Converged Optical & Packet Network
Owned by Christine Fierro + 2
Convergence of the optical & packet networks promises a future of simplified operations, powerful capabilities, and reduced costs.
Here's how WWT can help.
Start with a Briefing
Not sure where to start? Our experts can provide you with a briefing overview that touches on everything you need to know about Converged Optical & Packet Networking. We'll provide an overview of the major components, an assessment of available solutions, and example use-cases.
Request a Workshop
When you're ready to take the next step towards a Converged Optical & Packet architecture, our experts can facilitate a workshop that will assess your current network & investments, collaboratively determine your best path forward, and develop action items that will get you on your way.
New to IP routing or want to explore how Segment Routing works? Get hands-on experience with our routed transport network lab. From the ground up, this lab walks you through the steps to build a Segment Routing based transport network. Already know the basics? Check out our advanced Segment Routing labs found here.
Make it Happen
Implementing a converged optical & packet architecture can be a difficult task. Let WWT help accelerate the time to validate, deploy and migrate your services. WWT has the expertise and capabilities to deliver a complete end to end optical & packet architecture and can ease the path towards migration.
What is it? Why use it? How does it work?
Converged Optical & Packet architectures consolidate the function of the optical network into the routed network. Thanks to innovations in router ASIC design, routing software and optical transceiver technology it is now possible to provide both layers of the network within one platform.
The Converged Optical & Packet architecture removes the ROADM layer from the optical network and places the DWDM optics directly into the router. Then, utilizes relatively inexpensive, port dense core routers powered by Segment Routing technology, to manage all traffic paths — including pass-through traffic, where packets simply come into the router and are forwarded back out towards their destination.
Converged Optical & Packet architectures reduce the overall cost of the network, minimize points of failure, and simplifies operations. Please read on for a deeper understanding of why and how you would deploy a Converged Optical & Packet architecture.
For a more in-depth look at the technology behind the Converged architecture, please check out these articles:
The Converged Optical & Packet architecture is compelling for multiple reasons, including network cost reductions, network architecture simplification, and operational efficiency.
Cost reduction is realized in multiple ways:
- Less hardware to purchase.
- Lower support contract cost due to less hardware needed.
- Reduced space and power.
- Lower operation cost through a simplified network.
The Converged Optical & Packet architecture place's the transponder directly in the router via pluggable DWDM optics. Additionally, it moves the roles of path selection, path protection and traffic flow into the routed network. Since the router assumes most of the traditional optical network functions, the number and complexity of optical devices is significantly reduced.
DWDM still plays a role but in the form of low-cost, passive, point to point systems between adjacent router sites. Moreover, since todays pluggable DWDM optics support 400G data rates, you could initially deploy without a DWDM system. Then, as site-to-site traffic demand increases, a point-to-point DWDM system is easily added. It is a true flexible consumption model, only spend as your needs grow.
Space and power are reduced significantly as well because we no longer need a powered DWDM chassis/line card system. There are one rack-unit sized routers that can support this architecture, while advancements in ASIC and optic's technology has decreased overall power requirements dramatically.
Operationally, there is less hardware to manage and support; less hardware reduces the failure domain and provides for easier troubleshooting. A single management platform simplifies provisioning and monitoring. It enables your network engineering workforce to operate much more efficiently because the optical and routing domain are now one.
At its most basic layer, converged Optical & Packet would include routers and pluggable DWDM optics. In this scenario we would connect the routers with high-performance optics and then leverage Segment Routing. With routers leveraging inherent protocols such as Segment Routing, we are then able to switch the traffic to and through each router within the domain from source to destination. With traditional optical networking an operator would utilize ROADM capabilities at layer 1. This allows the system to send traffic to and from endpoints at an optical signal. With Segment Routing the same result is achieved; however, the signals are simply converted to an electrical signal as traffic switches through the router. Different strategies, same result. In the past this strategy was not realistic as the cost consideration was overwhelming. Today with 400G DCO/ZR/ZR+, and soon 800G optical interfaces, a converged Optical & Packet solution is an affordable strategy to consolidate multiple layers of the network and leverage efficiencies throughout.
1. Router to Router
A router-to-router solution is the simplest form of convergence. We have a high bandwidth optic DWDM optics connecting two or more router sites together. All the sub-rate client traffic (1/10/40/100/400G) then leverages Segment Routing to send the traffic to and from the source to destination. It is assumed that in this scenario the pluggable DWDM optic is riding on the outside plant pair of fiber. This strategy although simple and cost effective does not allow for much expansion on the outside plant fiber capacity. If there are no more strands of fiber to use between router end points, the network capacity would then be capped at the speed and rate of the pluggable DWDM optic. Additionally, there would be distance limitations. The best pluggable DWDM optics will be limited to approximately 80km. If additional distance is needed, amplification will need to be added to the solution (See scenario 3 – Router to Amplified).
2. Router to Filter
Stacking on from our router-to-router scenario we have the router to filter solution. As we learned with a router-to-router scenario we have a limitation in the ability to expand the capacity of the outside plant fiber. If an operator has the need for more than the allotted pluggable DWDM optic bandwidth, then an external multiplexing filter would allow for this functionality. External filters allow for many DWDM wavelengths to be multiplexed onto the outside plant pair of fiber. These filters can range from 4 channels up to 96, allowing for massive fiber capacity expansion. Filter solutions are generally small in rack unit footprint, ranging anywhere from one to two rack units, and are typically passive non powered devices. For these reasons, an external filter is an ideal solution for an operator to expand the capacity of their network over limited outside plant fiber or plan for future expansion.
3. Router to Amplified (can include filter)
Continuing from the router to filter scenario we have the router to amplified solution. In any optical transport situation light loss is a concern, as eventually light dissipates within a fiber core. Additionally, there is inherent loss when light goes from our router, out of the pluggable DWDM optic into the external filter (multiplexer), effectively shortening the potential distance of the link. Amplification is the answer to this problem. An external amplifier will take in the composite signal and boost the outgoing light to the next site. Since the light is amplified the effective distance (or reach) can be extended. External amplifiers are generally small in rack unit size and an effective solution to extend the distance between router end points.
4. Router to Add/Drop, ROADM, or CDC
Network operators have large complex networks with many services. In some cases, the operator may not have a requirement for a router to drop at every location around the network. This can be due to network design or service level requirements. In these situations, a hybrid design with converged routers and ROADM functionality are an ideal strategy. With ROADM passthrough capabilities the network only drops converged routers where needed. This is ideal for operators who are looking to transport long haul layer 1 services. In the diagram below we have a four (4) node ring. The operator has a need for a converged Optical & Packet solution only at sites 2, and 4. With ROADM passthrough capabilities we can send the router traffic through sites 1, and 3 at an optical level. This save the operator cost in not being forced to place devices and optics in locations that don't fit the operational goals. Infrastructures that typically support this type of architecture have long distances and a need to deliver separate layer 1 services that don't require direct interface with the converged solution.
Hardware and Software Components
Multiple factors, including distance between sites, capacity requirements, and traffic patterns influence the hardware strategy for a converged optical & packet architecture. The examples below illustrate common hardware architectures used to address various network conditions and requirements.
Router to Router: Short distance between locations, relatively low bandwidth requirements (400G or less)
Router to Filter: Short distance between locations, higher bandwidth requirements (greater than 400G)
Router to Amplified: Long distances between locations, higher bandwidth requirements (greater than 400G)
Router to ROADM: Long distances between locations, higher bandwidth requirements and more complex traffic patterns
Transit Agency Modernizes Core Network to Enhance Rider Experience
Is Hop-by-Hop the Future of Optical Transport Networking?
TEC37 Carrier Networking E04: Routing and Optical Technologies, Converged
TEC37 Carrier Networking Series E05: Routing and Optical Technologies, Converged - Part II
Frequently Asked Questions
A: No, the benefits of converged architectures exist for both greenfield as well as brownfield implementations. In fact, legacy architectures that do not rely on DWDM or packet technologies, such as SONET, are great candidates for migration. A converged architecture is a good fit anywhere that you require high & scalable bandwidth with rich packet services.
A: Not necessarily. If you already have a DWDM optical network today, then a good amount of that investment can likely be reused within a new converged architecture, leaving you with optimal Layer 1 capabilities—the best of both worlds. If you already have a packet network, then WWT experts can work with you to determine which investments are beneficial to a converged architecture. Reach out to us for more information, or schedule a briefing or workshop to explore this further.
A: No, a converged architecture can utilize an existing ROADM system allowing the operator to leverage the same reconfigurable wavelengths as before.
A: A converged architecture can leverage amplifiers and ROADM technologies to achieve greater distances.
A: We understand that the journey to a converged solution can sound daunting, however WWT is here to help. We have labs specifically created to take the complication out of layer 3 transport and are geared towards optical engineers. Additionally, there are management software solutions to provide the user the same look and feel of a traditional graphical user interface (GUI). Finally, WWT offers installation, lifecycle support, and custom training classes which are intended to get our customers "over the hump" in regards to understanding this type of solution.
A: Depending on the size of your network and the configuration, implementing a converged optical and packet solution can vary in complexity. WWT can help accelerate the time to validate, deploy and migrate your services to a new converged architecture.
A: A converged architecture is inherently scalable. Ideally, day one, an operator would implement a WDM multiplexer in which multiple wavelengths can be utilized. Moreover, routers can be added via router to router uplinks or new routers can be added in conjunction with an additional DWDM wavelength. DWDM capacity is allocated on a site to site basis, providing for a flexible and scalable growth model.