The rapid development of video, 5G, and cloud services has led the annual compound growth rate of global backbone network traffic to reach over 40%. In China, for example, in the last ten years, backbone network traffic has maintained an average annual growth rate of 47%. This growth puts huge pressure on backbone networks.
As a result, an important goal for transmission network development is quickly resolving these traffic problems by using the lowest possible investment amount. Research shows that traffic on backbone networks is unevenly distributed, and that it is heavy in some areas and light in others. Can an optimally designed network architecture be used to resolve these problems and reduce required investment amounts?
3D-Mesh backbone network is the solution to network congestion, and is the direction towards which existing backbone networks are evolving. The concept of 3D-Mesh network construction is not new, and has been widely applied to highways. Both highways and backbone networks are types of bearer networks. The primary difference between them is that the highways carry vehicle flows and backbone networks carry information flows. Overpasses are used to avoid highway congestion. Multi-plane construction at congested roads resolves traffic congestion, shortening driving time. In addition, driving safety is enhanced by the adoption of multiple lanes, minimizing investment amounts without affecting existing traffic. Similarly, congestion, latency, and reliability are also problems for backbone networks.
Overpasses have been used for decades. They are now widely deployed in major cities around the world and are common in large- and medium-sized cities. 3D-Mesh backbone networks emerged only a few years ago, but are developing rapidly. The widespread acceptance of this solution is clear from its deployment in China, India, and Europe, in national backbone networks and in metro core networks.
The biggest advantage of this solution over the traditional solution for capacity expansion is that it maximizes network resource utilization. Traditional backbone networks are ring networks. When capacity is insufficient, a new ring network is deployed on the old ring network. Although this relieves the traffic on the most congested links, it also unnecessarily expands the capacity of links with idle bandwidth. As a result, the resource utilization of the entire network becomes unbalanced and the required investment amount remains high. 3D-Mesh backbone network construction allows funds to be used where they are needed the most, and new links are established only for congested links. This achieves the best protection of investment and the fastest and most optimal ROI.
In addition to architecture changes, there are two key points to consider in the construction of 3D-Mesh backbone networks, as follows:
First is the collaboration of optical transport network (OTN) and the optical layer. The optical layer is used for the pass-through of large-granularity traffic and is most effective in reducing electrical-layer costs. OTN is used for the seamless switching and grooming of a large number of small-granularity services. The OTN + optical solution is similar to the separation of vehicles into speed-specific lanes on a highway, where lanes for small and large vehicles are separated. In this way, the resources of different channels can be used in a refined and efficient manner, maximizing the utilization and efficiency.
Second is the combination of ASON (automatic switching optical network) and SDN (software-defined networking). 3D-Mesh architecture is the basis for forming more protection routes of key links. ASON greatly improves network reliability. When SDN is enabled, traffic can be intelligently groomed and routes can be selected on demand by latency, further improving the saleable value of networks.
At last, backbone networks evolve towards the full mesh architecture. As described in this article, backbone networks also evolve towards 3D-Mesh networking. The combination of these two planes will transform the evolution of backbone networks into one that brings optimal ROI.