Why OTN keeps trucking on
As we approach the end of 2021, we mark 20 years since standards work began on optical transport networking (OTN) in the ITU-T Recommendation G.709. At that time, some declared SONET/SDH dead while others pointed to the reasons why it would remain for years. And remain it did.
Of course, optical transport technology evolved, and TDM standards that had enjoyed a long-life cycle now continue to decline in use. OTN evolved as both a natural TDM replacement and also a strong complement to the evolution of packet transport technologies. As Ethernet protocol speeds increased, OTN container payloads became commonly used through the higher rates, including 100 Gbps (ODU4). OTN became an ideal mechanism to load traffic of any type into a container for transport across the network.
Beyond 100GE, using OTN becomes a bit more complex for many reasons, including present market demand and the technical limitations of creating digital wrappers operating at rates above 100 Gbps. OTUCn is a good solution to these technical challenges. At the same time, 100GE and 400GE IP router interfaces are increasingly common, offering the option of traffic grooming at layer 3, complemented by ROADMs, steering traffic towards these high-capacity routers.
Economics will determine which solution becomes most common over time. Regardless, there’s a continuing need for OTN switching.
The need to share network infrastructure
Where’s this need for OTN switching coming from? The diversity of end-user applications and how they are connected by various service providers.
Communications networks now serve a wide range of users: traditional residential and business service, enterprise data, mobile network data, large enterprise data center interconnect (DCI), power utility data and so on. These users place various demands on the network infrastructure they tend to share.
The network infrastructure could be shared through a single communications service provider, through a wholesale-retail provider model or even, in a very large enterprise, through a shared network model, such as a national research network. Each user wants what they need for their specific application: capacity, latency, availability, security, and other factors form an SLA that the network must meet. OTN helps the service provider ensure this need is met.
The need for isolation: network slicing
More specifically, there’s been much discussion in recent years about network slicing: sharing physical network resources among various services, each with its own logical slice. Sounds easy enough but some services require more rigidly defined slices, isolated from the others.
This isolation can be accomplished through a layer 2 or 3 packet statistical multiplexing technique, resulting in a soft form of isolation, or using a layer 1 circuit-switched connection like OTN, resulting in hard isolation. These two techniques can also be combined to form a mix of isolation as appropriate for the services supported.
In this scenario, OTN is providing the hard isolation of all the MEF layer 2 services contained in the upper ODUflex-1 from the services in the lower ODUflex-2. The two cannot impact each other in any way yet are transported across the network within the larger ODU4 along with all of the determinism, monitoring and resiliency we’ve come to expect from OTN. These concepts are explained in more detail in a recent Nokia white paper.
Here, at the end of 2021, we see OTN as very relevant for many network operators. As operators look for ways to amortize their network costs across a wide set of end-user services and applications, their need to ensure isolation among network slices will require a blend of layer 1, 2 and 3 technologies. OTN is already playing a role.
Share your thoughts on this topic by joining the Twitter discussion with @nokianetworks using #optical #opticalnetworking