As bandwidth-intensive services such as video- and voice-over-IP emerge throughout the communications landscape, carriers are seeking to increase both the capacity and flexibility of their networks in the most economical and efficient way. Network operators need convergence solutions that enable them to provide the maximum number of services, but with minimal change to their installed infrastructure.
For example, traditional fixed-line operators are integrating residential broadband services, enterprise data connectivity and video services into their existing networks. Cable operators are also adding voice-over-IP (VoIP) and commercial services to their video and cable TV offerings. At the same time, both operators and enterprises are embarking on a huge migration from traditional services based on SDH/SONET, ATM and Frame Relay to Ethernet-based services that will carry voice, video and data traffic.
This migration is challenging operators to change their network architecture in order to create a single converged network, but at the same time to minimize the risk to their existing infrastructures, their revenue-generating services and, most importantly, their customers. Operators also want to introduce Ethernet-based services without stranding their investments in traditional networks, while also maintaining quality-of-service (QoS) and speeding up the time-to-market of new services.
So far, however, no technology has provided the functionality needed to gain widespread acceptance as a suitable convergence platform, especially when capital investment and operating costs are also taken into account. But one universal network standard - the Optical Transport Network (OTN) standard - is driving forward the introduction of new flexible technologies in the telecoms industry. OTN offers support of all existing services, including traditional voice and private lines, while providing operators with the flexibility to seamlessly roll out new Ethernet/IP services. As such, OTN will play a critical role in the networks of the future. OTN provides a crucial stepping stone in the evolution of transport networks, and promises to simplify convergence by making the optical network friendly to all service types. As a result, OTN has emerged as a must-have technology over the past year, particularly for operators looking to converge and deploy next-generation networks.
Convergence made easy
OTN enables convergence by providing an operational model for network administration, performance monitoring and fault isolation that is identical to the one used for SDH/SONET services. That comes as no surprise because it was created by the same industry standards body, the International Telecommunications Union (ITU).
However, while traditional SDH/SONET-based infrastructures based on strict time-division multiplexing (TDM) schemes do not enable more efficient use of bandwidth as data traffic increases, OTN improves transport network performance and facilitates the evolution to higher network bandwidth. OTN also offers unprecedented transparency to support all traffic types, including Ethernet-based storage services, as well as traditional services.
With OTN technology, multiple networks and services can be combined seamlessly onto a common, future-ready infrastructure. Because OTN was designed to be transparent to service types, all services are given individual treatment and any native functionality and performance are preserved without compromising the integrity of the underlying protocols. In short, OTN provides the features and durability needed to allow carriers to support and manage their networks effectively.
The best of both worlds
OTN is an ITU standard dating back to 1998. Also known as a "digital wrapper" technology - because of its ability to wrap any service into a digital optical container - OTN was unified from competing standards developed by both the ITU and the American National Standards Institute. According to ITU standards, an OTN is composed of a set of optical network elements that are connected by optical fibre links, providing transport, multiplexing, routing, management, supervision and survivability of optical channels, and the carrying of client signals.
Unlike SDH transport standards, OTN is a globally accepted standard that defines the management of next-generation optical networks. It is described by a host of ITU recommendations, starting at G.872, which illustrates the network architecture on OTN, and including G.709, which focuses on structure, interfaces and mapping. These recommendations provide equipment manufacturers with the tools needed to make interoperable products that allow carriers to build and manage ultrahigh-capacity optical networks, and to achieve end-to-end connectivity between optical transport elements in a global network.
Another distinguishing characteristic of OTN is its ability to provision transport for any digital signal, independent of any client-specific aspects. As such, according to the general functional modelling described in the ITU's recommendation G.805, the OTN boundary is placed in such a way that it includes the server-specific processes but not the client-specific processes.
In simpler terms, OTN is designed to manage the transmission of multiple wavelengths over a single fibre. It consists of a header in which overhead bytes are carried, a payload section and a trailer that performs forward-error correction (FEC) (figure 1). The payload section of OTN allows for all existing network protocols to be mapped without disruption, while FEC not only corrects errors but also enables carriers to provide different service-level agreements (SLAs) to their end-users. By minimizing the errors in the network, FEC is able to extend the reach of the fibre and allow faster transmission rates.
Prior to OTN, a common approach to networking involved connecting IP routers to multiservice provisioning platforms (MSPPs). But the MSPP can break the end-to-end communications link - which compromises topology discovery techniques that map out the network layout and terminate performance management data signals - when it removes the client SDH/SONET overhead bytes and terminates the data communications channel (DCC). These two elements are essential for maintaining client-management connectivity and inter-nodal communications across the network as they are used to determine network paths and link states. Loss of this information also prevents proper end-to-end management and makes it difficult to troubleshoot both customer and network problems.
OTN overcomes this problem by virtue of its transparency, which makes it an ideal technology for operators and enterprises to build converged networks. As transport networks become increasingly complex, MSPPs and routers lose this transparency and are not robust enough to support operators as they roll out new services. OTN also supports its own separate overhead for performance monitoring and fault signalling, as well as a general communications channel (GCC) for remote management, software downloads and other control functions.
Due to the fact that OTN specifications provide a robust management overhead analogous to SDH/SONET, network operators have no need to sacrifice the ability to manage at both the payload and service levels. In fact, an OTN payload can fully encapsulate an SDH/SONET frame without terminating the SONET DCC, which means that remote add-drop multiplexers can continue to be managed in the same way and topology discovery still works between customer equipment.
Service support
However, OTN does a lot more than transport SDH/SONET in a transparent way. It is also the only transport layer standard in the industry that can carry a fully managed 10 Gigabit Ethernet circuit from IP Ethernet switches and routers at full bandwidth, including the proprietary overhead associated with the many vendor-specific implementations.
Since OTN is highly effective in supporting asynchronous data services, it also supports the various speeds of Fibre Channel, ESCON and FICON that on their own are unable to deliver the level of performance monitoring and fault isolation needed to achieve a high-quality service. What's more, OTN brings those capabilities to asynchronous services without sacrificing the qualities that make those services attractive in the first place, such as low cost and ease of installation (figure 2).
OTN's inherent flexibility is enabled by its ability to extend transparency into the timing plane, which allows both synchronous and asynchronous signal types to be mixed on a common wavelength. Moreover, synchronous services with different clock sources can be transported side-by-side, which is not possible in a SDH/SONET network.
An OTN network is made up of several networking layers (figure 3). At the top of the stack is the service layer, which represents end-user services such as Gigabit Ethernet, SDH/SONET, Fibre Channel or any other protocol. Asynchronous services are also passed through a Generic Framing Procedure (GFP) mapper. The next layer, the optical channel payload virtual container (OPVC), maps the service into a uniform format and is the only layer that needs to change in order to support a new service type.
The output from the OPVC is mapped into a timeslot by the optical channel payload tributary unit (OPTU), which also performs timing adaptations to unify the clocking. All of these timeslots are held in the optical channel payload unit (OPU), while the optical channel data unit (ODU) provides the path-level transport functions of the OPU. Finally, the optical transport unit (OTU) provides the section-level overhead for the ODU and adds the general communications channel (GCC0) bytes. Altogether, the OTU is mapped into a wavelength or a WDM system by the physical layer.
OTN has a hierarchy just like SDH and SONET. An optical channel runs between anything that maps a service into an OTU signal. As defined by the ITU, OTU1 is the 2.7 Gbit/s signal designed to transport STM-16/OC-48 signals. OTU2 is the 10.7 Gbit/s signal designed to transport an STM-64, OC-192 or the physical layer of a 10 Gigabit Ethernet wide-area network, as well as Fibre Channel at 10 Gbit/s.
Unlike SDH/SONET, OTU2 is able to carry the physical layer of a 10 Gigabit Ethernet local-area network, even when IP/Ethernet switches and routers are operated at full line rate. This includes the support of proprietary overhead schemes that are implemented by some router vendors. Moreover, the OTU3 (40 Gbit/s) standard will easily extend these networks to support higher bandwidth requirements.
In order to support lower speed services, extensions to the standard make it possible to aggregate multiple sub-wavelength (ie. below 2.5 Gbit/s) services onto a single OTU wavelength with service level add-drop capability and bit-stream transparency. This approach dramatically lowers costs and also simplifies access and metro networks, where bandwidth demand is more granular and unpredictable as new packet-based services run side-by-side with legacy services.
OTN is emerging as the common optical-backbone network of the future. Its most valuable characteristics include the transparent support of legacy and packet services, its ability to extend SDH/SONET-like operations and fault management capabilities to other data protocols and its seamless support for 10 Gigabit Ethernet services. OTN also enables more efficient provisioning, switching and multiplexing of high-bandwidth services, such as IPTV, leading to improved wavelength usage.
OTN's ability to efficiently provision transport and manage Ethernet and other non-SDH/SONET data services makes it the ideal technology for carriers to build a single converged infrastructure that supports all types of applications and services. As the rapid migration to Ethernet-based services continues, network operators will increasingly turn to the capabilities offered by OTN.
Indeed, network convergence will only become more challenging as operators commit to building out next-generation networks. By deploying an OTN standard, operators will be able to continue offering value-added services, such as VoIP and video-on-demand, and be ahead of the game in an increasingly competitive industry.
• This article originally appeared in FibreSystems Europe in association with LIGHTWAVE Europe May 2006 p20.