February 22, 2019

What’s Wrong with Transport Routing?

By Gary Croke, Sr. Director of Marketing and Strategy

“Transport routing,” a term that may be new to many, simply refers to routing application in the transport (or backhaul) portion of the network. These transport networks can be based on fiber, microwave, millimeter wave or a combination of all these physical layer technologies. This short blog explores the issues with routing in the transport network and discusses a new solution to address those key issues.

Too expensive

The cost of routing, especially in transport networks, has been too high for too long. A few manufacturers have been demanding premiums partly because of lack of choices available and have been getting away with upsell pricing models that charge thousands of dollars for plug-in cards, software licenses, and support. It is common for a transport router to cost between $10,000 and $20,000, or even more in some applications, depending on configurations and capabilities – which in today’s business environment is too much.

What’s more, because most routers used in transport applications are built for multiple use cases or deployment scenarios, they support a very wide range of RFCs and features – most of which are not needed in the transport network. Only about 15% of features on these routers are ever deployed in the average transport domain thus users are ultimately paying manufacturers for features they will never use or need.

The vision of NFV/white box to lower cost is a good one however when it comes to transport routing, these technologies are not progressing fast enough. Something is needed in next 3 years to address these cost issues while these technologies come to market.

Optimized for Fiber Only

From a mobile network perspective, over 50% of the worlds cell sites are connected with microwave/millimeter wave. Transport networks in other verticals (public safety, utility, oil/gas, etc) tend to be built even more heavily on wireless technology. In these wireless transport scenarios, operators tend to deploy a microwave indoor unit alongside the transport router – which means extra cost to buy, deploy, and maintain the additional box. Today’s transport routers seem to be designed with only fiber networks in mind.

Unprepared for the future

Many of today’s routers that are built for access applications are not prepared for the capacity demands of future networks. The emergence of 5G as well as high definition video and security requirements will place demands on transport router infrastructure that many deployed devices are not prepared to handle. Lower cost routers deployed in the transport network can often be limited to the low 10’s of Gbps in throughput without option for high speed interfaces.

Beyond capacity, the evolution to SDN will be a step-wise migration and much of the installed base is not ready for this migration, lacking native Netconf/YANG interfaces and data models to simplify multi-vendor networks.

Something new is needed to fix these fundamental issues with transport routing.