Published on February 16, 2021
Are you ready for 400G technology?
We’re making greater use of data networks than ever. Think about it: how many of these network-dependent technologies do you see being used every day?
- Internet of Things (IoT)
- cloud-based computing
- EDGE computing
- virtual reality
- ultra-high-definition (4K and 8K) video streaming
- artificial intelligence
The answer: probably more than you think.
And these are just technologies we have today. Who knows what innovations are coming to drive network traffic up even more sharply? Plus, the 5G era is dawning, so expect this network usage trend to snowball.
That’s why network and data center operators are working with equipment manufacturers to validate new 400G Ethernet-capable network elements in their labs. The success of this lab work will help operators meet service level agreements with customers who expect faster, larger-capacity data flows while also lowering cost and power consumption per bit (and increasing their own profitability).
All of this means coming to grips with 400G.
How fast is 400G? A quick technical deep dive
The quick answer: the actual line rate of a 400G Ethernet link is 425 Gbit/s. Yes, you read that right.
The IEEE P802.3bs standard defines the technical specifications of 400G Ethernet data rate. The standard also establishes a mandatory forward error correction (FEC) mechanism to not only detect errors in the transmission but also correct them. The additional bits required for the FEC mechanism are included in the 425Gbit/s line rate.
What even makes 400Gbit/s transmission possible? It’s the adoption of 4-level pulse amplitude modulation (PAM4). Using PAM4, operators implement 8 lanes of 50G or 4 lanes of 100G for different form factors (i.e., OSFP and QSFP-DD). This novel optical transceiver architecture supports transmission of up to 400 Gbit/s Ethernet data over either multiwavelength or parallel fibers.
However, multilevel modulation creates higher analysis complexity, so we need advanced tools such as PAM4 eye diagram histograms and pre-emphasis and equalization capabilities to manage this complexity.
400G network deployment: how far have we come?
Network operators are currently testing new network devices such as switches and routers that can support 400G client interfaces as well as other 400G transport equipment. They are also upgrading their infrastructure. But this is not happening as quickly as it might.
400G network deployment: what are the challenges?
In a word: several.
Optical transceivers are the most critical element of this rollout. Their development maturity, availability, cost and reliability are not yet where operators want them to be.
Furthermore, many different types of transceivers are typically used in different environments over varying distances. QSFP56-DD and OSFP transceivers are being used in 400G networks in carrier labs and data centers while coherent optics are being deployed in data center interconnect and metro applications, and soon rolling out to submarine networks.
Meanwhile, in large data centers, port densification continues. Operations teams are increasing the capacity of aggregation links by upgrading them from 10GE to 25GE or 50GE, and 100GE to 200GE or 400GE links.
Network testing and monitoring have also become more important than ever. After all, the performance of 400G links is vital for the success of new 5G-enabled services. Latency measurements remain a key industry benchmark as operators assess the quality of experience offered by service providers. Optimal quality of experience also depends on data throughput, frame loss, latency variation (jitter) and other parameters.
With all the challenges inherent to 400G links, plug-and-play network deployment is just not realistic. Network operators will need to invest to make sure their networks efficiently support all services promised by new and emerging technologies. Investments must cover the physical (layer 1: fibers, connectors, connections, etc.) and higher level [Layer 2 (Data Link-Ethernet) and Layer 3 (Network – IP)] components.
400G network testing requirements: what are they?
Requirements differ based on context.
Testing in the field
Field technicians and engineers must upgrade their testing procedures to include elements required for 400G transmission. On the physical testing side, technicians must be able to assess connectors, interfaces and fibers for the higher performance standards required at 400G.
400G transceivers must be validated using criteria like:
- Bit error rate assessment pre- and post-FEC
- power consumption
- temperature monitoring
- physical evaluation of the signal using a histogram assessment or pre-emphasis/equalization
Testing in carrier labs
Carriers are receiving new components from network equipment manufacturers, but they must thoroughly evaluate and stress these components in their own labs before using them in a 400G network. So carriers test and benchmark using standardized test procedures based on, for instance, RFC2455 or bit error rate measurement.
The end customer is concerned about the performance of the service they receive. The quality of that service is determined by factors such as throughput, latency, jitter and frame loss. Service-level agreements define these parameters based on recognized, standardized procedures such as ITU-T Y.1564.
In order to meet SLA objectives, operators must validate not only Layer 2 Ethernet frame performance, but also Layer 3 IPv4 and IPv6 packet efficiency.
Once again: are you ready for 400G technology?
A colossal digital transformation is set to occur in the next 5 years, and the deployment of 400G technology is one of its most important enablers.
That’s why massive investment in solutions for 400G (and beyond) have already started. A solid, reliable network infrastructure based on 400G links is being deployed, and this will enable a virtually limitless number of network connections.
Get the right testing, monitoring and analysis tools
EXFO’s FTBx-88460 solution offers all the test applications and advanced tools you need to validate 400G networks. They are future-proof in that they support not only current optical interfaces but are also designed to handle new optical technologies as they become available.
EXFO’s FTBx-88460 solution offers the unique Open Transceiver System (OTS) that supports the QSFP-DD and OSFP 400G interfaces. It is also ready for newer technologies such as coherent ZR and ZR+ transceivers.
Diligent and reliable network infrastructure upgrades remain a perpetual challenge. EXFO’s solutions enable operators to meet that challenge successfully for years to come.