Light Speed Ahead - Optical Networking Trends in 2024

Overview

Optical Networking customers and situations are ever-growing and changing, but the advancements in 2024 are astounding, especially with the rise of AI. We here at WWT see a wide variety and scale of customers from the public sector enterprise, and service providers alike, all with a focus on their optical transport layer for varying reasons. Some are looking to increase their existing customer base by expanding fiber to neighboring counties. It could also be a large city looking to expand its capacity for police, fire, and city services. Maybe a large financial institution is looking to implement a highly redundant, encrypted next-generation DWDM solution with many terabits worth of capacity between data center locations. The point is that we can pick up on certain commonalities or trends with each engagement. These trends highlight the dynamic nature of optical networking in 2024. Operators and businesses must stay informed and adapt strategies to navigate this evolving landscape effectively.

Bandwidth explosion

AI is increasing bandwidth usage - Albeit a great tool, AI also drives more bandwidth to the network. Through the processing of running large language models for uploading and downloading results to generative AI creating large high-resolution videos and images. Additionally, AI-powered Internet of Things (IoT) devices can produce vast amounts of data that need transmissions to analyze AI algorithms. With this increase in bandwidth, operators need to understand the network impact and plan accordingly.
The rise of 5G, with its need for high-speed data transfer and near-instant response times, has significantly boosted optical networks. Because older technologies can't handle the vast amounts of data that 5G requires, fiber optical networking is the only real option. As more organizations use 5G, they'll need to upgrade their network infrastructure, including everything from the towers that connect to devices (fronthaul) to the lines that carry data long distances (backhaul). This means a lot of investment in fiber optic cables and the equipment that goes with them. Since 5G also requires more cell sites, that means laying even more fiber optic cable deeper into the network. Overall, 5G is a driving factor for optical networks creeping closer to the edge and a significant driver for increased bandwidth and capacity going forward.
XGS-PON and Optical Transport: XGS-PON fundamentally relies on optical transport networks to deliver high-bandwidth services to end users. The widespread adoption of XGS-PON will drive a growing need for more robust and scalable optical backbones to support the increased traffic. This pushes for more significant investment in optical infrastructure, including fiber networks and transmission equipment.
DOCSIS 4.0 and Optical Transport: While DOCSIS 4.0 primarily works over existing cable (HFC) infrastructure, backhaul connections to those networks are increasingly handled by optical transport. As DOCSIS 4.0 rolls out and bandwidth needs increase, the optical links connecting cable headends and aggregation points must be upgraded for capacity and speed.

AI and machine learning integration

AI is driving a significant impact and overall change in how operators have historically managed, delivered, and protected optical transport networks. The following are some contributing factors to those changes:

Network optimization and resource allocation - An operator can analyze traffic patterns using AI. With this, we can make real-time adjustments to bandwidth allocation within the layer 1, layer 2, and layer 3 domains. This will help ensure more efficient network bandwidth use while reducing bottlenecks and congestion.
Predictive Maintenance: AI can analyze historical data and network performance to predict potential failures before they occur. This allows for preventive maintenance, minimizing downtime, and ensuring network reliability.
Enhanced Security: AI can analyze network activity and identify potential security threats. This allows for proactive measures to be taken to protect against cyberattacks.
Improved Quality of Service (QoS): By prioritizing and optimizing data delivery, AI can ensure critical data, like video conferencing or medical applications, is delivered smoothly even when sharing the network with other types of traffic.

From automating routine network adjustments to predicting and preventing outages, AI and machine learning are becoming game-changers in optical networks. These powerful technologies pave the way for more intelligent traffic management, allowing data to flow more securely, efficiently, and reliably, ultimately boosting overall network performance and efficiency.

Open Optical Networking systems

Vendor lock-in has been a challenge in optical networking systems in the past. With Converged optical & packet systems adoption, operators are getting a feel for a more open architecture. By extracting traditional optical transport systems' transponded and muxponded layers, many operators are getting their first experience with a more open architecture. This allows operators to entertain separate routing vendors to house a pluggable coherent optic directly within the router, and it also allows operators to pick and choose which OEM is responsible for transporting those higher bandwidth coherent wavelengths on a per-span basis. This strategy drives the competitive nature of the network's core, reducing vendor lock-in.

Maturing beyond 400G to 1.2 and 1.6T

800G coherent pluggable optics are becoming available and will see increased deployments in 2024. This offers higher capacity per wavelength for data centers and metro/long-haul networks.

Beyond 800G: Still in its early stages, standards, and discussions are underway for this next-generation technology that promises higher capacity for the future. At the Optical Fiber Conference (OFC), I recently experienced multiple vendors introducing 1.2T pluggable optics showcasing impressive bandwidth and distance performance.

Ongoing focus on sustainability

Operators are increasingly seeking ways to reduce their environmental impact. A clear way this can be achieved is through the convergence of optical and packet layers. With this strategy, we have clearly seen a decrease in network devices and total power consumption.

Additionally, multi-layer domain controllers can be leveraged to help design and manage the overall strategy of converging the optical and packet layers, making it easier for the operator to manage a multi-layer - multi-vendor solution.

Summary

2024 marks a pivotal year for optical networking, driven by the insatiable demand for bandwidth. 5G rollout and the rise of AI are creating a surge in data traffic, pushing the industry towards robust fiber optic networks. Technologies like XGS-PON and DOCSIS 4.0 further emphasize the need for robust optical transport systems. Open optical networking systems are a welcome response, offering operators flexibility and potentially lower costs. This year also sees the arrival of 800G coherent pluggable optics, significantly boosting data transmission capacity. The industry is future-proofing itself with discussions on even higher capacities already underway. Sustainability is another key focus, with the convergence of optical and packet layers aiming to streamline networks and minimize their environmental footprint. Finally, AI and machine learning transform how networks are managed, delivering network optimization, predictive maintenance, and enhanced security. In essence, cutting-edge technologies enable optical networks to meet the ever-growing demand for bandwidth while becoming more efficient, reliable, and sustainable.