavelength, the elastic optical path network provides intermediate bandwidth of an appropriate size, such as 200 Gbit/s. For shorter optical paths, which suffer from less SNR degradation, we use a more spectrally efficient modulation format Finally, combined with elastic channel spacing, where the required minimum guard band is assigned between channels, we can utilize the excess channel spacing. In this way, elastic optical path networks accommodate a wide range of traffic in a highly spectrally efficient manner. combined with elastic channel spacing, where the required minimum guard band is assigned between channels, we can utilize the excess channel spacing.
- We’ve learnt from past experience that central controllers provide excellent resource utilization but such architectures tend to suffer from limited scalability. A hierarchically structured SDN architecture solves this limitation by creating regional and funtional domains. Those domains provide abstracted resource information to a central SDN controller, which is able to handle the multi-layer network end-to-end. Abstraction and hierarchies reduce complexity at the highest control layer. This allows SDN to scale even with central control. Why SDN? 1. Automation speeds up provisioning, minimizes faults and saves OPEX 2. Higher resource utilization achieves more revenue with less investment 3. Fast provisioning meets agility requirements of cloud-centric business services 4. Rapid service innovation gains service providers a competitive edge 5. Open interfaces and abstraction create scalability and manageability Why Transport SDN? 1. Optimizing resource allocation in multi-layer networks 2. Higher service availability with centrally controlled network restoration 3. Improving resource utilization by time sharing 4. Bandwidth slicing and self-provisioned services
- By extending SDN to the optical transport layer, resources of your optical network are virtualized and can be controlled along with switching, routing, storage and compute resources available in your network and data center. Virtualization and SDN control allow all resources to be dynamically allocated under the supervision of a centralized control system. Centralized orchestration of all resources is the prerequisite for optimized end-to-end, multi-layer data flows in your cloud-centric network. New connections can be brought up and torn down again in less than a minute, automatically or at the push of a button. Network Hypervisor enables segments of a virtual network to be controlled independently and provisioned dynamically. A hierarchically structured SDN control architecture is the only way to reduce complexity and enable scalability at the SDN control layer above. Network Hypervisor 1. Hiding complexity: Presents an abstracted view of optical transport network, off-loading the centralized SDN controller 2. Smooth network transformation: Enables combination of traditional operational processes with fully automated, open control of the optical layer 3. Proven solution: Successfully deployed in various proof-of- concept installations in combination with commercial orchestrators and SDN controllers
- Improving Resource Utilization Today, optical networks are operated in a static way with bandwidth being allocated for weeks or even months. There is, however, an increasing need for dynamically provisioned high-bandwidth services in real time. Media streams at sports events or climate data processing are just two scenarios with temporarily high-bandwidth connectivity requirements. Real-time provisioning allows resources to be shared among several users and applications. With SDN, resource utilization is improved and service providers are able to offer very attractive new bandwidth services. IP Offloading and Multi-Layer Network Optimization As IP traffic demand steadily grows, service providers find they need to continuously install additional router and transport network capacity. Intelligent traffic analysis tools can identify traffic flows, which do not need to be forwarded through intermediate routers and can be offloaded from the IP layer into the optical transmission network. This approach enables networks to scale more efficiently. Improving Service Availability Resilience in transport networks is mainly based on pre-provisioned protection mechanisms. Orchestrators have a holistic understanding of all network resources and can restore failed services based on any available resource rather than the limited set of pre-provisioned resources. Failures can be fixed by SDN technology using centrally hosted path computing engines (PCE). Previously, time-consuming manual interaction was needed to restore a network. Offering Self-Provisioned Bandwidth Services SDN technology creates an abstracted view of virtualized network resources. Those resources can represent the whole network or can be limited to just parts, also known as network slices. They can be allocated to a tenant, which might be an end-customer of a service provider or an operational department. Our network hypervisor is the instance that controls the bandwidth slices and provides access to the SDN controller through a standardized interface. The network hypervisor’s multi-tenant capability enables multiple SDN controllers to manage their network resources independently. The isolation of resource- and control-domains creates well separated and protected networks which meet the security requirements of cloud-centric services. Other cases:
- Optical transport equipment can look similar. And with increasing standardization of technology, you could be excused for thinking they are all pretty much the same. However, you need to be aware of other architectures that can have serious long-term ramifications for your business. With disaggregation, each layer of your transport network can evolve independently and be independently optimized for performance and cost. It prevents vendor lock-in by disaggregating the photonic layer f With disaggregation, wavelengths can originate from any DWDM-compliant transmission and routing equipment. It no longer prevents you from taking advantage of different technology life cycles from the terminal layer. The terminal layer sits hierarchically above the photonic layer and is where transmission modulation formats are assigned. With disaggregation, wavelengths can originate from any DWDM-compliant ransmission and routing equipment. It no longer prevents you from taking advantage of different technology lifecycles, which are typically five to seven years for the OLS and two to three years for the terminal layer.
- Open and programmable software architecture Standardized modeling of network elements is the ultimate goal to provide fully interoperable programmable network functions. The drive for open optical architectures also means a move away from closed, per-vendor, per-product management and control systems. With native support for YANG modeling, NETCONF/RESTCONF protocols and open APIs, the OLS can be consistently managed and controlled, thus minimizing your OPEX. Common, open and programmatic northbound interfaces and protocols abstract your network, enabling control by third-party SDN applications, creating a unified networking view and speeding up integration. They are a mandatory component when operating OLSs in your network infrastructure. Benefits of open line systems 1. Investment protection by network disaggregation 2. Each layer can evolve independently and be optimized for performance and cost 3. Standardized modeling of network elements and transceivers 4. Significant savings via footprint and power efficiencies 5. Open to lease spectrum to anyone with any terminal equipment
- Average framesize in today‘s internet traffic is about 300 to 400 bytes, source „NetworkWorld“.
- The enormous traffic increase observed during the past few years, as a result of the rapid adoption of broadband connectivity and the emergence of new bandwidth-hungry and Quality of Service (QoS) demanding applications and services, has set new requirements for the underlying network infrastructures. This is true for all parts of the network, either closer to the end users of towards the core network. Capacity improvements in optical networks have till now been addressed mostly by leveraging on the use of the channels available within the ITU-T Wavelength Division Multiplexing (WDM) grid to transmit signals employing fixed modulation formats and bit-rate serial interfaces (fixed transponders). It is evident that there is not much space available for further capacity growth, as the achieved spectral efficiencies are very close to the ultimate limits set by Shannon’s law.
- SDN use cases, architectures, protocols and data models are standardized and promoted by major standardization bodies and open communities such as Internet Engineering Task Force (IETF), Open Networking Foundation (ONF), Telecom Infra Project (TIP), or CORD.
- Cloud providers -> Why open life systems -> vendor A, B, C