Bandwidth Virtualization Enables a Programmable Optical Network

Transcription

1 Bandwidth Virtualization Enables a Programmable Optical Network Bandwidth Virtualization TM in Digital Optical Networks provides a scalable, cost-effective and easy to manage service deliver architecture that contains built-in investment protection for emerging technology enhancements. Based on the use of Photonic Integrated Circuit and Digital Optical Network TM technology, Bandwidth Virtualization decouples the service layer from the optical transmission layer, providing operators with software-based flexibility for provisioning new services; unconstrained reconfigurability, and rapid service turn-up. In addition Bandwidth Virtualization alleviates the complexity, time and cost limitations of hardware-based service deployment. Bandwidth Virtualization provides an avenue for carriers to implement a programmable optical network to rapidly and efficiently deploy new and differentiated services and elevates the service offering beyond commoditized pointpoint transport services.

2 Introduction As Internet Protocol (IP) network traffic continues to increase, with growth rates of 75% - 125% per year being a common consensus estimate, network operators are grappling with a range of implications for their networks. These challenges are diverse, and include accommodating higher rate 40Gb/s interfaces from IP core routers, which are expected to further scale to 100Gb/s in the near term; supporting a wide range of service bit rates, from 1Gb/s to 100Gb/s, and protocols, such as SONET/SDH, OTN, Fibre Channel and video; maximizing market competitiveness through speed of service turn-up; all the while needing to ensure operational simplicity for service planning, engineering, deployment and growth. Addressing these challenges with today s technology in many cases imposes significant constraints and costs on network operators. Current optical transport networks are often built using Wavelength Division Multiplexing (WDM) and all-optical Reconfigurable Optical Add/Drop Multiplexer (ROADM) technologies to maximize fiber capacity and maximize service reconfigurability, respectively. Unfortunately, conventional WDM/ROADM systems also suffer from architectural issues that may impose important limitations on service providers. For example, in a typical WDM network provisioning a new service involves connecting the client interface, for example a 10Gb/s port from an IP router, to a transponder that maps that service over a wavelength for transmission across the optical network (see figure 1). In conventional transponder-based WDM networks, each service is thus directly coupled to a specific wavelength, and service activation speed is limited by analog optical engineering and deployment of all the equipment necessary to ensure its successful end-to-end transmission. Figure 1: Conventional WDM systems directly couple service provisioning to wavelength engineering and turn-up. A new service therefore requires a new wavelength deployment, which in turn requires the management of many complexities such as: (a) determining the availability of an end-to-end wavelength without wavelength blocking, or implementing wavelength conversion in case there is blocking; (b) validating optical link engineering rules to ensure appropriate optical parameters such as Optical Signal-to-Noise ratio (OSNR), chromatic dispersion and polarization mode 2

3 dispersion (PMD) are within specifications, and deploying dispersion compensation and/or regenerators in case any of these parameters exceed specifications; (c) managing the planning and engineering complexity associated with multiplexing lowerrate services onto a wavelength using muxponders; (d) planning for the lead times and availability of transponders and filters from the equipment supplier, and; (e) performing often complex system turn-up procedures such as power balancing the testing the wavelength to confirm reliable end-to-end transmission performance. As can be easily inferred from the above list, dealing with all these issues can be complex, costly and time-consuming, and adversely impacts operators business objectives. In addition to slow service activation, the fixed service-wave association of a transponder limits the service flexibility and evolution to new service interfaces and higher data rates. Operators seeking to offer new service products, for example support for ultra-high bandwidth services at 40Gb/s, and in the near future 100Gb/s Ethernet (100GbE), requires them to operate their WDM networks at significantly higher bit rates per wavelength. This in turn requires the WDM systems to accommodate, and thus compensate for, the many optical impairments that scale, often exponentially, with bit rate. For network operators this has meant that their existing WDM transport systems, often originally designed for transmission at 10Gb/s per wavelength, must either be overbuilt with new WDM systems, or upgraded to now support transmission at 40Gb/s per wavelength. Network overbuilds with new WDM systems can be costly and time-consuming. Upgrading existing networks to support higher-rate transmission, while more cost-effective short-term, may require extensive optical link re-engineering, and often imposes sacrifices in the reach of WDM systems. All such issues are expected to be even more challenging for transmission at 100Gb/s, and will need to be re-visited again once 100GbE services become reality. All of these challenges accumulate, thereby increasing operational complexity for carriers due to the many truck-rolls required for manual interventions, requiring higher capital outlay for network overbuilds or upgrades, significantly lengthened provisioning cycles for deployment of new capacity, and creating inflexibility for support of new service types. This may limit an operator s ability to easily and speedily support new orders from existing or potential customers, and thereby use speed of service as a competitive differentiator. Building a Programmable Optical Network Ideally, network operators should be able to deploy additional capacity and support new service types quickly and easily, with a minimum of manual intervention, hardware deployment, and engineering complexity. In such a case, service deployment would be more of a matter of software-enabled network reconfiguration, thereby ushering in the concept of a programmable optical network (see figure 2). Such a programmable optical network is possible through the use of a network architecture called Bandwidth Virtualization. Bandwidth Virtualization allows end-to-end service provisioning to be de-coupled from the complex link-by-link optical wavelength engineering of WDM systems, thereby allowing a variety of services, ranging from sub-wavelength data rates to superwavelength data rates, to be quickly and simply provisioned and transported over a common WDM network operating at a data rate optimized for the lowest network cost. 3

4 Figure 2: Key elements of a programmable optical network include a pool of WDM bandwidth, integrated digital switching, multi-service client interfaces and software intelligence. Here sub-wavelength service refers to service data rates that are a fraction of the nominal data rate of a wavelength in the WDM line, whereas a super-wavelength service has a data rate higher than the wavelength data rate on the WDM line. In practice, electrical or digital multiplexing is used to either map multiple sub-wavelength services into a common wavelength, or a super-wavelength service across from multiple wavelengths which are bonded to provide the required bandwidth to support end-to-end transmission. To be realized, Bandwidth Virtualization requires the convergence of several key architectural elements within the WDM network, including: Service-ready capacity between nodes that is cost-optimized, scalable, pre-tested and ready for service activation. Large-scale photonic integration provides an ideal platform for this form of consolidated, cost-effective WDM capacity deployment. Integrated bandwidth management that consolidates high-capacity WDM transport with digital bandwidth management to enable remote, reconfigurable mapping of any service to any available line capacity. Multi-service/protocol client interfaces that are independent and de-coupled from the WDM line optics, thereby enabling any sub-wavelength or super-wavelength service to be mapped into the available WDM line capacity Software Intelligence using a GMPLS control plane to allow automated, remote and reconfigurable end-to-end service provisioning and routing, without the need for manual interventions or truck-rolls at intermediate sites. 4

5 Fundamentally, Bandwidth Virtualization enables services, of any type or bit rate, to be delivered using a pool of WDM line side bandwidth, rather than be coupled to a specific wavelength and line rate as in a conventional WDM network. Figure 3 illustrates how Bandwidth Virtualization enables a de-coupling of service provisioning from optical link engineering to allow any service to be mapped to an available pool of WDM line side bandwidth. Figure 3: WDM with Bandwidth Virtualization de-couples service provisioning from the underlying WDM link capacity, enabling rapid turn-up of any service. Such a network is characterized by a simple plug-and-play any service, anywhere approach to optical networking driven by software-initiated reconfiguration rather than hardware-driven engineering, installation and manual intervention. This in turn provides a range of benefits to network operators, including: New service support: Bandwidth Virtualization enables new high bandwidth services such as 40Gb/s and 100Gb/s, to be transported over the same line system using a common, costeffective bit rate on the WDM line, and without re-engineering the deployed WDM line system. By contrast, transponder based WDM systems require the WDM line to be engineered in response to the highest supported service rate, and thereby incur any resulting cost or complexity imposed by optical link engineering requirements across all services. Fast service deployment: Bandwidth Virtualization, by leveraging software intelligence and eliminating the dependency between service deployment and optical network re-engineering, allows new services to be quickly provisioned over existing infrastructure simply through the connection of a client interface at each end of the service path. This allows operators to quickly respond to new bandwidth requests, provide market-leading support for new services, and use these as competitive differentiators in a market-place that is otherwise characterized by price-based competition and commoditization. Operational ease: Bandwidth Virtualization allows rapid and simple service activation as client interfaces are added only at the network end-points regardless of service type, without the need to upgrade network resources or implement truck-rolls to handle wavelength blocking or link re-engineering, thereby allowing new services to be provisioned quickly over service-ready transmission capacity. 5

6 Capital Efficiency: Bandwidth Virtualization allows operators to select the most costefficient WDM line capacity and bit rate independent of service type, and operate over existing infrastructures. This avoids the need to re-engineer the optical line (ie: extra regens, complex chromatic or PMD compensation, wavelength conversion, back-back muxponder, etc), improves capital efficiency by avoiding line overbuilds when provisioning new highbandwidth services, and enables better wave fill and avoids stranded capacity for subwavelength services. The above benefits of Bandwidth Virtualization address the limitations of conventional transponder-based WDM systems that restrict operator s ability to simply, efficiently and quickly offer new services. A summary of how a WDM network with Bandwidth Virtualization compares to a conventional WDM network is shown in Table 1 below. Service Reconfigurability WDM Line System Economics Service Activation Bandwidth Efficiency Table 1 Transponder-based WDM Service data rate change requires re-engineering WDM line or new line system Line system tied to highest service rate Expensive for high data rate services Service activation closely tied to link engineering Slow and cumbersome, and possibly costly Stranded bandwidth for subwavelength services delivered over muxponders WDM with Bandwidth Virtualization No WDM line re-engineering needed WDM capacity costoptimized independent of service mix Service activation decoupled from link engineering Rapid, just-in-time model No stranded bandwidth - Services can tap available bandwidth pool; Fractional services flexibly multiplexed into WDM pipe Implementing Bandwidth Virtualization Infinera enables the realization of a programmable optical network using Bandwidth Virtualization through the convergence of several key and innovative technologies that span optical components, system design and network architecture. These include Photonic Integrated Circuits (PICs), Digital Optical Networks, and software protocols. First, the recent development and wide-spread adoption of large-scale Photonic Integrated Circuits (PICs) provides a key technological tool that allows the concept of a programmable optical network based on Bandwidth Virtualization to be practically realized. Large-scale PICs integrate dozens to hundreds of optical components such as lasers, modulators, detectors, attenuators, multiplexers/de-multiplexers and optical amplifiers into a single device, and are conceptually very similar to an electronic IC. PICs operating with 10 wavelengths at 10Gb/s per wavelength and having a total WDM capacity of 100Gb/s per device have been widely deployed in optical transport networks since 2004, while 6

7 recent R&D efforts have demonstrated PICs capable of total aggregate data rates up to 1.6Tb/s per device. This demonstrates the potential for large-scale photonic integration to enable even greater capacity and functional integration in the future. Photonic integration thus plays a key role in enabling Bandwidth Virtualization by consolidating multiple wavelength channels into a WDM system on a chip having 100Gb/s of aggregate capacity. This provides the required pool of line-side WDM bandwidth that can be economically pre-deployed, ready for service, and over which any service can be mapped. In addition to providing order-of-magnitude improvements in the size, power consumption and reliability, PICs allow system designers to economically perform Optical-Electrical-Optical (OEO) conversion at any node in the network, thereby permitting the use of digital electronics, rather than analog optics, to perform feature-rich and value-added functions such as reconfigurable subwavelength digital add/drop, multiplexing, and bandwidth management, along with digital protection and performance monitoring and support for new service features such as fast digital protection, GMPLS restoration, and Layer 1 optical virtual private networks (O-VPNs). This enables the concept of a Digital Optical Network that combines the functionality and benefits of integrated bandwidth management with the cost-effective bandwidth scalability of a WDM network. Integrated sub-wavelength digital bandwidth management implemented in the Infinera Digital Optical Network means that the PIC-based WDM line bandwidth can be deployed today in increments of 100Gb/s, and then virtualized and managed with a granularity of 2.5Gb/s (ODU1). This allows any service, from sub-wavelength Gigabit Ethernet (GbE) or 2.5Gb/s SONET/SDH, to wavelength services at 10Gb/s, and super-wavelength services such as 40Gb/s and 100GbE, to be flexibly and reconfigurably assigned to the virtual bandwidth pool. Figure 4: Schematic representation of the Digital Virtual Concatenation protocol. Complementing the Infinera Digital Optical Network is the development and use of a Digital Virtual Concatenation TM (DVC) protocol which performs the mapping of any service onto the available pool of PIC-based WDM bandwidth. An example of how the DVC protocol used in the Infinera DTN to transport a 40Gb/s service is shown in figure 4. In this case a 40Gb/s client signal is connected to a tributary adapter module (TAM) on the DTN, and inter-leaved using DVC into four 7

8 10Gb/s payloads (4 x ODU2). The 4 x ODU2 are converted into 4 x OTU2V by adding a high-gain FEC, and transmitted on four out of the ten wavelengths originating from the 100Gb/s PIC. At the receive end, the 4 x ODU2 are de-mapped from the 4 x OTU2V and re-assembled into the OC- 768/STM-256 signal. A similar approach would be used for transmission of either sub-wavelength services, in which case multiple services would be mapped into a single 10Gb/s wavelength, or for transmission of other super-wavelength services such as 100GbE, which would be mapped over all ten wavelengths on the PIC. The Infinera Digital Optical Network consolidates all the above technological and architectural elements into a single network platform that combines service-ready WDM capacity and integrated digital bandwidth management with embedded software intelligence to automate end-to-end service provisioning. This in turn allows operators to realize in practice a new programmable optical networking paradigm that speeds service delivery while significantly reducing operating complexity. Transporting 40Gb/s & 100Gb/s Services Using Bandwidth Virtualization Bandwidth Virtualization implemented over an Infinera Digital Optical Network has been used to successfully transmit both 40Gb/s and 100Gb/s services over long-distance networks spanning thousands of kilometers in several field trials, as described below. In a first field trial demonstration of the use of Bandwidth Virtualization, a 40Gb/s service was transmitted over a record distance of 8,477km on a trans-oceanic network. The trans-oceanic network comprised terrestrial links from Frankfurt, Germany connecting via Paris, France and London, England to a submarine cable system connecting the UK to the USA, with a final terrestrial link connecting from the US-based cable landing site to New York City. In this case Bandwidth Virtualization was used to map a 40Gb/s interface from a Juniper T640 router in Frankfurt across four wavelengths transmitted from a 100Gb/s PIC on an Infinera Digital Optical Network, which was then re-assembled as a 40Gb/s service in New York City, thereby providing demonstration of the first 40Gb/s IP link across the Atlantic. Bandwidth Virtualization was also used to enable the first-ever transmission of pre-standard 100GbE service over a wide-area network (WAN) as part of a field trial which took place in November 2006 during the SC06 International Conference on High-Performance Computing, Networking, Storage and Analysis. In this case, the 100GbE signal was mapped over ten wavelengths transmitted from a PIC-based Digital Optical Network, and transmitted from Tampa, Florida to Houston, Texas and back again across fiber network provided by Level(3) Communications, over a total distance of 4000 km. The demonstration represented the first time a pre-standard 100GbE signal was successfully transmitted through a live production network across the WAN, and was important in showing that 100 Gigabit Ethernet technology is not only viable but can be implemented in existing WDM networks engineered to support with 10Gb/s data rates per wavelength. Conclusion The need to provide rapid service delivery of a wide range of services from 1Gb/s to 100Gb/s, is driving operators to look at solutions that will meet these needs while being capital and operationally efficient. Bandwidth Virtualization provides a new network architecture, enabled by the use of large-scale PICs and Digital Optical Networks, which allows the end-to-end provisioning and management of high-bandwidth services to be de-coupled from the underlying optical 8

9 transmission engineering. Thus Bandwidth Virtualization helps network operators transform provisioning of new services from what used to be a hardware and engineering-driven operation into a software-driven activity, resulting in greater speed and flexibility, lower costs, and effectively enables the practical implementation of a programmable optical layer. 9

10 Infinera Corporation 169 Java Drive Sunnyvale, CA USA Telephone: Fax: Have a question about Infinera s products or services? Please contact us via the addresses below. Americas: Asia & Pacific Rim: Europe, Middle East, and Africa: General sales-am@infinera.com sales-apac@infinera.com sales-emea@infinera.com info@infinera.com Specifications subject to change without notice. Document Number: Copyright 2007 Infinera Corporation. All rights reserved. Infinera, Infinera DTN, IQ, Bandwidth Virtualization, Digital Virtual Concatenation and Infinera Digital Optical Network are trademarks of Infinera Corporation. 10