As optical networking technology continues to advance, it often seems that there is a huge emphasis on network speed above all else.

We have seen this regarding the 100G (100 Gbps) connections already in use, with 400G products now reaching scalability. We expect the same with 800G as it moves forward.

Earlier this year, Nokia successfully tested a 1 terabit (Tbps) connection in a 7950 XRS router over a pair of single-mode fibers; the first commercially viable lab trial at this speed. Even more impressive is the 1 petabit (Pbps) connection over 50km that has been successfully tested in a strictly controlled lab setting in the past.

Is this significant? Absolutely. The rate of advancement is staggering. For example, the leap from 1 Gb to 1 Tb is a thousandfold increase, and 1 Pb is an additional 1,000 times faster than 1 Tb. Therefore, a live 1 Gb link in use today is 1 million times slower than the experimental 1 Pb connection mentioned above. For context, the transmission of 1 Pbps is the equivalent of sending 5,000 two-hour HDTV videos in a single second.

Beyond Network Speed

But speed is not the only factor to consider.

In addition to fast speeds, the highest reliability possible is a critical objective for network operators. They typically strive for the “five nines,” or 99.999%, of uptime. What does that mean when applied to real applications? Since there are 8,760 hours in a year, 99.999% uptime means there should be no more than 8 minutes and 45 seconds of network downtime in an entire year — not much room for error. This is why operators have to build layers of redundancy into their networks.

5G service poses more reliability challenges. The promise of 5G is dependent on Gigabit speed and ubiquitous coverage over a low-latency network. As mentioned above, the speed is easily achievable. However, ubiquity and latency present unique challenges. To reach the “five nines” goal and achieve ubiquity, network designs need to evolve from current fixed tower designs to macro and microcells. As a result, 5G services are currently only available from roughly 12% to 17% in the U.S. In addition, without proper design and management, the fronthaul architecture needed to achieve the ubiquity target can add latency.

While speed is a significant factor in network performance, it is part of a more complex equation. Ultimately, the application of quality high-speed interconnections over a diverse fiber infrastructure is required to provide a robust 5G network. Therefore, the need for speed is key. However, it will take excellent planning and execution to resolve the fronthaul connectivity challenges to achieve the desired ubiquity, latency, and “five nines” of reliability.